diff --git a/.vscode/launch.json b/.vscode/launch.json index 954e5aabe..b54767e3e 100644 --- a/.vscode/launch.json +++ b/.vscode/launch.json @@ -5,7 +5,7 @@ "version": "0.2.0", "configurations": [ { - "name": "Next.js: debug full stack", + "name": "Next.js: debug full stack (Chrome)", "type": "node", "request": "launch", "program": "${workspaceFolder}/node_modules/next/dist/bin/next", @@ -20,6 +20,22 @@ }, "cwd": "${workspaceFolder}" }, + { + "name": "Next.js: debug full stack (Edge)", + "type": "node", + "request": "launch", + "program": "${workspaceFolder}/node_modules/next/dist/bin/next", + "runtimeArgs": ["--inspect"], + "skipFiles": ["/**"], + "serverReadyAction": { + "action": "debugWithEdge", + "killOnServerStop": true, + "pattern": "- Local:.+(https?://.+)", + "uriFormat": "%s", + "webRoot": "${workspaceFolder}" + }, + "cwd": "${workspaceFolder}" + }, { "type": "node", "request": "launch", diff --git a/app/(apis)/llm/chat/completions/analytics.ts b/app/(apis)/llm/chat/completions/analytics.ts new file mode 100644 index 000000000..14e92e0f2 --- /dev/null +++ b/app/(apis)/llm/chat/completions/analytics.ts @@ -0,0 +1,40 @@ +import { after } from 'next/server' +import config from '@/payload.config' +import { getPayload } from 'payload' + +type EventType = 'llm.tool-use' | 'llm.completion' + +// Must at least have a user, everything else is optional. +type Metadata = { + user: string, + [key: string]: any +} + +export function createDataPoint(type: EventType, metadata: Metadata, data: Record) { + // Send analytics after the request is done. + // This is the newer version of waitUntil. + after( + async () => { + const payload = await getPayload({ config }) + + try { + await payload.create({ + collection: 'events', + data: { + type, + source: 'llm.do', + data, + metadata, + // Do More Work tenant. + tenant: '67eff7d61cb630b09c9de598' + } + }) + } catch (error) { + console.error( + '[ANALYTICS] Error creating log', + error + ) + } + } + ) +} \ No newline at end of file diff --git a/app/(apis)/llm/chat/completions/route.ts b/app/(apis)/llm/chat/completions/route.ts index 000406525..901c4b80d 100644 --- a/app/(apis)/llm/chat/completions/route.ts +++ b/app/(apis)/llm/chat/completions/route.ts @@ -1,6 +1,15 @@ + import { auth } from '@/auth' -import { streamText, generateObject, streamObject, generateText, resolveConfig, createLLMProvider } from '@/pkgs/ai-providers/src' -import { CoreMessage, jsonSchema, createDataStreamResponse, tool } from 'ai' +import { createLLMProvider, generateObject, generateText, streamObject, streamText } from '@/pkgs/ai-providers/src' +import { getModel } from '@/pkgs/language-models' +import { CoreMessage, createDataStreamResponse, jsonSchema, tool } from 'ai' +import { convertIncomingSchema } from './schema' +import { schemas } from './schemas' +import { createDataPoint } from './analytics' + +import { alterSchemaForOpenAI } from '@/pkgs/ai-providers/src/providers/openai' + +import { convertJSONSchemaToOpenAPISchema } from '@/pkgs/ai-providers/src/providers/google' export const maxDuration = 600 export const dynamic = 'force-dynamic' @@ -16,11 +25,29 @@ type OpenAICompatibleRequest = { stream?: boolean; response_format?: any; tools?: any; +} + +type LLMCompatibleRequest = { /* * If true, the response will be streamed as a data stream response * This is used by the useChat hook in the client */ useChat?: boolean; + /** + * Object used to represent mixins for the getModel function. + * Allows you to control the model via JS rather than a string. + */ + modelOptions?: { + providerPriorities?: ('cost' | 'throughput' | 'latency')[], + tools?: string[], + } +} + +const ErrorResponse = (message: string, code: number = 400) => { + return Response.json({ + success: false, + error: message + }, { status: code }) } export async function POST(req: Request) { @@ -36,18 +63,17 @@ export async function POST(req: Request) { if (apiKey) { // Remove the Bearer prefix - apiKey = apiKey.split(' ')[1] - .replace('sk-do-', 'sk-or-') + apiKey = apiKey.split(' ')[1].replace('sk-do-', 'sk-or-') // Make sure the API key is valid const identifyUser = async (offset: number = 0) => { - const res = await fetch(`https://openrouter.ai/api/v1/keys?offset=${ offset }`, { + const res = await fetch(`https://openrouter.ai/api/v1/keys?offset=${offset}`, { headers: { - 'Authorization': `Bearer ${ process.env.OPENROUTER_PROVISIONING_KEY }` - } + Authorization: `Bearer ${process.env.OPENROUTER_PROVISIONING_KEY}`, + }, }) - .then(x => x.json()) - .then(x => x.data) + .then((x) => x.json()) + .then((x) => x.data) if (res.length === 0) { return null @@ -61,14 +87,12 @@ export async function POST(req: Request) { if (!keyMatch) { // Loop again until we find a match or we've ran out - return await identifyUser( - offset + res.length - ) + return await identifyUser(offset + res.length) } return { authenticationType: 'apiKey', - email: keyMatch.name + email: keyMatch.name, } } @@ -81,16 +105,14 @@ export async function POST(req: Request) { session = { user: { id: user.email, - email: user.email + email: user.email, }, - expires: new Date(Date.now() + 1000 * 60 * 60 * 24 * 30).toISOString() + expires: new Date(Date.now() + 1000 * 60 * 60 * 24 * 30).toISOString(), } } // Support both GET and POST requests - let postData: OpenAICompatibleRequest = { - model: 'openai/gpt-4.1', - } + let postData: Partial = {} try { postData = await req.json() @@ -98,68 +120,183 @@ export async function POST(req: Request) { // Mixin query string into the post data postData = { ...postData, - ...Object.fromEntries(qs.entries()) + ...Object.fromEntries(qs.entries()), } - } catch (error) { // Convert the query string into an object postData = { - model: 'openai/gpt-4.1', - ...Object.fromEntries(qs.entries()) + ...Object.fromEntries(qs.entries()), } } - const { - model = 'openai/gpt-4.1', - messages, - prompt, - system, - temperature, - max_tokens, - top_p, - stream, + const { model, prompt, system, stream, tools: userTools, ...rest } = postData as OpenAICompatibleRequest + + // Overwritable variables + let { response_format, - tools: userTools + messages } = postData as OpenAICompatibleRequest + // llm.do superset OpenAI standard + const { + modelOptions, + useChat + } = postData as LLMCompatibleRequest + if (!prompt && !messages) { - return new Response('No prompt or messages provided', { status: 400 }) + return ErrorResponse('No prompt or messages provided') } + // Fix messages to be in the VerceL AI SDK format. + // Most notibly, we need to fix files coming in as OpenAI compatible + // and translate it for the user. + + if (messages) { + // Swap tool role for assistant role. + // Tool responses from user tools comes in as tool role, which is not compatible with + // the AI SDK. + // @ts-expect-error - Read above + messages = messages.map((message) => { + if (message.role === 'tool') { + return { + ...message, + role: 'assistant' + } + } + + return message + }) + + if (!messages) { + return ErrorResponse('No messages provided') + } + + // Check to see if this message is a file, and in the open ai format. + // First get the indexes of file messages + const fileMessageIndexes: number[] = [] + + for (const [index, message] of messages.entries()) { + // @ts-expect-error - An error is expected because its not the right type, but we're fixing it. + if (message.content[0]?.type === 'file' && (message.content[0] as any).file.file_data) { + fileMessageIndexes.push(index) + } + } + + let tempMessages: CoreMessage[] = [] + + for (const index of Array.from({ length: messages.length }, (_, i) => i)) { + const message = messages[index] + // const file = message.content[0] as unknown as { + // type: 'file' + // file: { + // filename: string + // file_data: string + // } + // } + + // Translate the file to the VerceL AI SDK format + if (fileMessageIndexes.includes(index)) { + tempMessages.push({ + role: message.role as 'user', + content: [ + { + type: 'file', + // @ts-expect-error - Read above + data: message.content[0].file.file_data, + mimeType: 'application/pdf', + }, + ], + }) + } else { + tempMessages.push(message) + } + } + + messages = tempMessages + } + + console.log(postData) + const llm = createLLMProvider({ - baseURL: 'https://gateway.ai.cloudflare.com/v1/b6641681fe423910342b9ffa1364c76d/ai-functions/openrouter', + baseURL: 'https://gateway.ai.cloudflare.com/v1/b6641681fe423910342b9ffa1364c76d/ai-testing/openrouter', apiKey: apiKey, headers: { 'HTTP-Referer': 'http://workflows.do', - 'X-Title': 'Workflows.do' - } + 'X-Title': 'Workflows.do', + }, }) - const llmModel = llm(model) + // @ts-expect-error - Object is coming from the client + const llmModel = llm(model, modelOptions) + + const { parsed: parsedModel, ...modelData } = getModel(model, modelOptions ?? {}) + + if (!modelData.slug) { + return Response.json({ + success: false, + type: 'MODEL_NOT_FOUND', + error: `Model ${ model } not found` + }, { status: 404 }) + } + + const fixSchema = (schema: any) => { + switch (modelData.author) { + case 'openai': + return alterSchemaForOpenAI(schema) + case 'google': + return convertJSONSchemaToOpenAPISchema(schema) + default: + return schema + } + } + + if (response_format) { + response_format = fixSchema(convertIncomingSchema(response_format)) + } + + console.log( + parsedModel + ) + + // Only run this if we dont already have a response_format + if (!response_format && parsedModel.outputSchema) { + if (parsedModel.outputSchema !== 'JSON') { + // If its an internal schema, like W2, then use that instead of Schema.org. + if (schemas[parsedModel.outputSchema]) { + response_format = schemas[parsedModel.outputSchema] + } else { + const schema = await fetch(`https://cdn.jsdelivr.net/gh/charlestati/schema-org-json-schemas/schemas/${parsedModel.outputSchema}.schema.json`).then((x) => x.json()) + + response_format = jsonSchema(fixSchema(schema)) + } + } + } // Fix user tools to be able to be used by our system const tools: Record = {} - for (const [name, toolData] of Object.entries((userTools ?? {}) as Record)) { + for (const [_name, toolData] of Object.entries((userTools ?? {}) as Record)) { tools[toolData.function.name] = tool({ type: 'function', description: toolData.function.description, - parameters: jsonSchema(toolData.function.parameters) + parameters: jsonSchema(fixSchema(toolData.function.parameters)), }) } const openAiResponse = (result: any) => { return Response.json({ - id: result.id, + id: result.id || `msg-${ Math.random().toString(36).substring(2, 15) }`, object: 'llm.completion', created: Date.now(), model, + provider: modelData.provider, + parsed: parsedModel, choices: [ { message: { content: result.text, role: 'assistant', - tool_calls: result.toolCalls.map((toolCall: any) => ({ + tool_calls: (result.toolCalls || []).map((toolCall: any) => ({ index: 0, id: toolCall.id, type: 'function', @@ -173,78 +310,237 @@ export async function POST(req: Request) { finish_reason: 'stop' } ], - usage: { + usage: result.usage ? { prompt_tokens: result.usage.prompt_tokens, completion_tokens: result.usage.completion_tokens, total_tokens: result.usage.total_tokens - } + } : undefined }) } - if (stream) { - if (response_format) { - const result = await streamObject({ - model: llmModel, - system, - messages, - prompt, - user: session?.user.email || '', - // @ts-expect-error - Type error to be fixed. - schema: jsonSchema(response_format), - onError({ error }) { - console.error(error); // your error logging logic here + const openAIStreamableResponse = (textStream: any) => { + // We need to replicate this response format: + // data: {"id": "chatcmpl-81ac59df-6615-4967-9462-a0d4bcb002dd", "model": "llama3.2-3b-it-q6", "created": 1733773199, "object": "chat.completion.chunk", "choices": [{"index": 0, "delta": {"content": " any"}, "logprobs": null, "finish_reason": null}]} + // data: {"id":"gen-1746649993-JcAnN9JWfGSdco3C13ad","provider":"Google AI Studio","model":"google/gemini-2.0-flash-lite-001","object":"chat.completion.chunk","created":1746649993,"choices":[{"index":0,"delta":{"role":"assistant","content":"Okay"},"finish_reason":null,"native_finish_reason":null,"logprobs":null}]} + + return createDataStreamResponse({ + execute: async (dataStream) => { + const id = `chatcmpl-${Math.random().toString(36).substring(2, 15)}` + + for await (const chunk of textStream) { + const openAICompatibleChunk = { + id, + model, + created: Date.now(), + object: 'chat.completion.chunk', + choices: [ + { + index: 0, + delta: { + content: chunk, + }, + logprobs: null, + finish_reason: null, + }, + ], + } + + // @ts-expect-error - We're using this for a different type than what it was built for + dataStream.write(`data: ${JSON.stringify(openAICompatibleChunk)}\n\n`) } - }) - return result.toTextStreamResponse() - } else { - - const result = await streamText({ - model: llmModel, - system, - messages, - prompt, + // Send the stop reason chunk + dataStream.write( + // @ts-expect-error - We're using this for a different type than what it was built for + `data: ${JSON.stringify({ + id: `chatcmpl-${Math.random().toString(36).substring(2, 15)}`, + model, + created: Date.now(), + object: 'chat.completion.chunk', + choices: [], + finish_reason: 'stop', + })}\n`, + ) + }, + }) + } + + console.log('Using', stream ? 'streaming' : 'non-streaming', 'with', response_format ? 'response_format' : 'no response_format') + + console.log('Using tools', tools) + + const onTool = async (tool: string, args: any, result: object) => { + createDataPoint( + 'llm.tool-use', + { user: session?.user.email || '', - maxSteps: 50 - }) - - // We need to support both streaming and useChat use cases. - if (postData.useChat) { - return result.toDataStreamResponse() + tool, + args + }, + result + ) + } + + try { + if (stream) { + if (response_format || parsedModel.outputSchema === 'JSON') { + let generateObjectError: (error: string) => void = () => {} + const generateObjectErrorPromise = new Promise((resolve) => { + generateObjectError = resolve + }) + + const result = await streamObject({ + ...rest, + model: llmModel, + modelOptions, + system, + messages, + prompt, + user: session?.user.email || '', + schema: parsedModel.outputSchema === 'JSON' ? undefined : jsonSchema(response_format), + // @ts-expect-error - Type error to be fixed. + output: parsedModel.outputSchema === 'JSON' ? 'no-schema' : undefined, + onError({ error }) { + console.error(error) // your error logging logic here + generateObjectError(JSON.stringify(error)) + }, + onTool + }) + + if (useChat) { + return createDataStreamResponse({ + execute: async (dataStream) => { + const textStream = result.textStream + + // When this promise resolves, it will have an error. + // We need to pass this error to the client so it can be displayed. + generateObjectErrorPromise.then((error) => { + dataStream.write(`0:"${error?.replaceAll('"', '\\"')}"\n`) + }) + + // Simulate a message id + dataStream.write(`f:{"messageId":"msg-${Math.random().toString(36).substring(2, 15)}"}\n`) + + const formatChunk = (chunk: string) => { + // Make sure that the chunk can be parsed as JSON + return chunk.replaceAll('"', '\\"').replaceAll('\n', '\\n') + } + + dataStream.write(`0:"\`\`\`json\\n"\n`) + + for await (const chunk of textStream) { + dataStream.write(`0:"${formatChunk(chunk)}"\n`) + } + + dataStream.write(`0:"\\n\`\`\`"\n`) + + // // Fixes usagePromise not being exposed via the types + // const usage = (result as any).usagePromise.status.value as { + // promptTokens: number + // completionTokens: number + // totalTokens: number + // } + + //dataStream.write(`e:{"finishReason":"stop","usage":{"promptTokens":2217,"completionTokens":70},"isContinued":false}\n`) + //dataStream.write(`d:{"finishReason":"stop","usage":{"promptTokens":2367,"completionTokens":89}}\n`) + }, + }) + } else { + const response = result.toTextStreamResponse() + + response.headers.set('Content-Type', 'application/json; charset=utf-8') + + return response + } + } else { + const result = await streamText({ + ...rest, + model: llmModel, + modelOptions, + system, + messages, + prompt, + user: session?.user.email || '', + maxSteps: 50, + onTool + }) + + // We need to support both streaming and useChat use cases. + if (useChat) { + return result.toDataStreamResponse() + } else { + return openAIStreamableResponse(result.textStream) + } + } + } else { + if (response_format) { + const schema = jsonSchema(response_format) + + const result = await generateObject({ + ...rest, + model: llmModel, + modelOptions, + system, + messages, + prompt, + user: session?.user.email || '', + mode: 'json', + // @ts-expect-error - Type error to be fixed. + schema, + onTool + }) + + // @ts-expect-error - TS doesnt like us adding random properties to the result. + // But this is needed to trick our openAI response API into thinking its text. + result.text = JSON.stringify(result.object) + + return openAiResponse(result) } else { - return result.toTextStreamResponse() + const result = await generateText({ + ...rest, + model: llmModel, + modelOptions, + system, + messages, + prompt, + user: session?.user.email || '', + maxSteps: 10, + tools, + onTool + }) + + return openAiResponse(result) } } - } else { - if (response_format) { - const result = await generateObject({ - model: llmModel, - system, - messages, - prompt, - user: session?.user.email || '', - mode: 'json', - // @ts-expect-error - Type error to be fixed. - schema: jsonSchema(response_format) - }) + } catch (e) { - return Response.json({ - data: result.object - }) - } else { - const result = await generateText({ - model: llmModel, - system, - messages, - prompt, - user: session?.user.email || '', - maxSteps: 10, - tools - }) + console.error(e) + + switch ((e as { type: string }).type) { + case 'AI_PROVIDERS_TOOLS_REDIRECT': + const error = e as { + type: 'AI_PROVIDERS_TOOLS_REDIRECT' + apps: string[] + connectionRequests: { + app: string + redirectUrl: string + }[] + } - return openAiResponse(result) + return Response.json({ + success: false, + type: 'TOOLS_REDIRECT', + error: `To continue with this request, please authorize the following apps: ${error.apps.join(', ')}`, + connectionRequests: error.connectionRequests + }, { status: 400 }) + default: + return Response.json({ + success: false, + type: 'INTERNAL_SERVER_ERROR', + error: 'An error occurred while processing your request. This has been logged and will be investigated. Please try again later.' + }, { status: 500 }) } } } -export const GET = POST \ No newline at end of file +export const GET = POST diff --git a/app/(apis)/llm/chat/completions/schema.ts b/app/(apis)/llm/chat/completions/schema.ts new file mode 100644 index 000000000..1613c1548 --- /dev/null +++ b/app/(apis)/llm/chat/completions/schema.ts @@ -0,0 +1,11 @@ +import { jsonSchema } from 'ai' + +export function convertIncomingSchema(schema: any) { + // A function to convert any schema to a json schema + // Supports OpenAI, llm.do, and normal JSON Schema. + + if (schema.type === 'json_schema' && schema.json_schema) { + // OpenAI compatible schema. + return schema.json_schema.schema + } +} \ No newline at end of file diff --git a/app/(apis)/llm/chat/completions/schemas/index.ts b/app/(apis)/llm/chat/completions/schemas/index.ts new file mode 100644 index 000000000..98a147869 --- /dev/null +++ b/app/(apis)/llm/chat/completions/schemas/index.ts @@ -0,0 +1,6 @@ +import { W2 } from './w2.schema' + +// Wrap all our custom schemas together into one object. +export const schemas = { + W2 +} as Record diff --git a/app/(apis)/llm/chat/completions/schemas/w2.schema.ts b/app/(apis)/llm/chat/completions/schemas/w2.schema.ts new file mode 100644 index 000000000..b9f6d275e --- /dev/null +++ b/app/(apis)/llm/chat/completions/schemas/w2.schema.ts @@ -0,0 +1,88 @@ +export const W2 = { + "type": "object", + "properties": { + "employeeFirstName": { + "type": "string" + }, + "employeeLastName": { + "type": "string" + }, + "employeeAddress": { + "type": "string" + }, + "employeeCity": { + "type": "string" + }, + "employeeState": { + "type": "string" + }, + "employeeZip": { + "type": "string" + }, + "employeeSSN": { + "type": "string" + }, + "wagesBox1": { + "type": "number" + }, + "taxWithheldBox2": { + "type": "number" + }, + "ssWagesBox3": { + "type": "number" + }, + "ssWithheldBox4": { + "type": "number" + }, + "medicareWagesBox5": { + "type": "number" + }, + "medicareWithheldBox6": { + "type": "number" + }, + "employerEIN": { + "type": "string" + }, + "employerName": { + "type": "string" + }, + "employerAddress": { + "type": "string" + }, + "employerCity": { + "type": "string" + }, + "employerState": { + "type": "string" + }, + "employerZip": { + "type": "string" + }, + "taxYear": { + "type": "string" + } + }, + "required": [ + "employeeFirstName", + "employeeLastName", + "employeeAddress", + "employeeCity", + "employeeState", + "employeeZip", + "employeeSSN", + "wagesBox1", + "taxWithheldBox2", + "ssWagesBox3", + "ssWithheldBox4", + "medicareWagesBox5", + "medicareWithheldBox6", + "employerEIN", + "employerName", + "employerAddress", + "employerCity", + "employerState", + "employerZip", + "taxYear" + ], + "additionalProperties": false +} \ No newline at end of file diff --git a/app/(apis)/llm/models/route.ts b/app/(apis)/llm/models/route.ts new file mode 100644 index 000000000..fd96c54d1 --- /dev/null +++ b/app/(apis)/llm/models/route.ts @@ -0,0 +1,26 @@ +import { models } from '@/pkgs/language-models' + +const uniquePerField = (array: any[],field: string) => { + // Returns all values inside an array + // based on the field value, ensuring that only one of that field value + // is returned. + const unique = new Set() + return array.filter((item) => { + if (unique.has(item[field])) return false + unique.add(item[field]) + return true + }) +} + +export async function GET(req: Request) { + return Response.json({ + object: 'list', + data: uniquePerField(models, 'slug').map((model) => ({ + id: model.slug, + object: 'model', + created: model.createdAt, + owned_by: model.author, + permission: [] + })) + }) +} \ No newline at end of file diff --git a/app/(apis)/models/tools/[model]/route.ts b/app/(apis)/models/tools/[model]/route.ts index 00f7e9455..673e9967e 100644 --- a/app/(apis)/models/tools/[model]/route.ts +++ b/app/(apis)/models/tools/[model]/route.ts @@ -67,14 +67,14 @@ export const GET = API(async (request, { db, user, origin, url, domain, params } const resolvedConfig = await resolveConfig({ model: modelName, - user: user.email || 'connor@driv.ly' + user: user.email }) const parsed = parse(modelName) const composio = new Composio({ apiKey: process.env.COMPOSIO_API_KEY }) const connections = await composio.connectedAccounts.list({ - user_uuid: user.email || 'connor@driv.ly' + user_uuid: user.email }) const composioToolset = new VercelAIToolSet({ diff --git a/app/(apis)/route.ts b/app/(apis)/route.ts index 143ad985c..507fa5255 100644 --- a/app/(apis)/route.ts +++ b/app/(apis)/route.ts @@ -78,33 +78,42 @@ export const GET = API(async (request, { db, user, origin, url, domain, payload } for (const d of filteredDomains) { - if (d.endsWith('.do')) { - let isInCategory = false - for (const sites of Object.values(siteCategories)) { - if (sites.includes(d)) { - isInCategory = true - break - } - } - - if (!isInCategory) { - const siteName = d.replace('.do', '') - const description = getDomainDescription(d) || '' - const siteTitle = `${titleCase(siteName)}${description ? ` - ${description}` : ''}` - - let category = 'Other' + if (typeof d !== 'string') { + console.error(`Invalid domain encountered: ${d}`, typeof d) + continue + } - if (collectionSlugs.includes(siteName)) { - category = 'Collections' - if (!formattedSites['Collections']) { - formattedSites['Collections'] = {} + if (d.endsWith('.do')) { + try { + let isInCategory = false + for (const sites of Object.values(siteCategories)) { + if (sites.includes(d)) { + isInCategory = true + break } } - if (!formattedSites[category]) { - formattedSites[category] = {} + if (!isInCategory) { + const siteName = d.replace('.do', '') + const description = getDomainDescription(d) || '' + const siteTitle = `${titleCase(siteName)}${description ? ` - ${description}` : ''}` + + let category = 'Other' + + if (collectionSlugs.includes(siteName)) { + category = 'Collections' + if (!formattedSites['Collections']) { + formattedSites['Collections'] = {} + } + } + + if (!formattedSites[category]) { + formattedSites[category] = {} + } + formattedSites[category][siteTitle] = formatWithOptions(`sites/${siteName}`, d) } - formattedSites[category][siteTitle] = formatWithOptions(`sites/${siteName}`, d) + } catch (error) { + console.error(`Error processing domain ${d}:`, error) } } } diff --git a/app/(sites)/sites/gpt.do/components/chat.tsx b/app/(sites)/sites/gpt.do/components/chat.tsx index 8c146a244..2a7898d0a 100644 --- a/app/(sites)/sites/gpt.do/components/chat.tsx +++ b/app/(sites)/sites/gpt.do/components/chat.tsx @@ -65,7 +65,9 @@ export const Chat = ({ id, initialChatModel, initialVisibilityType, availableMod model: selectedModelId.value, selectedVisibilityType: initialVisibilityType, output: outputFormat, - tools: tools ? { [tools]: true } : {}, + modelOptions: { + tools: tools ? [tools] : undefined + }, system: systemPrompt, temp: temperature, seed: seed, diff --git a/collections/admin/APIKeys.ts b/collections/admin/APIKeys.ts index cdb9a64c8..ba708cecc 100644 --- a/collections/admin/APIKeys.ts +++ b/collections/admin/APIKeys.ts @@ -1,3 +1,4 @@ +import { createKey, findKey, getKey } from '@/lib/openrouter' import type { CollectionConfig } from 'payload' export const APIKeys: CollectionConfig = { @@ -13,9 +14,13 @@ export const APIKeys: CollectionConfig = { disableLocalStrategy: true, }, fields: [ - { name: 'name', type: 'text' }, + { name: 'name', type: 'text', required: true }, + { name: 'user', type: 'relationship', relationTo: 'users' }, + { name: 'organization', type: 'relationship', relationTo: 'organizations' }, { name: 'email', type: 'text' }, { name: 'description', type: 'text' }, + { name: 'key', type: 'text' }, + { name: 'hash', type: 'text' }, { name: 'url', type: 'text' }, { name: 'cfWorkerDomains', @@ -33,4 +38,44 @@ export const APIKeys: CollectionConfig = { ], }, ], + hooks: { + beforeOperation: [ + async ({ operation, args }) => { + const data = args.data + if (operation === 'create') { + if (data.key) { + const key = await findKey(data.key) + if (key) data.hash = key.hash + } + if (!data.hash) { + const { key, hash } = await createKey({ name: data.name, limit: 1 }) + data.key = key + data.hash = hash + } + } + return args + }, + ], + }, + endpoints: [ + { + path: '/:id/credit', + method: 'get', + handler: async ({ routeParams = {}, payload }) => { + const { id } = routeParams + if (typeof id !== 'string' && typeof id !== 'number') { + return Response.json({ error: 'API key ID is required' }, { status: 400 }) + } + const apiKey = await payload.findByID({ + collection: 'apikeys', + id, + }) + if (!apiKey?.key) { + return Response.json({ error: 'API key not found' }, { status: 404 }) + } + const usage = await getKey(apiKey.key) + return Response.json({ credit: usage.limit_remaining }) + }, + }, + ], } diff --git a/collections/observability/Events.ts b/collections/observability/Events.ts index a3a8d822a..712c47b99 100644 --- a/collections/observability/Events.ts +++ b/collections/observability/Events.ts @@ -7,7 +7,7 @@ export const Events: CollectionConfig = { useAsTitle: 'type', description: 'Records of all significant occurrences within the platform', }, - access: { create: () => false, update: () => false, delete: () => false }, + access: { update: () => false, delete: () => false }, fields: [ { name: 'type', type: 'text' }, { name: 'source', type: 'text' }, diff --git a/lib/openrouter.ts b/lib/openrouter.ts new file mode 100644 index 000000000..ef1e94341 --- /dev/null +++ b/lib/openrouter.ts @@ -0,0 +1,98 @@ +const headers = { + Authorization: `Bearer ${process.env.OPENROUTER_PROVISIONING_KEY}`, +} + +export async function identifyUser(apiKey: string) { + const keyMatch = await findKey(apiKey) + return keyMatch + ? { + authenticationType: 'apiKey', + email: keyMatch.name, + } + : null +} + +// Make sure the API key is valid +export async function findKey(apiKey: string) { + // We can only match on the first 3 characters, and the last 3 of the API key. + const keyAfterIntro = apiKey.split('-v1-')[1] + const fixedApiKey = keyAfterIntro.slice(0, 3) + '...' + keyAfterIntro.slice(-3) + + // Loop until we find a match or we've ran out + let keyMatch + let resLength = 0 + let offset = 0 + do { + offset += resLength + const res = await fetch(`https://openrouter.ai/api/v1/keys?include_disabled=false&offset=${offset}`, { headers }) + .then((x) => x.json()) + .then((x) => x.data as KeyDetails[]) + resLength = res.length + if (resLength) keyMatch = res.find((key) => key.label?.split('-v1-')[1] === fixedApiKey) + } while (!keyMatch && resLength > 0) + return keyMatch || null +} + +export async function createKey(details: { name: string; limit?: number }) { + return await fetch(`https://openrouter.ai/api/v1/keys`, { + headers, + method: 'POST', + body: JSON.stringify(details), + }) + .then((x) => x.json()) + .then((x) => x.data as KeyDetails) +} + +export async function getKeyDetails(hash: string) { + return await fetch(`https://openrouter.ai/api/v1/keys/${hash}`, { headers }) + .then((x) => x.json()) + .then((x) => x.data as KeyDetails) +} + +export async function updateKeyDetails( + hash: string, + details: { + name?: string + disabled?: boolean + limit?: number + }, +) { + return await fetch(`https://openrouter.ai/api/v1/keys/${hash}`, { + headers, + method: 'PATCH', + body: JSON.stringify(details), + }) + .then((x) => x.json()) + .then((x) => x.data as KeyDetails) +} + +export async function getKey(apiKey: string) { + return await fetch('https://openrouter.ai/api/v1/key', { headers: { Authorization: `Bearer ${apiKey}` } }) + .then((x) => x.json()) + .then( + (x) => + x.data as { + label: string + usage: number + is_free_tier: boolean + is_provisioning_key: boolean + rate_limit: { + requests: number + interval: string + } + limit?: number + limit_remaining?: number + }, + ) +} + +export type KeyDetails = { + name: string + label?: string + limit?: number + disabled?: boolean + created_at?: string + updated_at?: string + hash?: string + key?: string +} diff --git a/package.json b/package.json index 588aae25a..8090163d1 100644 --- a/package.json +++ b/package.json @@ -55,6 +55,7 @@ "prettier-fix": "prettier --write \"**/*.{js,ts,tsx,md,mdx}\"", "test": "turbo test", "test:watch": "vitest", + "test:sdk": "vitest run tests/sdk", "test:unit": "vitest run --exclude tests/e2e/ --exclude tests/chromatic/", "test:e2e": "vitest run tests/e2e-playwright", "test:visual": "playwright test tests/chromatic", @@ -316,7 +317,7 @@ "jsxSingleQuote": true, "jsxBracketSameLine": true }, - "packageManager": "pnpm@10.10.0", + "packageManager": "pnpm@10.11.0", "nextBundleAnalysis": { "budget": 358400, "budgetPercentIncreaseRed": 20, diff --git a/payload.config.ts b/payload.config.ts index f1ab318be..a2c93032d 100644 --- a/payload.config.ts +++ b/payload.config.ts @@ -1,14 +1,13 @@ // storage-adapter-import-placeholder // import { payloadAgentPlugin } from '@drivly/payload-agent' // import { betterAuthPlugin } from '@drivly/better-payload-auth/plugin' -import { openapi } from 'payload-oapi' import { mongooseAdapter } from '@payloadcms/db-mongodb' import { resendAdapter } from '@payloadcms/email-resend' import { payloadCloudPlugin } from '@payloadcms/payload-cloud' -import { stripePlugin } from '@payloadcms/plugin-stripe' import { multiTenantPlugin } from '@payloadcms/plugin-multi-tenant' +import { stripePlugin } from '@payloadcms/plugin-stripe' import { lexicalEditor } from '@payloadcms/richtext-lexical' -import { createHooksQueuePlugin } from './src/utils/hooks-queue' +import { openapi } from 'payload-oapi' // import workosPlugin from './pkgs/payload-workos' import path from 'path' import { buildConfig } from 'payload' @@ -16,6 +15,8 @@ import { buildConfig } from 'payload' import sharp from 'sharp' import { fileURLToPath } from 'url' import { collections } from './collections' +import { getKeyDetails, updateKeyDetails } from './lib/openrouter' +import type { Apikey } from './payload.types' import { tasks, workflows } from './tasks' const filename = fileURLToPath(import.meta.url) @@ -118,20 +119,88 @@ export default buildConfig({ stripePlugin({ stripeSecretKey: process.env.STRIPE_SECRET_KEY || '', stripeWebhooksEndpointSecret: process.env.STRIPE_WEBHOOK_SECRET, + webhooks: async ({ event, stripe, pluginConfig: stripeConfig, payload }) => { + await payload.db.create({ + collection: 'events', + data: { + source: 'stripe', + type: event.type, + metadata: event.data.object, + }, + }) + if (event.type === 'checkout.session.completed') { + const { customer_details, amount_total } = event.data.object + const { + docs: [key], + } = await payload.db.find({ + collection: 'apikeys', + where: { + email: { + equals: customer_details?.email, + }, + hash: { + exists: true, + }, + }, + }) + if (key?.hash) { + let details = await getKeyDetails(key.hash) + details = await updateKeyDetails(key.hash, { limit: (details.limit || 0) + amount_total / 100 }) + } + } else { + console.log(event.type, JSON.stringify(event.data.object, null, 2)) + } + }, + // sync: [ + // { + // stripeResourceTypeSingular: 'customer', + // stripeResourceType: 'customers', + // collection: 'connectAccounts', + // fields: [ + // { + // fieldPath: 'stripeAccountId', + // stripeProperty: 'id', + // }, + // { + // fieldPath: 'email', + // stripeProperty: 'email', + // }, + // { + // fieldPath: 'name', + // stripeProperty: 'name', + // }, + // ], + // }, + // { + // stripeResourceTypeSingular: 'product', + // stripeResourceType: 'products', + // collection: 'billingPlans', + // fields: [ + // { + // fieldPath: 'stripeProductId', + // stripeProperty: 'id', + // }, + // { + // fieldPath: 'name', + // stripeProperty: 'name', + // }, + // ], + // }, + // ], }), - createHooksQueuePlugin({ - 'nouns.beforeChange': 'inflectNouns', - 'nouns.afterChange': 'inflectNouns', - 'verbs.beforeChange': 'conjugateVerbs', - 'verbs.afterChange': 'conjugateVerbs', - 'functions.afterChange': ['processCodeFunction', 'generateFunctionExamples'], - 'searches.beforeChange': 'generateEmbedding', - 'searches.afterChange': ['searchThings', 'hybridSearchThings'], - 'actions.afterChange': 'executeFunction', - 'events.afterChange': 'deliverWebhook', - 'tasks.afterChange': ['syncTaskToLinear'], - 'tasks.afterDelete': ['deleteLinearIssue'], - }), + // createHooksQueuePlugin({ + // 'nouns.beforeChange': 'inflectNouns', + // 'nouns.afterChange': 'inflectNouns', + // 'verbs.beforeChange': 'conjugateVerbs', + // 'verbs.afterChange': 'conjugateVerbs', + // 'functions.afterChange': ['processCodeFunction', 'generateFunctionExamples'], + // 'searches.beforeChange': 'generateEmbedding', + // 'searches.afterChange': ['searchThings', 'hybridSearchThings'], + // 'actions.afterChange': 'executeFunction', + // 'events.afterChange': 'deliverWebhook', + // 'tasks.afterChange': ['syncTaskToLinear'], + // 'tasks.afterDelete': ['deleteLinearIssue'], + // }), multiTenantPlugin({ tenantSelectorLabel: 'Project', collections: { diff --git a/payload.types.ts b/payload.types.ts index 47a6a93b0..91e088a53 100644 --- a/payload.types.ts +++ b/payload.types.ts @@ -2599,9 +2599,13 @@ export interface Webhook { */ export interface Apikey { id: string; - name?: string | null; + name: string; + user?: (string | null) | User; + organization?: (string | null) | Organization; email?: string | null; description?: string | null; + key?: string | null; + hash?: string | null; url?: string | null; /** * Domains of authorized Cloudflare Workers @@ -4299,8 +4303,12 @@ export interface WebhooksSelect { */ export interface ApikeysSelect { name?: T; + user?: T; + organization?: T; email?: T; description?: T; + key?: T; + hash?: T; url?: T; cfWorkerDomains?: | T diff --git a/pkgs/ai-providers/src/ai.ts b/pkgs/ai-providers/src/ai.ts index bd4938695..ffb1bcc5f 100644 --- a/pkgs/ai-providers/src/ai.ts +++ b/pkgs/ai-providers/src/ai.ts @@ -29,29 +29,40 @@ const camelCaseToScreamingSnakeCase = (str: string) => { .replace(/([A-Z])/g, '_$1').toUpperCase() } -type GenerateTextOptions = Omit[0], 'model'> & { +type ProvidersGenerateMixin = { model: (string & {}) | LanguageModelV1 user?: string openrouterApiKey?: string + modelOptions?: any + /** + * Report when a tool is called, used for analytics. + * @param tool + * @param args + * @returns + */ + onTool?: (tool: string, args: any, result: any) => void } -type GenerateObjectOptions = Omit[0], 'model'> & { - model: (string & {}) | LanguageModelV1 - user?: string - openrouterApiKey?: string -} +type GenerateTextOptions = Omit[0], 'model'> & ProvidersGenerateMixin -type StreamObjectOptions = Omit[0], 'model'> & { - model: (string & {}) | LanguageModelV1 - user?: string - openrouterApiKey?: string +type GenerateObjectOptions = Omit[0], 'model'> & ProvidersGenerateMixin + +type StreamObjectOptions = Omit[0], 'model'> & ProvidersGenerateMixin + +export type AIToolRedirectError = Error & { + type: 'AI_PROVIDERS_TOOLS_REDIRECT' + connectionRequests: { + app: string + redirectUrl: string + }[] + apps: string[] } // Generates a config object from export async function resolveConfig(options: GenerateTextOptions) { // If options.model is a string, use our llm provider. if (typeof options.model === 'string') { - options.model = model(options.model) + options.model = model(options.model, options.modelOptions || {}) } // @ts-expect-error - This is out of spec for LanguageModelV1, but we need to know if this is the LLMProvider @@ -60,34 +71,113 @@ export async function resolveConfig(options: GenerateTextOptions) { // @ts-expect-error - We know this property exists, but TS doesnt const parsedModel = options.model?.resolvedModel - if (parsedModel.parsed?.tools && Object.keys(parsedModel.parsed.tools).length > 0) { + const toolNames = Array.isArray(parsedModel.parsed.tools) ? parsedModel.parsed.tools : Object.keys(parsedModel.parsed.tools || {}) + + console.log( + 'Using tools', + toolNames, + parsedModel.parsed + ) + + if (parsedModel.parsed?.tools && toolNames.length > 0) { if (!options.user) { throw new Error('user is required when using tools') } - const toolNames = Object.keys(parsedModel.parsed.tools) - - const composio = new Composio({ apiKey: process.env.COMPOSIO_API_KEY }) - const connections = await composio.connectedAccounts.list({ - user_uuid: options.user - }) - - const composioToolset = new VercelAIToolSet({ - apiKey: process.env.COMPOSIO_API_KEY, - connectedAccountIds: connections.items - .map(connection => [connection.appName, connection.id]) - .reduce((acc, [app, id]) => ({ ...acc, [app]: id }), {}) - }) - - const apps = toolNames.map(name => name.split('.')[0]) + options.tools = options.tools ?? {} - const tools = await composioToolset.getTools({ - apps, - actions: toolNames.map(name => camelCaseToScreamingSnakeCase(name)), - }) + if (toolNames.length > 0) { + const composio = new Composio({ apiKey: process.env.COMPOSIO_API_KEY }) - options.tools = options.tools ?? {} - options.tools = { ...options.tools, ...tools } + const connections = await composio.connectedAccounts.list({ + entityId: options.user, + status: 'ACTIVE' + }) + + // Return a completion error if the user tries to use a app + // that they have not yet connected. Inside this error, we will include a + // redirect link to add the app. + + const activeApps = connections.items.map(connection => connection.appName) + let missingApps: string[] = Array.from(new Set(toolNames.map((x: string) => x.split('.')[0]).filter((app: string) => !activeApps.includes(app)))) + + const appMetadata = await Promise.all(missingApps.map(async app => { + return composio.apps.get({ + appKey: app as string + }) + })) + + // Mixin a filter to remove apps that dont have any auth. + missingApps = missingApps.filter(app => !appMetadata.find(x => x.key === app)?.no_auth) + + if (missingApps.length > 0) { + const connectionRequests = [] + + // Dont make a new connection request if we've already got one pending. + const pendingConnections = await composio.connectedAccounts.list({ + entityId: options.user, + status: 'INITIATED' + }).then(x => x.items) + + for (const app of missingApps) { + if (pendingConnections.find(x => x.appName === app)) { + console.debug( + '[COMPOSIO] Found existing connection request for', + app + ) + + const connection = pendingConnections.find(x => x.appName === app) + + connectionRequests.push({ + app: connection?.appName as string, + redirectUrl: connection?.connectionParams?.redirectUrl as string + }) + } else { + const integration = await composio.integrations.create({ + name: app, + appUniqueKey: app, + useComposioAuth: true, + forceNewIntegration: true, + }) + + const connection = await composio.connectedAccounts.initiate({ + integrationId: integration.id, + entityId: options.user + }) + + connectionRequests.push({ + app: app as string, + redirectUrl: connection.redirectUrl as string + }) + } + } + + const error = new Error(`Missing access to apps: ${missingApps.join(', ')}.`) as AIToolRedirectError + + error.type = 'AI_PROVIDERS_TOOLS_REDIRECT' + error.connectionRequests = connectionRequests + error.apps = missingApps + + throw error + } + + const composioToolset = new VercelAIToolSet({ + apiKey: process.env.COMPOSIO_API_KEY, + connectedAccountIds: connections.items + .map(connection => [connection.appName, connection.id]) + .reduce((acc, [app, id]) => ({ ...acc, [app]: id }), {}) + }) + + const apps = toolNames.map((name: string) => name.split('.')[0]) + + const tools = await composioToolset.getTools({ + apps, + actions: toolNames.map((name: string) => camelCaseToScreamingSnakeCase(name)), + }) + + + options.tools = { ...options.tools, ...tools } + } if (parsedModel?.parsed?.tools?.fetch) { options.tools.fetchWebsiteContents = fetchWebsiteContents as Tool @@ -104,7 +194,7 @@ export async function resolveConfig(options: GenerateTextOptions) { if (parsedModel.provider?.slug === 'openAi') { // We need to amend composio tools for OpenAI usage. - for (const [name, tool] of Object.entries(tools)) { + for (const [name, tool] of Object.entries(options.tools)) { options.tools[name] = { ...tool, parameters: { @@ -120,13 +210,33 @@ export async function resolveConfig(options: GenerateTextOptions) { `[TOOL:${name}]`, args ) - // @ts-expect-error - TS doesnt like us calling this function even though it exists. - return tool.execute(args) + + try { + // @ts-expect-error - TS doesnt like us calling this function even though it exists. + const result = await tool.execute(args) + + if (options.onTool) { + options.onTool(name, args, { + success: true, + result + }) + } + + return result + } catch (error) { + if (options.onTool) { + options.onTool(name, args, { + success: false, + error: JSON.parse(JSON.stringify(error)) + }) + } + throw error + } } } } } else { - for (const [name, tool] of Object.entries(tools)) { + for (const [name, tool] of Object.entries(options.tools)) { options.tools[name] = { ...tool, execute: async (args: any) => { @@ -134,8 +244,28 @@ export async function resolveConfig(options: GenerateTextOptions) { `[TOOL:${name}]`, args ) - // @ts-expect-error - TS doesnt like us calling this function even though it exists. - return tool.execute(args) + + try { + // @ts-expect-error - TS doesnt like us calling this function even though it exists. + const result = await tool.execute(args) + + if (options.onTool) { + options.onTool(name, args, { + success: true, + result + }) + } + + return result + } catch (error) { + if (options.onTool) { + options.onTool(name, args, { + success: false, + error: JSON.parse(JSON.stringify(error)) + }) + } + throw error + } } } } @@ -143,7 +273,6 @@ export async function resolveConfig(options: GenerateTextOptions) { // Apply model author specific fixes if (parsedModel.author == 'google') { - // For each tool, we need to replace the jsonSchema with a google compatible one. for (const toolName in options.tools) { options.tools[toolName].parameters.jsonSchema = convertJSONSchemaToOpenAPISchema(options.tools[toolName].parameters.jsonSchema) @@ -151,14 +280,8 @@ export async function resolveConfig(options: GenerateTextOptions) { } if (parsedModel.author == 'openai') { - // For each tool, we need to replace the jsonSchema with a google compatible one. for (const toolName in options.tools) { - - console.log( - options.tools[toolName].parameters.jsonSchema - ) - options.tools[toolName].parameters.jsonSchema = alterSchemaForOpenAI(options.tools[toolName].parameters.jsonSchema) } } diff --git a/pkgs/ai-providers/src/index.ts b/pkgs/ai-providers/src/index.ts index 8a356bfe9..50aaba1cc 100644 --- a/pkgs/ai-providers/src/index.ts +++ b/pkgs/ai-providers/src/index.ts @@ -1,2 +1,4 @@ export * from './provider' -export * from './ai' \ No newline at end of file +export * from './ai' +export * from './providers/google' +export * from './providers/openai' \ No newline at end of file diff --git a/pkgs/ai-providers/src/provider.ts b/pkgs/ai-providers/src/provider.ts index 0ce454469..1d9d87075 100644 --- a/pkgs/ai-providers/src/provider.ts +++ b/pkgs/ai-providers/src/provider.ts @@ -1,8 +1,7 @@ -import { LanguageModelV1, LanguageModelV1CallWarning, LanguageModelV1FinishReason, LanguageModelV1StreamPart } from '@ai-sdk/provider' - -import { createOpenAI } from '@ai-sdk/openai' -import { createGoogleGenerativeAI } from '@ai-sdk/google' +import { LanguageModelV1 } from '@ai-sdk/provider' import { createAnthropic } from '@ai-sdk/anthropic' +import { createGoogleGenerativeAI } from '@ai-sdk/google' +import { createOpenAI } from '@ai-sdk/openai' import { getModel, getModels, Model } from 'language-models' // Not in use for the inital release. @@ -113,11 +112,15 @@ class LLMProvider implements LanguageModelV1 { this.modelId = modelId this.options = options ?? {} this.config = config ?? {} - this.resolvedModel = getModel(modelId) + // @ts-expect-error - Type is wrong + this.resolvedModel = getModel(modelId, options) - if (!this.resolvedModel.slug) { - throw new Error(`Model ${modelId} not found`) - } + console.log( + 'MODEL', + this.resolvedModel, + 'OPTIONS', + options + ) this.apiKey = this.config?.apiKey @@ -129,11 +132,11 @@ class LLMProvider implements LanguageModelV1 { get provider() { let provider = 'openrouter' - // Access provider property which is added by getModel but not in the Model type - const providerSlug = this.resolvedModel.provider?.slug - return provider + // // Access provider property which is added by getModel but not in the Model type + // const providerSlug = this.resolvedModel.provider?.slug + // switch (providerSlug) { // case 'openAi': // provider = 'openai' @@ -159,7 +162,7 @@ class LLMProvider implements LanguageModelV1 { get supportsImageUrls() { // Depending on the model, we may or may not support image urls - return this.resolvedModel.inputModalities.includes('image') + return this.resolvedModel.inputModalities?.includes('image') } // Fix Anthropic's default object generation mode. @@ -170,7 +173,7 @@ class LLMProvider implements LanguageModelV1 { async doGenerate(options: Parameters[0]): Promise>> { // Access providerModelId which is added by getModel but not in the Model type - const modelSlug = this.provider == 'openrouter' ? this.resolvedModel.slug : (this.resolvedModel as any).provider?.providerModelId + const modelSlug = this.provider == 'openrouter' ? this.resolvedModel.slug || this.modelId : (this.resolvedModel as any).provider?.providerModelId let modelConfigMixin = {} @@ -186,7 +189,7 @@ class LLMProvider implements LanguageModelV1 { const provider = createOpenAI({ baseURL: this.config?.baseURL || 'https://gateway.ai.cloudflare.com/v1/b6641681fe423910342b9ffa1364c76d/ai-functions/openrouter', apiKey: this.apiKey, - headers: this.config?.headers, + headers: this.config?.headers }) return await provider(modelSlug, modelConfigMixin).doGenerate(options) diff --git a/pkgs/ai-providers/src/providers/openai.ts b/pkgs/ai-providers/src/providers/openai.ts index e4f4c9f2a..6ff4ce316 100644 --- a/pkgs/ai-providers/src/providers/openai.ts +++ b/pkgs/ai-providers/src/providers/openai.ts @@ -37,14 +37,28 @@ export function alterSchemaForOpenAI( if (description) result.description = description; if (required) result.required = required; - if (format) result.format = format; + // OpenAI doesn't allow format in schema properties if (constValue !== undefined) { result.enum = [constValue]; } - // Handle type - if (type) { + // Process properties first so we can check if it exists for type inference + if (properties != null) { + result.properties = Object.entries(properties).reduce( + (acc, [key, value]) => { + acc[key] = alterSchemaForOpenAI(value); + return acc; + }, + {} as Record, + ); + } + + // Handle type - ensure a type is always present + // If it has properties, it MUST be an object regardless of the specified type + if (properties && Object.keys(properties).length > 0) { + result.type = 'object' + } else if (type) { if (Array.isArray(type)) { if (type.includes('null')) { result.type = type.filter(t => t !== 'null')[0]; @@ -57,6 +71,9 @@ export function alterSchemaForOpenAI( } else { result.type = type; } + } else { + // Default to string if no type is specified and no properties exist + result.type = 'string'; } // Handle enum @@ -64,54 +81,44 @@ export function alterSchemaForOpenAI( result.enum = enumValues; } - if (properties != null) { - result.properties = Object.entries(properties).reduce( - (acc, [key, value]) => { - acc[key] = alterSchemaForOpenAI(value); - return acc; - }, - {} as Record, - ); - } - if (items) { result.items = Array.isArray(items) - ? items.map(alterSchemaForOpenAI) + ? alterSchemaForOpenAI(items[0]) // Just use the first item schema : alterSchemaForOpenAI(items); } - if (allOf) { - result.allOf = allOf.map(alterSchemaForOpenAI); + // Handle combinators by taking only the first schema + if (allOf && allOf.length > 0) { + const firstSchema = alterSchemaForOpenAI(allOf[0]); + if (typeof firstSchema === 'object') { + Object.assign(result, firstSchema); + // Re-check if we need to force type to object after merging + if (result.properties && Object.keys(result.properties as object).length > 0) { + result.type = 'object'; + } + } } - if (anyOf) { - // Handle cases where anyOf includes a null type - if ( - anyOf.some( - schema => typeof schema === 'object' && schema?.type === 'null', - ) - ) { - const nonNullSchemas = anyOf.filter( - schema => !(typeof schema === 'object' && schema?.type === 'null'), - ); - - if (nonNullSchemas.length === 1) { - // If there's only one non-null schema, convert it and make it nullable - const converted = alterSchemaForOpenAI(nonNullSchemas[0]); - if (typeof converted === 'object') { - result.nullable = true; - Object.assign(result, converted); - } - } else { - // If there are multiple non-null schemas, keep them in anyOf - result.anyOf = nonNullSchemas.map(alterSchemaForOpenAI); - result.nullable = true; + + if (anyOf && anyOf.length > 0) { + const firstSchema = alterSchemaForOpenAI(anyOf[0]); + if (typeof firstSchema === 'object') { + Object.assign(result, firstSchema); + // Re-check if we need to force type to object after merging + if (result.properties && Object.keys(result.properties as object).length > 0) { + result.type = 'object'; } - } else { - result.anyOf = anyOf.map(alterSchemaForOpenAI); } } - if (oneOf) { - result.oneOf = oneOf.map(alterSchemaForOpenAI); + + if (oneOf && oneOf.length > 0) { + const firstSchema = alterSchemaForOpenAI(oneOf[0]); + if (typeof firstSchema === 'object') { + Object.assign(result, firstSchema); + // Re-check if we need to force type to object after merging + if (result.properties && Object.keys(result.properties as object).length > 0) { + result.type = 'object'; + } + } } const illegalKeys = [ @@ -120,7 +127,8 @@ export function alterSchemaForOpenAI( 'minimum', 'maximum', 'minLength', - 'maxLength' + 'maxLength', + 'format' // Added format to the list of illegal keys ] for (const key of illegalKeys) { @@ -134,6 +142,11 @@ export function alterSchemaForOpenAI( result.required = Object.keys(result.properties || {}) + // Final check to ensure type is correct based on properties + if (result.properties && Object.keys(result.properties as object).length > 0) { + result.type = 'object' + } + return result; } diff --git a/pkgs/language-models/changelog.md b/pkgs/language-models/changelog.md new file mode 100644 index 000000000..721132c2b --- /dev/null +++ b/pkgs/language-models/changelog.md @@ -0,0 +1,31 @@ +# 2025-05-15 17:02:35 + +### Removed models: +- qwen/qwen2.5-7b-instruct + +# 2025-05-15 16:02:43 +### Added models: +- qwen/qwen2.5-7b-instruct + + +# 2025-05-12 19:03:25 + +### Removed models: +- google/gemini-flash-1.5-8b-exp +# 2025-05-10 02:01:58 + +### Removed models: +- qwen/qwen-2.5-vl-72b-instruct +# 2025-05-10 01:01:55 +### Added models: +- qwen/qwen-2.5-vl-72b-instruct + +# 2025-05-10 00:01:35 +### Added models: +- nousresearch/deephermes-3-mistral-24b-preview +### Removed models: +- qwen/qwen-2.5-vl-72b-instruct +# 2025-05-09 01:02:07 + +### Removed models: +- google/gemini-pro-vision diff --git a/pkgs/language-models/generate/build-models.ts b/pkgs/language-models/generate/build-models.ts index 05c5bb71e..2181bd9c3 100644 --- a/pkgs/language-models/generate/build-models.ts +++ b/pkgs/language-models/generate/build-models.ts @@ -31,6 +31,60 @@ async function fetchProviders(slug: string) { return camelCaseDeep(response.data || []) } +const authorIconCache = new Map() + +async function getModelAuthorIcon(model: string): Promise { + const [ author, modelName ] = model.split('/') + if (authorIconCache.has(author)) { + console.log( + '[ICONS] Used cache for author:', author + ) + + return authorIconCache.get(author) as string + } + + const html = await fetch(`https://openrouter.ai/${author}`).then(res => res.text()) + + // 'https://t0.gstatic.com/faviconV2', + // '/images/icons/' + + const patterns = [ + { + pattern: '/images/icons/', + getUrl: (icon: string) => { + return 'https://openrouter.ai/images/icons/' + icon + } + }, + { + pattern: 'https://t0.gstatic.com/faviconV2', + getUrl: (icon: string) => { + // @ts-expect-error - replaceAll exists, no clue why TS is complaining + return 'https://t0.gstatic.com/faviconV2' + icon.replaceAll('\u0026', '&') + } + } + ] + + const activePattern = patterns.find(p => html.includes(p.pattern)) + + if (!activePattern) { + console.log( + '[ICONS] Falling back as we cant find a icon for', model + ) + + return '' + } + + const icon = html + .split(activePattern.pattern)[1] + .split('\\"')[0] + + const url = activePattern.getUrl(icon) + + authorIconCache.set(author, url) + + return url +} + async function main() { try { const URL = 'https://openrouter.ai/api/frontend/models/find?order=top-weekly' @@ -56,7 +110,11 @@ async function main() { const models = data['top-weekly'] - const modelsData = models.models.map((model) => { + const allProviders = await fetch('https://openrouter.ai/api/frontend/all-providers') + .then(res => res.json()) + .then(data => data.data) + + const modelsData = models.models.map((model) => { if (model.slug in overwrites) { console.log(`Overwriting model ${model.slug} with custom data`, overwrites[model.slug]) const tempModel: Record = flatten(model) @@ -95,6 +153,16 @@ async function main() { let completed = 0 for (const model of modelsData) { + console.log( + `[ICONS] Fetching author icon for ${model.slug}...` + ) + + model.authorIcon = await getModelAuthorIcon(model.slug) + + console.log( + '[ICONS] Found author icon', model.authorIcon + ) + console.log(`[PROVIDERS] Fetching provider metadata for ${model.permaslug}...`) const providers = await fetchProviders(model.permaslug) @@ -112,8 +180,15 @@ async function main() { ) } + let icon = allProviders.find(p => p.displayName === provider.providerDisplayName)?.icon?.url || '' + + if (icon.includes('/images/icons/')) { + icon = `https://openrouter.ai${icon}` + } + return { name: provider.providerDisplayName, + icon, slug: camelCase(providerName), quantization: provider.quantization, context: provider.contextLength, @@ -124,8 +199,8 @@ async function main() { supportedParameters: model.slug === 'anthropic/claude-3.7-sonnet' ? ['max_tokens', 'temperature', 'stop', 'tools', 'tool_choice'] : provider.supportedParameters, inputCost: priceToDollars(provider.pricing.prompt), outputCost: priceToDollars(provider.pricing.completion), - throughput: provider.stats?.[0]?.p50Throughput, - latency: provider.stats?.[0]?.p50Latency, + throughput: provider.stats?.p50Throughput, + latency: provider.stats?.p50Latency, } }) @@ -135,11 +210,53 @@ async function main() { // Write to models.json in src directory const { resolve } = await import('node:path') - const { writeFileSync } = await import('node:fs') + const { writeFileSync, readFileSync, existsSync } = await import('node:fs') const outputPath = resolve('./src/models.js') + writeFileSync(outputPath, `export default ${JSON.stringify({ models: modelsData }, null, 2)}`) + if (existsSync(outputPath)) { + // Read the file first, we want to get a list of models that were added in this run. + const existingModels = JSON.parse(readFileSync(outputPath, 'utf8').replace('export default ', '')) + const newModels = modelsData.filter((model) => !existingModels.models.some((m) => m.permaslug === model.permaslug)) + const removedModels = existingModels.models.filter((model) => !modelsData.some((m) => m.permaslug === model.permaslug)) + + const saveRead = (path: string): string => { + try { + return readFileSync(path, 'utf8') + } catch (error) { + return '' + } + } + + const logPath = resolve('./changelog.md') + const currentLogMarkdown = saveRead(logPath) + + if (newModels.length || removedModels.length) { + + const addedModelsSegment = newModels.length > 0 ? `### Added models: + ${newModels.map((m) => `- ${m.slug}`).join('\n')}` : '' + + const removedModelsSegment = removedModels.length > 0 ? `### Removed models: + ${removedModels.map((m) => `- ${m.slug}`).join('\n')}` : '' + + writeFileSync( + logPath, + // Append the new models to the file. + // Title should be the date and hour of the run. + (`# ${new Date().toISOString().split('T')[0]} ${new Date().toISOString().split('T')[1].split('.')[0]} + ${addedModelsSegment} + ${removedModelsSegment} + + ${currentLogMarkdown}`) + // @ts-expect-error - replaceAll exists, no clue why TS is complaining + .replaceAll(' ', '') + ) + } + + } + console.log(`Models data written to ${outputPath}`) } catch (error) { console.error('Error fetching or writing models data:', error) diff --git a/pkgs/language-models/package.json b/pkgs/language-models/package.json index 06100d064..b91313e34 100644 --- a/pkgs/language-models/package.json +++ b/pkgs/language-models/package.json @@ -11,7 +11,8 @@ "lint": "tsc --noEmit", "prepublishOnly": "pnpm clean && pnpm build", "test": "vitest", - "typecheck": "tsc --noEmit" + "typecheck": "tsc --noEmit", + "cli": "npx tsx src/cli.ts" }, "type": "module", "packageManager": "pnpm@10.6.5", diff --git a/pkgs/language-models/src/cli.ts b/pkgs/language-models/src/cli.ts new file mode 100644 index 000000000..d907d1f1b --- /dev/null +++ b/pkgs/language-models/src/cli.ts @@ -0,0 +1,11 @@ +// A CLI interface for the parser. + +import { getModel } from './parser' + +// Get the model name from the command line arguments +const modelName = process.argv[2] + +// Get the model from the parser +const model = getModel(modelName) + +console.log(model) \ No newline at end of file diff --git a/pkgs/language-models/src/models.d.ts b/pkgs/language-models/src/models.d.ts index 7f7c2b7de..72e1a068f 100644 --- a/pkgs/language-models/src/models.d.ts +++ b/pkgs/language-models/src/models.d.ts @@ -107,6 +107,7 @@ interface ModelBase { warningMessage: string | null permaslug: string reasoningConfig: ReasoningConfig | null + authorIcon: string | null } // Define Endpoint structure, referencing ModelBase for its 'model' property diff --git a/pkgs/language-models/src/models.js b/pkgs/language-models/src/models.js index ee77c9ccb..7c9152bb3 100644 --- a/pkgs/language-models/src/models.js +++ b/pkgs/language-models/src/models.js @@ -191,15 +191,17 @@ export default { }, "sorting": { "topWeekly": 0, - "newest": 237, - "throughputHighToLow": 160, - "latencyLowToHigh": 59, - "pricingLowToHigh": 147, - "pricingHighToLow": 173 + "newest": 236, + "throughputHighToLow": 128, + "latencyLowToHigh": 55, + "pricingLowToHigh": 144, + "pricingHighToLow": 176 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", "slug": "openAi", "quantization": "unknown", "context": 128000, @@ -233,10 +235,13 @@ export default { "structured_outputs" ], "inputCost": 0.15, - "outputCost": 0.6 + "outputCost": 0.6, + "throughput": 60.7505, + "latency": 440 }, { "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", "slug": "azure", "quantization": null, "context": 128000, @@ -267,7 +272,9 @@ export default { "structured_outputs" ], "inputCost": 0.15, - "outputCost": 0.6 + "outputCost": 0.6, + "throughput": 156.592, + "latency": 1510 } ] }, @@ -431,7 +438,7 @@ export default { "supportsToolParameters": true, "supportsReasoning": true, "supportsMultipart": true, - "limitRpm": null, + "limitRpm": 500, "limitRpd": null, "hasCompletions": false, "hasChatCompletions": true, @@ -444,14 +451,16 @@ export default { "sorting": { "topWeekly": 1, "newest": 108, - "throughputHighToLow": 150, - "latencyLowToHigh": 204, + "throughputHighToLow": 148, + "latencyLowToHigh": 250, "pricingLowToHigh": 269, "pricingHighToLow": 41 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", "slug": "vertex", "quantization": null, "context": 200000, @@ -476,10 +485,13 @@ export default { "tool_choice" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 56.618, + "latency": 1913 }, { "name": "Amazon Bedrock", + "icon": "https://openrouter.ai/images/icons/Bedrock.svg", "slug": "amazonBedrock", "quantization": null, "context": 200000, @@ -504,10 +516,13 @@ export default { "tool_choice" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 35.9425, + "latency": 1886.5 }, { "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", "slug": "anthropic", "quantization": null, "context": 200000, @@ -532,10 +547,13 @@ export default { "tool_choice" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 57.987, + "latency": 1626 }, { "name": "Google Vertex (Europe)", + "icon": "", "slug": "googleVertex (europe)", "quantization": null, "context": 200000, @@ -560,7 +578,9 @@ export default { "tool_choice" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 50.2515, + "latency": 2288.5 } ] }, @@ -742,14 +762,16 @@ export default { "sorting": { "topWeekly": 2, "newest": 118, - "throughputHighToLow": 29, - "latencyLowToHigh": 79, + "throughputHighToLow": 31, + "latencyLowToHigh": 78, "pricingLowToHigh": 128, "pricingHighToLow": 190 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [ { "name": "Google AI Studio", + "icon": "https://openrouter.ai/images/icons/GoogleAIStudio.svg", "slug": "google", "quantization": null, "context": 1048576, @@ -780,10 +802,13 @@ export default { "structured_outputs" ], "inputCost": 0.1, - "outputCost": 0.4 + "outputCost": 0.4, + "throughput": 154.537, + "latency": 550 }, { "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", "slug": "vertex", "quantization": null, "context": 1000000, @@ -814,7 +839,9 @@ export default { "structured_outputs" ], "inputCost": 0.15, - "outputCost": 0.6 + "outputCost": 0.6, + "throughput": 163.8315, + "latency": 479 } ] }, @@ -997,14 +1024,16 @@ export default { "sorting": { "topWeekly": 3, "newest": 41, - "throughputHighToLow": 65, - "latencyLowToHigh": 76, - "pricingLowToHigh": 144, + "throughputHighToLow": 52, + "latencyLowToHigh": 92, + "pricingLowToHigh": 141, "pricingHighToLow": 144 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [ { "name": "Vertex Non-Thinking", + "icon": "", "slug": "vertexNonThinking", "quantization": null, "context": 1048576, @@ -1032,10 +1061,13 @@ export default { "structured_outputs" ], "inputCost": 0.15, - "outputCost": 0.6 + "outputCost": 0.6, + "throughput": 100.101, + "latency": 593 }, { "name": "AI Studio Non-Thinking", + "icon": "", "slug": "aiStudioNonThinking", "quantization": null, "context": 1048576, @@ -1063,7 +1095,274 @@ export default { "structured_outputs" ], "inputCost": 0.15, - "outputCost": 0.6 + "outputCost": 0.6, + "throughput": 127.5345, + "latency": 624 + } + ] + }, + { + "slug": "google/gemini-2.5-pro-preview", + "hfSlug": "", + "updatedAt": "2025-05-07T00:41:53.500835+00:00", + "createdAt": "2025-05-07T00:41:53+00:00", + "hfUpdatedAt": null, + "name": "Google: Gemini 2.5 Pro Preview", + "shortName": "Gemini 2.5 Pro Preview", + "author": "google", + "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.", + "modelVersionGroupId": null, + "contextLength": 1048576, + "inputModalities": [ + "text", + "image", + "file" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Gemini", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": "", + "permaslug": "google/gemini-2.5-pro-preview-03-25", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "9d2cac4d-81d4-4e67-ac7a-6c73040655ee", + "name": "Google | google/gemini-2.5-pro-preview-03-25", + "contextLength": 1048576, + "model": { + "slug": "google/gemini-2.5-pro-preview", + "hfSlug": "", + "updatedAt": "2025-05-07T00:41:53.500835+00:00", + "createdAt": "2025-05-07T00:41:53+00:00", + "hfUpdatedAt": null, + "name": "Google: Gemini 2.5 Pro Preview", + "shortName": "Gemini 2.5 Pro Preview", + "author": "google", + "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.", + "modelVersionGroupId": null, + "contextLength": 1048576, + "inputModalities": [ + "text", + "image", + "file" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Gemini", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": "", + "permaslug": "google/gemini-2.5-pro-preview-03-25", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "google/gemini-2.5-pro-preview", + "modelVariantPermaslug": "google/gemini-2.5-pro-preview-03-25", + "providerName": "Google", + "providerInfo": { + "name": "Google", + "displayName": "Google Vertex", + "slug": "google-vertex", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://cloud.google.com/terms/", + "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", + "dataPolicyUrl": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/abuse-monitoring", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "freeModels": { + "training": true, + "retainsPrompts": true + } + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": false, + "isAbortable": false, + "moderationRequired": false, + "group": "Google", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": "https://status.cloud.google.com/products/sdXM79fz1FS6ekNpu37K/history", + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/GoogleVertex.svg" + } + }, + "providerDisplayName": "Google Vertex", + "providerModelId": "gemini-2.5-pro-preview-05-06", + "providerGroup": "Google", + "quantization": null, + "variant": "standard", + "isFree": false, + "canAbort": false, + "maxPromptTokens": null, + "maxCompletionTokens": 65535, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "tools", + "tool_choice", + "stop", + "seed", + "response_format", + "structured_outputs" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://cloud.google.com/terms/", + "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", + "dataPolicyUrl": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/abuse-monitoring", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "freeModels": { + "training": true, + "retainsPrompts": true + }, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "pricing": { + "prompt": "0.00000125", + "completion": "0.00001", + "image": "0.00516", + "request": "0", + "inputCacheRead": "0.00000031", + "inputCacheWrite": "0.000001625", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [ + { + "type": "prompt-threshold", + "threshold": 200000, + "prompt": "0.0000025", + "completions": "0.000015", + "inputCacheRead": "0.000000625", + "inputCacheWrite": "0.000002875" + } + ], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": true, + "supportsReasoning": true, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": false, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 4, + "newest": 2, + "throughputHighToLow": 96, + "latencyLowToHigh": 274, + "pricingLowToHigh": 242, + "pricingHighToLow": 76 + }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", + "providers": [ + { + "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", + "slug": "vertex", + "quantization": null, + "context": 1048576, + "maxCompletionTokens": 65535, + "providerModelId": "gemini-2.5-pro-preview-05-06", + "pricing": { + "prompt": "0.00000125", + "completion": "0.00001", + "image": "0.00516", + "request": "0", + "inputCacheRead": "0.00000031", + "inputCacheWrite": "0.000001625", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "tools", + "tool_choice", + "stop", + "seed", + "response_format", + "structured_outputs" + ], + "inputCost": 1.25, + "outputCost": 10, + "throughput": 80.49, + "latency": 2356 + }, + { + "name": "Google AI Studio", + "icon": "https://openrouter.ai/images/icons/GoogleAIStudio.svg", + "slug": "google", + "quantization": null, + "context": 1048576, + "maxCompletionTokens": 65536, + "providerModelId": "gemini-2.5-pro-preview-05-06", + "pricing": { + "prompt": "0.00000125", + "completion": "0.00001", + "image": "0.00516", + "request": "0", + "inputCacheRead": "0.00000031", + "inputCacheWrite": "0.000001625", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "tools", + "tool_choice", + "stop", + "seed", + "response_format", + "structured_outputs" + ], + "inputCost": 1.25, + "outputCost": 10, + "throughput": 84.177, + "latency": 4064.5 } ] }, @@ -1223,16 +1522,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 4, + "topWeekly": 5, "newest": 75, - "throughputHighToLow": 242, - "latencyLowToHigh": 137, + "throughputHighToLow": 245, + "latencyLowToHigh": 140, "pricingLowToHigh": 33, "pricingHighToLow": 148 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", "context": 163840, @@ -1261,10 +1562,13 @@ export default { "min_p" ], "inputCost": 0.3, - "outputCost": 0.88 + "outputCost": 0.88, + "throughput": 23.2175, + "latency": 832 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": null, "context": 128000, @@ -1295,10 +1599,13 @@ export default { "logit_bias" ], "inputCost": 0.33, - "outputCost": 1.3 + "outputCost": 1.3, + "throughput": 21.7775, + "latency": 1159 }, { "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", "slug": "klusterAi", "quantization": null, "context": 163840, @@ -1315,13 +1622,27 @@ export default { }, "supportedParameters": [ "max_tokens", - "temperature" + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", + "seed" ], "inputCost": 0.33, - "outputCost": 1.4 + "outputCost": 1.4, + "throughput": 17.584, + "latency": 1120 }, { "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", "slug": "lambda", "quantization": "fp8", "context": 163840, @@ -1352,10 +1673,13 @@ export default { "response_format" ], "inputCost": 0.34, - "outputCost": 0.88 + "outputCost": 0.88, + "throughput": 26.208, + "latency": 872 }, { "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", "slug": "inferenceNet", "quantization": "fp8", "context": 128000, @@ -1384,10 +1708,13 @@ export default { "top_logprobs" ], "inputCost": 0.45, - "outputCost": 1.45 + "outputCost": 1.45, + "throughput": 11.414, + "latency": 1775 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 163840, @@ -1416,10 +1743,13 @@ export default { "top_logprobs" ], "inputCost": 0.5, - "outputCost": 1.5 + "outputCost": 1.5, + "throughput": 18.944, + "latency": 916.5 }, { "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", "slug": "parasail", "quantization": "fp8", "context": 163840, @@ -1444,10 +1774,13 @@ export default { "top_k" ], "inputCost": 0.74, - "outputCost": 1.5 + "outputCost": 1.5, + "throughput": 26.983, + "latency": 822 }, { "name": "CentML", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://centml.ai/&size=256", "slug": "centMl", "quantization": "fp8", "context": 163840, @@ -1475,10 +1808,13 @@ export default { "repetition_penalty" ], "inputCost": 0.8, - "outputCost": 0.8 + "outputCost": 0.8, + "throughput": 25.065, + "latency": 917.5 }, { "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", "slug": "fireworks", "quantization": null, "context": 163840, @@ -1511,10 +1847,13 @@ export default { "top_logprobs" ], "inputCost": 0.9, - "outputCost": 0.9 + "outputCost": 0.9, + "throughput": 64.5325, + "latency": 739 }, { "name": "GMICloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://gmicloud.ai/&size=256", "slug": "gmiCloud", "quantization": "fp8", "context": 131072, @@ -1538,10 +1877,13 @@ export default { "seed" ], "inputCost": 0.9, - "outputCost": 0.9 + "outputCost": 0.9, + "throughput": 57.8755, + "latency": 832 }, { "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", "slug": "hyperbolic", "quantization": "fp8", "context": 163840, @@ -1572,10 +1914,13 @@ export default { "repetition_penalty" ], "inputCost": 1.25, - "outputCost": 1.25 + "outputCost": 1.25, + "throughput": 23.3085, + "latency": 1810 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "fp8", "context": 131072, @@ -1604,10 +1949,13 @@ export default { "response_format" ], "inputCost": 1.25, - "outputCost": 1.25 + "outputCost": 1.25, + "throughput": 24.0765, + "latency": 2668 }, { "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", "slug": "sambaNova", "quantization": null, "context": 32768, @@ -1630,10 +1978,13 @@ export default { "stop" ], "inputCost": 3, - "outputCost": 4.5 + "outputCost": 4.5, + "throughput": 166.299, + "latency": 2382.5 }, { "name": "DeepSeek", + "icon": "https://openrouter.ai/images/icons/DeepSeek.png", "slug": "deepSeek", "quantization": "fp8", "context": 64000, @@ -1662,186 +2013,164 @@ export default { "top_logprobs" ], "inputCost": 0.27, - "outputCost": 1.1 + "outputCost": 1.1, + "throughput": 19.7295, + "latency": 4492 } ] }, { - "slug": "google/gemini-2.5-pro-preview", - "hfSlug": "", - "updatedAt": "2025-05-07T00:41:53.500835+00:00", - "createdAt": "2025-05-07T00:41:53+00:00", + "slug": "deepseek/deepseek-chat-v3-0324", + "hfSlug": "deepseek-ai/DeepSeek-V3-0324", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-03-24T13:59:15.252028+00:00", "hfUpdatedAt": null, - "name": "Google: Gemini 2.5 Pro Preview", - "shortName": "Gemini 2.5 Pro Preview", - "author": "google", - "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.", - "modelVersionGroupId": null, - "contextLength": 1048576, + "name": "DeepSeek: DeepSeek V3 0324", + "shortName": "DeepSeek V3 0324", + "author": "deepseek", + "description": "DeepSeek V3, a 685B-parameter, mixture-of-experts model, is the latest iteration of the flagship chat model family from the DeepSeek team.\n\nIt succeeds the [DeepSeek V3](/deepseek/deepseek-chat-v3) model and performs really well on a variety of tasks.", + "modelVersionGroupId": "be67b3ba-9d99-440c-ae90-6514d99b93ed", + "contextLength": 163840, "inputModalities": [ - "text", - "image", - "file" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Gemini", + "group": "DeepSeek", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "google/gemini-2.5-pro-preview-03-25", + "warningMessage": null, + "permaslug": "deepseek/deepseek-chat-v3-0324", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "9d2cac4d-81d4-4e67-ac7a-6c73040655ee", - "name": "Google | google/gemini-2.5-pro-preview-03-25", - "contextLength": 1048576, + "id": "820376cb-f110-4d56-ab52-5bd6ca269420", + "name": "DeepInfra | deepseek/deepseek-chat-v3-0324", + "contextLength": 163840, "model": { - "slug": "google/gemini-2.5-pro-preview", - "hfSlug": "", - "updatedAt": "2025-05-07T00:41:53.500835+00:00", - "createdAt": "2025-05-07T00:41:53+00:00", + "slug": "deepseek/deepseek-chat-v3-0324", + "hfSlug": "deepseek-ai/DeepSeek-V3-0324", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-03-24T13:59:15.252028+00:00", "hfUpdatedAt": null, - "name": "Google: Gemini 2.5 Pro Preview", - "shortName": "Gemini 2.5 Pro Preview", - "author": "google", - "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.", - "modelVersionGroupId": null, - "contextLength": 1048576, + "name": "DeepSeek: DeepSeek V3 0324", + "shortName": "DeepSeek V3 0324", + "author": "deepseek", + "description": "DeepSeek V3, a 685B-parameter, mixture-of-experts model, is the latest iteration of the flagship chat model family from the DeepSeek team.\n\nIt succeeds the [DeepSeek V3](/deepseek/deepseek-chat-v3) model and performs really well on a variety of tasks.", + "modelVersionGroupId": "be67b3ba-9d99-440c-ae90-6514d99b93ed", + "contextLength": 131072, "inputModalities": [ - "text", - "image", - "file" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Gemini", + "group": "DeepSeek", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "google/gemini-2.5-pro-preview-03-25", + "warningMessage": null, + "permaslug": "deepseek/deepseek-chat-v3-0324", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "google/gemini-2.5-pro-preview", - "modelVariantPermaslug": "google/gemini-2.5-pro-preview-03-25", - "providerName": "Google", + "modelVariantSlug": "deepseek/deepseek-chat-v3-0324", + "modelVariantPermaslug": "deepseek/deepseek-chat-v3-0324", + "providerName": "DeepInfra", "providerInfo": { - "name": "Google", - "displayName": "Google Vertex", - "slug": "google-vertex", + "name": "DeepInfra", + "displayName": "DeepInfra", + "slug": "deepinfra", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://cloud.google.com/terms/", - "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", - "dataPolicyUrl": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/abuse-monitoring", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "freeModels": { - "training": true, - "retainsPrompts": true + "retainsPrompts": false } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, + "hasCompletions": true, + "isAbortable": true, "moderationRequired": false, - "group": "Google", + "group": "DeepInfra", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.cloud.google.com/products/sdXM79fz1FS6ekNpu37K/history", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/GoogleVertex.svg" + "url": "/images/icons/DeepInfra.webp" } }, - "providerDisplayName": "Google Vertex", - "providerModelId": "gemini-2.5-pro-preview-05-06", - "providerGroup": "Google", - "quantization": null, + "providerDisplayName": "DeepInfra", + "providerModelId": "deepseek-ai/DeepSeek-V3-0324", + "providerGroup": "DeepInfra", + "quantization": "fp8", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 65535, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "tools", - "tool_choice", "stop", - "seed", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", "response_format", - "structured_outputs" + "top_k", + "seed", + "min_p" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://cloud.google.com/terms/", - "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", - "dataPolicyUrl": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/abuse-monitoring", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "freeModels": { - "training": true, - "retainsPrompts": true + "retainsPrompts": false }, "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, "pricing": { - "prompt": "0.00000125", - "completion": "0.00001", - "image": "0.00516", + "prompt": "0.0000003", + "completion": "0.00000088", + "image": "0", "request": "0", - "inputCacheRead": "0.00000031", - "inputCacheWrite": "0.000001625", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [ - { - "type": "prompt-threshold", - "threshold": 200000, - "prompt": "0.0000025", - "completions": "0.000015", - "inputCacheRead": "0.000000625", - "inputCacheWrite": "0.000002875" - } - ], + "variablePricings": [], "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": true, + "supportsToolParameters": false, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { "supportedParameters": {}, @@ -1851,27 +2180,27 @@ export default { }, "sorting": { "topWeekly": 5, - "newest": 2, - "throughputHighToLow": 83, - "latencyLowToHigh": 284, - "pricingLowToHigh": 242, - "pricingHighToLow": 76 + "newest": 75, + "throughputHighToLow": 245, + "latencyLowToHigh": 140, + "pricingLowToHigh": 33, + "pricingHighToLow": 148 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [ { - "name": "Google Vertex", - "slug": "vertex", - "quantization": null, - "context": 1048576, - "maxCompletionTokens": 65535, - "providerModelId": "gemini-2.5-pro-preview-05-06", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "fp8", + "context": 163840, + "maxCompletionTokens": null, + "providerModelId": "deepseek-ai/DeepSeek-V3-0324", "pricing": { - "prompt": "0.00000125", - "completion": "0.00001", - "image": "0.00516", + "prompt": "0.0000003", + "completion": "0.00000088", + "image": "0", "request": "0", - "inputCacheRead": "0.00000031", - "inputCacheWrite": "0.000001625", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -1880,241 +2209,68 @@ export default { "max_tokens", "temperature", "top_p", - "tools", - "tool_choice", "stop", - "seed", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", "response_format", - "structured_outputs" + "top_k", + "seed", + "min_p" ], - "inputCost": 1.25, - "outputCost": 10 + "inputCost": 0.3, + "outputCost": 0.88, + "throughput": 23.2175, + "latency": 832 }, { - "name": "Google AI Studio", - "slug": "google", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", "quantization": null, - "context": 1048576, - "maxCompletionTokens": 65536, - "providerModelId": "gemini-2.5-pro-preview-05-06", + "context": 128000, + "maxCompletionTokens": 16000, + "providerModelId": "deepseek/deepseek-v3-0324", "pricing": { - "prompt": "0.00000125", - "completion": "0.00001", - "image": "0.00516", + "prompt": "0.00000033", + "completion": "0.0000013", + "image": "0", "request": "0", - "inputCacheRead": "0.00000031", - "inputCacheWrite": "0.000001625", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "tools", - "tool_choice", "stop", + "frequency_penalty", + "presence_penalty", "seed", - "response_format", - "structured_outputs" - ], - "inputCost": 1.25, - "outputCost": 10 - } - ] - }, - { - "slug": "google/gemini-2.5-pro-exp-03-25", - "hfSlug": "", - "updatedAt": "2025-05-12T16:02:53.80897+00:00", - "createdAt": "2025-03-25T17:01:39.919989+00:00", - "hfUpdatedAt": null, - "name": "Google: Gemini 2.5 Pro Experimental", - "shortName": "Gemini 2.5 Pro Experimental", - "author": "google", - "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.", - "modelVersionGroupId": null, - "contextLength": 1048576, - "inputModalities": [ - "text", - "image", - "file" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Gemini", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": "Update: Google has severly lowered rate limits on this model. Original: Due to extremely high demand, the [Gemini 2.5 Pro Experimental](/google/gemini-2.5-pro-exp-03-25) model is now strictly limited to **1 request per minute and 1000 requests per day** (including errors). Frequent `429` errors are expected. To maintain reliable performance, please switch to the [paid Gemini 2.5 Pro endpoint](/google/gemini-2.5-pro-preview-03-25). Your credits (such as the $10 minimum purchase) can be used directly on the paid endpoint without affecting your free-tier quotas.", - "permaslug": "google/gemini-2.5-pro-exp-03-25", - "reasoningConfig": null, - "features": null, - "endpoint": { - "id": "e5b7a905-dd02-4f6e-b61b-3aa1a1b8b9c8", - "name": "Google AI Studio | google/gemini-2.5-pro-exp-03-25", - "contextLength": 1048576, - "model": { - "slug": "google/gemini-2.5-pro-exp-03-25", - "hfSlug": "", - "updatedAt": "2025-05-12T16:02:53.80897+00:00", - "createdAt": "2025-03-25T17:01:39.919989+00:00", - "hfUpdatedAt": null, - "name": "Google: Gemini 2.5 Pro Experimental", - "shortName": "Gemini 2.5 Pro Experimental", - "author": "google", - "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.", - "modelVersionGroupId": null, - "contextLength": 1048576, - "inputModalities": [ - "text", - "image", - "file" - ], - "outputModalities": [ - "text" + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" ], - "hasTextOutput": true, - "group": "Gemini", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": "Update: Google has severly lowered rate limits on this model. Original: Due to extremely high demand, the [Gemini 2.5 Pro Experimental](/google/gemini-2.5-pro-exp-03-25) model is now strictly limited to **1 request per minute and 1000 requests per day** (including errors). Frequent `429` errors are expected. To maintain reliable performance, please switch to the [paid Gemini 2.5 Pro endpoint](/google/gemini-2.5-pro-preview-03-25). Your credits (such as the $10 minimum purchase) can be used directly on the paid endpoint without affecting your free-tier quotas.", - "permaslug": "google/gemini-2.5-pro-exp-03-25", - "reasoningConfig": null, - "features": null - }, - "modelVariantSlug": "google/gemini-2.5-pro-exp-03-25", - "modelVariantPermaslug": "google/gemini-2.5-pro-exp-03-25", - "providerName": "Google AI Studio", - "providerInfo": { - "name": "Google AI Studio", - "displayName": "Google AI Studio", - "slug": "google-ai-studio", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://cloud.google.com/terms/", - "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 55 - }, - "freeModels": { - "training": true, - "retainsPrompts": true, - "retentionDays": 55 - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, - "moderationRequired": false, - "group": "Google AI Studio", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/GoogleAIStudio.svg" - } - }, - "providerDisplayName": "Google AI Studio", - "providerModelId": "gemini-2.5-pro-exp-03-25", - "providerGroup": "Google AI Studio", - "quantization": null, - "variant": "standard", - "isFree": true, - "canAbort": false, - "maxPromptTokens": null, - "maxCompletionTokens": 65536, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "tools", - "tool_choice", - "stop", - "seed", - "response_format", - "structured_outputs" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://cloud.google.com/terms/", - "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 55 - }, - "freeModels": { - "training": true, - "retainsPrompts": true, - "retentionDays": 55 - }, - "training": true, - "retainsPrompts": true, - "retentionDays": 55 - }, - "pricing": { - "prompt": "0", - "completion": "0", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": 1, - "limitRpd": 2000, - "hasCompletions": false, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null + "inputCost": 0.33, + "outputCost": 1.3, + "throughput": 21.7775, + "latency": 1159 }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 6, - "newest": 72, - "throughputHighToLow": 6, - "latencyLowToHigh": 310, - "pricingLowToHigh": 31, - "pricingHighToLow": 279 - }, - "providers": [ { - "name": "Google AI Studio", - "slug": "google", + "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", + "slug": "klusterAi", "quantization": null, - "context": 1048576, - "maxCompletionTokens": 65536, - "providerModelId": "gemini-2.5-pro-exp-03-25", + "context": 163840, + "maxCompletionTokens": 163840, + "providerModelId": "deepseek-ai/DeepSeek-V3-0324", "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.00000033", + "completion": "0.0000014", "image": "0", "request": "0", "webSearch": "0", @@ -2125,26 +2281,33 @@ export default { "max_tokens", "temperature", "top_p", - "tools", - "tool_choice", "stop", - "seed", - "response_format", - "structured_outputs" + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", + "seed" ], - "inputCost": 0, - "outputCost": 0 + "inputCost": 0.33, + "outputCost": 1.4, + "throughput": 17.584, + "latency": 1120 }, { - "name": "Google Vertex", - "slug": "vertex", - "quantization": null, - "context": 1048576, - "maxCompletionTokens": 65535, - "providerModelId": "gemini-2.5-pro-exp-03-25", + "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", + "slug": "lambda", + "quantization": "fp8", + "context": 163840, + "maxCompletionTokens": 163840, + "providerModelId": "deepseek-v3-0324", "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.00000034", + "completion": "0.00000088", "image": "0", "request": "0", "webSearch": "0", @@ -2152,314 +2315,28 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "tools", - "tool_choice", "stop", + "frequency_penalty", + "presence_penalty", "seed", - "response_format", - "structured_outputs" - ], - "inputCost": 0, - "outputCost": 0 - } - ] - }, - { - "slug": "deepseek/deepseek-chat-v3-0324", - "hfSlug": "deepseek-ai/DeepSeek-V3-0324", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-24T13:59:15.252028+00:00", - "hfUpdatedAt": null, - "name": "DeepSeek: DeepSeek V3 0324", - "shortName": "DeepSeek V3 0324", - "author": "deepseek", - "description": "DeepSeek V3, a 685B-parameter, mixture-of-experts model, is the latest iteration of the flagship chat model family from the DeepSeek team.\n\nIt succeeds the [DeepSeek V3](/deepseek/deepseek-chat-v3) model and performs really well on a variety of tasks.", - "modelVersionGroupId": "be67b3ba-9d99-440c-ae90-6514d99b93ed", - "contextLength": 163840, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "DeepSeek", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "deepseek/deepseek-chat-v3-0324", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "820376cb-f110-4d56-ab52-5bd6ca269420", - "name": "DeepInfra | deepseek/deepseek-chat-v3-0324", - "contextLength": 163840, - "model": { - "slug": "deepseek/deepseek-chat-v3-0324", - "hfSlug": "deepseek-ai/DeepSeek-V3-0324", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-24T13:59:15.252028+00:00", - "hfUpdatedAt": null, - "name": "DeepSeek: DeepSeek V3 0324", - "shortName": "DeepSeek V3 0324", - "author": "deepseek", - "description": "DeepSeek V3, a 685B-parameter, mixture-of-experts model, is the latest iteration of the flagship chat model family from the DeepSeek team.\n\nIt succeeds the [DeepSeek V3](/deepseek/deepseek-chat-v3) model and performs really well on a variety of tasks.", - "modelVersionGroupId": "be67b3ba-9d99-440c-ae90-6514d99b93ed", - "contextLength": 131072, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "DeepSeek", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "deepseek/deepseek-chat-v3-0324", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "deepseek/deepseek-chat-v3-0324", - "modelVariantPermaslug": "deepseek/deepseek-chat-v3-0324", - "providerName": "DeepInfra", - "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", - "baseUrl": "url", - "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", - "paidModels": { - "training": false, - "retainsPrompts": false - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "DeepInfra", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/DeepInfra.webp" - } - }, - "providerDisplayName": "DeepInfra", - "providerModelId": "deepseek-ai/DeepSeek-V3-0324", - "providerGroup": "DeepInfra", - "quantization": "fp8", - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", - "paidModels": { - "training": false, - "retainsPrompts": false - }, - "training": false, - "retainsPrompts": false - }, - "pricing": { - "prompt": "0.0000003", - "completion": "0.00000088", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 4, - "newest": 75, - "throughputHighToLow": 242, - "latencyLowToHigh": 137, - "pricingLowToHigh": 33, - "pricingHighToLow": 148 - }, - "providers": [ - { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "fp8", - "context": 163840, - "maxCompletionTokens": null, - "providerModelId": "deepseek-ai/DeepSeek-V3-0324", - "pricing": { - "prompt": "0.0000003", - "completion": "0.00000088", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "inputCost": 0.3, - "outputCost": 0.88 - }, - { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": null, - "context": 128000, - "maxCompletionTokens": 16000, - "providerModelId": "deepseek/deepseek-v3-0324", - "pricing": { - "prompt": "0.00000033", - "completion": "0.0000013", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" - ], - "inputCost": 0.33, - "outputCost": 1.3 - }, - { - "name": "kluster.ai", - "slug": "klusterAi", - "quantization": null, - "context": 163840, - "maxCompletionTokens": 163840, - "providerModelId": "deepseek-ai/DeepSeek-V3-0324", - "pricing": { - "prompt": "0.00000033", - "completion": "0.0000014", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature" - ], - "inputCost": 0.33, - "outputCost": 1.4 - }, - { - "name": "Lambda", - "slug": "lambda", - "quantization": "fp8", - "context": 163840, - "maxCompletionTokens": 163840, - "providerModelId": "deepseek-v3-0324", - "pricing": { - "prompt": "0.00000034", - "completion": "0.00000088", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format" + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" ], "inputCost": 0.34, - "outputCost": 0.88 + "outputCost": 0.88, + "throughput": 26.208, + "latency": 872 }, { "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", "slug": "inferenceNet", "quantization": "fp8", "context": 128000, @@ -2488,10 +2365,13 @@ export default { "top_logprobs" ], "inputCost": 0.45, - "outputCost": 1.45 + "outputCost": 1.45, + "throughput": 11.414, + "latency": 1775 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 163840, @@ -2520,10 +2400,13 @@ export default { "top_logprobs" ], "inputCost": 0.5, - "outputCost": 1.5 + "outputCost": 1.5, + "throughput": 18.944, + "latency": 916.5 }, { "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", "slug": "parasail", "quantization": "fp8", "context": 163840, @@ -2548,10 +2431,13 @@ export default { "top_k" ], "inputCost": 0.74, - "outputCost": 1.5 + "outputCost": 1.5, + "throughput": 26.983, + "latency": 822 }, { "name": "CentML", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://centml.ai/&size=256", "slug": "centMl", "quantization": "fp8", "context": 163840, @@ -2579,10 +2465,13 @@ export default { "repetition_penalty" ], "inputCost": 0.8, - "outputCost": 0.8 + "outputCost": 0.8, + "throughput": 25.065, + "latency": 917.5 }, { "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", "slug": "fireworks", "quantization": null, "context": 163840, @@ -2615,10 +2504,13 @@ export default { "top_logprobs" ], "inputCost": 0.9, - "outputCost": 0.9 + "outputCost": 0.9, + "throughput": 64.5325, + "latency": 739 }, { "name": "GMICloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://gmicloud.ai/&size=256", "slug": "gmiCloud", "quantization": "fp8", "context": 131072, @@ -2642,10 +2534,13 @@ export default { "seed" ], "inputCost": 0.9, - "outputCost": 0.9 + "outputCost": 0.9, + "throughput": 57.8755, + "latency": 832 }, { "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", "slug": "hyperbolic", "quantization": "fp8", "context": 163840, @@ -2676,10 +2571,13 @@ export default { "repetition_penalty" ], "inputCost": 1.25, - "outputCost": 1.25 + "outputCost": 1.25, + "throughput": 23.3085, + "latency": 1810 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "fp8", "context": 131072, @@ -2708,10 +2606,13 @@ export default { "response_format" ], "inputCost": 1.25, - "outputCost": 1.25 + "outputCost": 1.25, + "throughput": 24.0765, + "latency": 2668 }, { "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", "slug": "sambaNova", "quantization": null, "context": 32768, @@ -2734,10 +2635,13 @@ export default { "stop" ], "inputCost": 3, - "outputCost": 4.5 + "outputCost": 4.5, + "throughput": 166.299, + "latency": 2382.5 }, { "name": "DeepSeek", + "icon": "https://openrouter.ai/images/icons/DeepSeek.png", "slug": "deepSeek", "quantization": "fp8", "context": 64000, @@ -2766,102 +2670,107 @@ export default { "top_logprobs" ], "inputCost": 0.27, - "outputCost": 1.1 + "outputCost": 1.1, + "throughput": 19.7295, + "latency": 4492 } ] }, { - "slug": "meta-llama/llama-3.3-70b-instruct", - "hfSlug": "meta-llama/Llama-3.3-70B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-06T17:28:57.828422+00:00", + "slug": "google/gemini-2.5-pro-exp-03-25", + "hfSlug": "", + "updatedAt": "2025-05-12T16:02:53.80897+00:00", + "createdAt": "2025-03-25T17:01:39.919989+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.3 70B Instruct", - "shortName": "Llama 3.3 70B Instruct", - "author": "meta-llama", - "description": "The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.\n\nSupported languages: English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.\n\n[Model Card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/MODEL_CARD.md)", - "modelVersionGroupId": "397604e2-45fa-454e-a85d-9921f5138747", - "contextLength": 131072, + "name": "Google: Gemini 2.5 Pro Experimental", + "shortName": "Gemini 2.5 Pro Experimental", + "author": "google", + "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.", + "modelVersionGroupId": null, + "contextLength": 1048576, "inputModalities": [ - "text" + "text", + "image", + "file" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Gemini", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "meta-llama/llama-3.3-70b-instruct", + "warningMessage": "Update: Google has severly lowered rate limits on this model. Original: Due to extremely high demand, the [Gemini 2.5 Pro Experimental](/google/gemini-2.5-pro-exp-03-25) model is now strictly limited to **1 request per minute and 1000 requests per day** (including errors). Frequent `429` errors are expected. To maintain reliable performance, please switch to the [paid Gemini 2.5 Pro endpoint](/google/gemini-2.5-pro-preview-03-25). Your credits (such as the $10 minimum purchase) can be used directly on the paid endpoint without affecting your free-tier quotas.", + "permaslug": "google/gemini-2.5-pro-exp-03-25", "reasoningConfig": null, - "features": {}, + "features": null, "endpoint": { - "id": "e3b0a527-44d6-4ea6-9ec2-6a6416a84c7c", - "name": "DeepInfra | meta-llama/llama-3.3-70b-instruct", - "contextLength": 131072, + "id": "e5b7a905-dd02-4f6e-b61b-3aa1a1b8b9c8", + "name": "Google AI Studio | google/gemini-2.5-pro-exp-03-25", + "contextLength": 1048576, "model": { - "slug": "meta-llama/llama-3.3-70b-instruct", - "hfSlug": "meta-llama/Llama-3.3-70B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-06T17:28:57.828422+00:00", + "slug": "google/gemini-2.5-pro-exp-03-25", + "hfSlug": "", + "updatedAt": "2025-05-12T16:02:53.80897+00:00", + "createdAt": "2025-03-25T17:01:39.919989+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.3 70B Instruct", - "shortName": "Llama 3.3 70B Instruct", - "author": "meta-llama", - "description": "The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.\n\nSupported languages: English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.\n\n[Model Card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/MODEL_CARD.md)", - "modelVersionGroupId": "397604e2-45fa-454e-a85d-9921f5138747", - "contextLength": 131072, + "name": "Google: Gemini 2.5 Pro Experimental", + "shortName": "Gemini 2.5 Pro Experimental", + "author": "google", + "description": "Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.", + "modelVersionGroupId": null, + "contextLength": 1048576, "inputModalities": [ - "text" + "text", + "image", + "file" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Gemini", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "meta-llama/llama-3.3-70b-instruct", + "warningMessage": "Update: Google has severly lowered rate limits on this model. Original: Due to extremely high demand, the [Gemini 2.5 Pro Experimental](/google/gemini-2.5-pro-exp-03-25) model is now strictly limited to **1 request per minute and 1000 requests per day** (including errors). Frequent `429` errors are expected. To maintain reliable performance, please switch to the [paid Gemini 2.5 Pro endpoint](/google/gemini-2.5-pro-preview-03-25). Your credits (such as the $10 minimum purchase) can be used directly on the paid endpoint without affecting your free-tier quotas.", + "permaslug": "google/gemini-2.5-pro-exp-03-25", "reasoningConfig": null, - "features": {} + "features": null }, - "modelVariantSlug": "meta-llama/llama-3.3-70b-instruct", - "modelVariantPermaslug": "meta-llama/llama-3.3-70b-instruct", - "providerName": "DeepInfra", + "modelVariantSlug": "google/gemini-2.5-pro-exp-03-25", + "modelVariantPermaslug": "google/gemini-2.5-pro-exp-03-25", + "providerName": "Google AI Studio", "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", + "name": "Google AI Studio", + "displayName": "Google AI Studio", + "slug": "google-ai-studio", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://cloud.google.com/terms/", + "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 55 + }, + "freeModels": { + "training": true, + "retainsPrompts": true, + "retentionDays": 55 } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, + "hasCompletions": false, + "isAbortable": false, "moderationRequired": false, - "group": "DeepInfra", + "group": "Google AI Studio", "editors": [], "owners": [], "isMultipartSupported": true, @@ -2869,51 +2778,53 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/DeepInfra.webp" + "url": "/images/icons/GoogleAIStudio.svg" } }, - "providerDisplayName": "DeepInfra", - "providerModelId": "meta-llama/Llama-3.3-70B-Instruct-Turbo", - "providerGroup": "DeepInfra", - "quantization": "fp8", + "providerDisplayName": "Google AI Studio", + "providerModelId": "gemini-2.5-pro-exp-03-25", + "providerGroup": "Google AI Studio", + "quantization": null, "variant": "standard", - "isFree": false, - "canAbort": true, + "isFree": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 16384, + "maxCompletionTokens": 65536, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "tools", + "tool_choice", "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", "seed", - "min_p" + "response_format", + "structured_outputs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://cloud.google.com/terms/", + "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 55 }, - "training": false, - "retainsPrompts": false + "freeModels": { + "training": true, + "retainsPrompts": true, + "retentionDays": 55 + }, + "training": true, + "retainsPrompts": true, + "retentionDays": 55 }, "pricing": { - "prompt": "0.00000008", - "completion": "0.00000025", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -2927,34 +2838,37 @@ export default { "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, + "limitRpm": 1, + "limitRpd": 2000, + "hasCompletions": false, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 8, - "newest": 157, - "throughputHighToLow": 231, - "latencyLowToHigh": 19, - "pricingLowToHigh": 56, - "pricingHighToLow": 208 + "topWeekly": 7, + "newest": 72, + "throughputHighToLow": 308, + "latencyLowToHigh": 315, + "pricingLowToHigh": 31, + "pricingHighToLow": 279 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/Llama-3.3-70B-Instruct-Turbo", + "name": "Google AI Studio", + "icon": "https://openrouter.ai/images/icons/GoogleAIStudio.svg", + "slug": "google", + "quantization": null, + "context": 1048576, + "maxCompletionTokens": 65536, + "providerModelId": "gemini-2.5-pro-exp-03-25", "pricing": { - "prompt": "0.00000008", - "completion": "0.00000025", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -2962,33 +2876,32 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "tools", + "tool_choice", "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", "seed", - "min_p" + "response_format", + "structured_outputs" ], - "inputCost": 0.08, - "outputCost": 0.25 + "inputCost": 0, + "outputCost": 0, + "throughput": 178.311, + "latency": 7343 }, { - "name": "kluster.ai", - "slug": "klusterAi", - "quantization": "fp8", - "context": 131000, - "maxCompletionTokens": 131000, - "providerModelId": "klusterai/Meta-Llama-3.3-70B-Instruct-Turbo", + "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", + "slug": "vertex", + "quantization": null, + "context": 1048576, + "maxCompletionTokens": 65535, + "providerModelId": "gemini-2.5-pro-exp-03-25", "pricing": { - "prompt": "0.00000008", - "completion": "0.00000035", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -2997,20 +2910,243 @@ export default { }, "supportedParameters": [ "max_tokens", - "temperature" + "temperature", + "top_p", + "tools", + "tool_choice", + "stop", + "seed", + "response_format", + "structured_outputs" ], - "inputCost": 0.08, - "outputCost": 0.35 - }, - { - "name": "inference.net", - "slug": "inferenceNet", - "quantization": "fp8", - "context": 128000, + "inputCost": 0, + "outputCost": 0, + "throughput": 483.67, + "latency": 16045 + } + ] + }, + { + "slug": "meta-llama/llama-3.3-70b-instruct", + "hfSlug": "meta-llama/Llama-3.3-70B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-12-06T17:28:57.828422+00:00", + "hfUpdatedAt": null, + "name": "Meta: Llama 3.3 70B Instruct", + "shortName": "Llama 3.3 70B Instruct", + "author": "meta-llama", + "description": "The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.\n\nSupported languages: English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.\n\n[Model Card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/MODEL_CARD.md)", + "modelVersionGroupId": "397604e2-45fa-454e-a85d-9921f5138747", + "contextLength": 131000, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Llama3", + "instructType": "llama3", + "defaultSystem": null, + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "meta-llama/llama-3.3-70b-instruct", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "f85e853d-675d-4025-a0e1-1a0798723540", + "name": "Kluster | meta-llama/llama-3.3-70b-instruct", + "contextLength": 131000, + "model": { + "slug": "meta-llama/llama-3.3-70b-instruct", + "hfSlug": "meta-llama/Llama-3.3-70B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-12-06T17:28:57.828422+00:00", + "hfUpdatedAt": null, + "name": "Meta: Llama 3.3 70B Instruct", + "shortName": "Llama 3.3 70B Instruct", + "author": "meta-llama", + "description": "The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.\n\nSupported languages: English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.\n\n[Model Card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/MODEL_CARD.md)", + "modelVersionGroupId": "397604e2-45fa-454e-a85d-9921f5138747", + "contextLength": 131072, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Llama3", + "instructType": "llama3", + "defaultSystem": null, + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "meta-llama/llama-3.3-70b-instruct", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "meta-llama/llama-3.3-70b-instruct", + "modelVariantPermaslug": "meta-llama/llama-3.3-70b-instruct", + "providerName": "Kluster", + "providerInfo": { + "name": "Kluster", + "displayName": "kluster.ai", + "slug": "klusterai", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://www.kluster.ai/terms-of-use", + "privacyPolicyUrl": "https://www.kluster.ai/privacy-policy", + "paidModels": { + "training": false, + "retainsPrompts": false + } + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": false, + "isAbortable": true, + "moderationRequired": false, + "group": "Kluster", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256" + } + }, + "providerDisplayName": "kluster.ai", + "providerModelId": "klusterai/Meta-Llama-3.3-70B-Instruct-Turbo", + "providerGroup": "Kluster", + "quantization": "fp8", + "variant": "standard", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": 131000, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", + "seed" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://www.kluster.ai/terms-of-use", + "privacyPolicyUrl": "https://www.kluster.ai/privacy-policy", + "paidModels": { + "training": false, + "retainsPrompts": false + }, + "training": false, + "retainsPrompts": false + }, + "pricing": { + "prompt": "0.00000007", + "completion": "0.00000033", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": false, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 8, + "newest": 157, + "throughputHighToLow": 235, + "latencyLowToHigh": 88, + "pricingLowToHigh": 56, + "pricingHighToLow": 211 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", + "providers": [ + { + "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", + "slug": "klusterAi", + "quantization": "fp8", + "context": 131000, + "maxCompletionTokens": 131000, + "providerModelId": "klusterai/Meta-Llama-3.3-70B-Instruct-Turbo", + "pricing": { + "prompt": "0.00000007", + "completion": "0.00000033", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", + "seed" + ], + "inputCost": 0.07, + "outputCost": 0.33, + "throughput": 33.095, + "latency": 571.5 + }, + { + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "fp8", + "context": 131072, "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/llama-3.3-70b-instruct/fp-8", + "providerModelId": "meta-llama/Llama-3.3-70B-Instruct-Turbo", "pricing": { - "prompt": "0.0000001", + "prompt": "0.00000008", "completion": "0.00000025", "image": "0", "request": "0", @@ -3019,25 +3155,28 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "structured_outputs", - "response_format", "stop", "frequency_penalty", "presence_penalty", - "seed", + "repetition_penalty", + "response_format", "top_k", - "min_p", - "logit_bias", - "top_logprobs" + "seed", + "min_p" ], - "inputCost": 0.1, - "outputCost": 0.25 + "inputCost": 0.08, + "outputCost": 0.25, + "throughput": 34.0525, + "latency": 268 }, { "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", "slug": "lambda", "quantization": "fp8", "context": 131072, @@ -3066,10 +3205,13 @@ export default { "response_format" ], "inputCost": 0.12, - "outputCost": 0.3 + "outputCost": 0.3, + "throughput": 60.483, + "latency": 510 }, { "name": "Phala", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://phala.network/&size=256", "slug": "phala", "quantization": null, "context": 131072, @@ -3099,10 +3241,13 @@ export default { "repetition_penalty" ], "inputCost": 0.12, - "outputCost": 0.35 + "outputCost": 0.35, + "throughput": 29.689, + "latency": 749 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "bf16", "context": 131072, @@ -3133,10 +3278,13 @@ export default { "logit_bias" ], "inputCost": 0.13, - "outputCost": 0.39 + "outputCost": 0.39, + "throughput": 76.6895, + "latency": 783 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 131072, @@ -3165,10 +3313,13 @@ export default { "top_logprobs" ], "inputCost": 0.13, - "outputCost": 0.4 + "outputCost": 0.4, + "throughput": 38.8, + "latency": 660 }, { "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", "slug": "parasail", "quantization": "fp8", "context": 131072, @@ -3193,10 +3344,13 @@ export default { "top_k" ], "inputCost": 0.28, - "outputCost": 0.78 + "outputCost": 0.78, + "throughput": 81.678, + "latency": 542 }, { "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", "slug": "cloudflare", "quantization": "fp8", "context": 24000, @@ -3222,10 +3376,13 @@ export default { "presence_penalty" ], "inputCost": 0.29, - "outputCost": 2.25 + "outputCost": 2.25, + "throughput": 35.689, + "latency": 658 }, { "name": "CentML", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://centml.ai/&size=256", "slug": "centMl", "quantization": "bf16", "context": 131072, @@ -3255,10 +3412,13 @@ export default { "repetition_penalty" ], "inputCost": 0.35, - "outputCost": 0.35 + "outputCost": 0.35, + "throughput": 70.0325, + "latency": 734.5 }, { "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", "slug": "hyperbolic", "quantization": "fp8", "context": 131072, @@ -3289,10 +3449,13 @@ export default { "repetition_penalty" ], "inputCost": 0.4, - "outputCost": 0.4 + "outputCost": 0.4, + "throughput": 36.439, + "latency": 1267 }, { "name": "Atoma", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://atoma.network/&size=256", "slug": "atoma", "quantization": "fp8", "context": 104962, @@ -3313,10 +3476,13 @@ export default { "top_p" ], "inputCost": 0.4, - "outputCost": 0.4 + "outputCost": 0.4, + "throughput": 28.218, + "latency": 936 }, { "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", "slug": "groq", "quantization": null, "context": 131072, @@ -3347,10 +3513,13 @@ export default { "seed" ], "inputCost": 0.59, - "outputCost": 0.79 + "outputCost": 0.79, + "throughput": 343.6785, + "latency": 463 }, { "name": "Friendli", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://friendli.ai/&size=256", "slug": "friendli", "quantization": null, "context": 131072, @@ -3382,10 +3551,13 @@ export default { "structured_outputs" ], "inputCost": 0.6, - "outputCost": 0.6 + "outputCost": 0.6, + "throughput": 95.6525, + "latency": 818 }, { "name": "NextBit", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nextbit256.com/&size=256", "slug": "nextBit", "quantization": "fp8", "context": 32768, @@ -3411,10 +3583,13 @@ export default { "structured_outputs" ], "inputCost": 0.6, - "outputCost": 0.75 + "outputCost": 0.75, + "throughput": 32.64, + "latency": 2383 }, { "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", "slug": "sambaNova", "quantization": "bf16", "context": 131072, @@ -3437,10 +3612,13 @@ export default { "stop" ], "inputCost": 0.6, - "outputCost": 1.2 + "outputCost": 1.2, + "throughput": 309.927, + "latency": 2159.5 }, { "name": "Cerebras", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cerebras.ai/&size=256", "slug": "cerebras", "quantization": "fp16", "context": 32000, @@ -3469,10 +3647,13 @@ export default { "top_logprobs" ], "inputCost": 0.85, - "outputCost": 1.2 + "outputCost": 1.2, + "throughput": 4775.985, + "latency": 279 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "fp8", "context": 131072, @@ -3501,10 +3682,13 @@ export default { "response_format" ], "inputCost": 0.88, - "outputCost": 0.88 + "outputCost": 0.88, + "throughput": 94.488, + "latency": 559 }, { "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", "slug": "fireworks", "quantization": "fp16", "context": 131072, @@ -3535,216 +3719,23 @@ export default { "top_logprobs" ], "inputCost": 0.9, - "outputCost": 0.9 - } - ] - }, - { - "slug": "google/gemini-flash-1.5-8b", - "hfSlug": null, - "updatedAt": "2025-04-04T21:52:07.548792+00:00", - "createdAt": "2024-10-03T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Google: Gemini 1.5 Flash 8B", - "shortName": "Gemini 1.5 Flash 8B", - "author": "google", - "description": "Gemini Flash 1.5 8B is optimized for speed and efficiency, offering enhanced performance in small prompt tasks like chat, transcription, and translation. With reduced latency, it is highly effective for real-time and large-scale operations. This model focuses on cost-effective solutions while maintaining high-quality results.\n\n[Click here to learn more about this model](https://developers.googleblog.com/en/gemini-15-flash-8b-is-now-generally-available-for-use/).\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).", - "modelVersionGroupId": "3a412ab9-b077-48de-884c-90843f7abbf2", - "contextLength": 1000000, - "inputModalities": [ - "text", - "image" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Gemini", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "google/gemini-flash-1.5-8b", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "6f393253-cfcc-468e-a875-f5747a647953", - "name": "Google AI Studio | google/gemini-flash-1.5-8b", - "contextLength": 1000000, - "model": { - "slug": "google/gemini-flash-1.5-8b", - "hfSlug": null, - "updatedAt": "2025-04-04T21:52:07.548792+00:00", - "createdAt": "2024-10-03T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Google: Gemini 1.5 Flash 8B", - "shortName": "Gemini 1.5 Flash 8B", - "author": "google", - "description": "Gemini Flash 1.5 8B is optimized for speed and efficiency, offering enhanced performance in small prompt tasks like chat, transcription, and translation. With reduced latency, it is highly effective for real-time and large-scale operations. This model focuses on cost-effective solutions while maintaining high-quality results.\n\n[Click here to learn more about this model](https://developers.googleblog.com/en/gemini-15-flash-8b-is-now-generally-available-for-use/).\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).", - "modelVersionGroupId": "3a412ab9-b077-48de-884c-90843f7abbf2", - "contextLength": 1000000, - "inputModalities": [ - "text", - "image" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Gemini", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "google/gemini-flash-1.5-8b", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "google/gemini-flash-1.5-8b", - "modelVariantPermaslug": "google/gemini-flash-1.5-8b", - "providerName": "Google AI Studio", - "providerInfo": { - "name": "Google AI Studio", - "displayName": "Google AI Studio", - "slug": "google-ai-studio", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://cloud.google.com/terms/", - "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 55 - }, - "freeModels": { - "training": true, - "retainsPrompts": true, - "retentionDays": 55 - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, - "moderationRequired": false, - "group": "Google AI Studio", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/GoogleAIStudio.svg" - } - }, - "providerDisplayName": "Google AI Studio", - "providerModelId": "gemini-1.5-flash-8b-001", - "providerGroup": "Google AI Studio", - "quantization": "unknown", - "variant": "standard", - "isFree": false, - "canAbort": false, - "maxPromptTokens": null, - "maxCompletionTokens": 8192, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "tools", - "tool_choice", - "seed", - "response_format", - "structured_outputs" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://cloud.google.com/terms/", - "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 55 - }, - "freeModels": { - "training": true, - "retainsPrompts": true, - "retentionDays": 55 - }, - "training": false, - "retainsPrompts": true, - "retentionDays": 55 - }, - "pricing": { - "prompt": "0.0000000375", - "completion": "0.00000015", - "image": "0", - "request": "0", - "inputCacheRead": "0.00000001", - "inputCacheWrite": "0.0000000583", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [ - { - "type": "prompt-threshold", - "threshold": 128000, - "prompt": "0.000000075", - "completions": "0.0000003", - "inputCacheRead": "0.00000001", - "inputCacheWrite": "0.0000000583" - } - ], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": false, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null + "outputCost": 0.9, + "throughput": 109.5785, + "latency": 3019.5 }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 9, - "newest": 193, - "throughputHighToLow": 13, - "latencyLowToHigh": 10, - "pricingLowToHigh": 93, - "pricingHighToLow": 226 - }, - "providers": [ { - "name": "Google AI Studio", - "slug": "google", - "quantization": "unknown", - "context": 1000000, - "maxCompletionTokens": 8192, - "providerModelId": "gemini-1.5-flash-8b-001", + "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", + "slug": "inferenceNet", + "quantization": "fp8", + "context": 128000, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/llama-3.3-70b-instruct/fp-8", "pricing": { - "prompt": "0.0000000375", - "completion": "0.00000015", + "prompt": "0.0000001", + "completion": "0.00000025", "image": "0", "request": "0", - "inputCacheRead": "0.00000001", - "inputCacheWrite": "0.0000000583", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -3753,17 +3744,21 @@ export default { "max_tokens", "temperature", "top_p", + "structured_outputs", + "response_format", "stop", "frequency_penalty", "presence_penalty", - "tools", - "tool_choice", "seed", - "response_format", - "structured_outputs" + "top_k", + "min_p", + "logit_bias", + "top_logprobs" ], - "inputCost": 0.04, - "outputCost": 0.15 + "inputCost": 0.1, + "outputCost": 0.25, + "throughput": 13.164, + "latency": 1625 } ] }, @@ -3927,7 +3922,7 @@ export default { "supportsToolParameters": true, "supportsReasoning": true, "supportsMultipart": true, - "limitRpm": null, + "limitRpm": 500, "limitRpd": null, "hasCompletions": false, "hasChatCompletions": true, @@ -3940,14 +3935,16 @@ export default { "sorting": { "topWeekly": 1, "newest": 108, - "throughputHighToLow": 150, - "latencyLowToHigh": 204, + "throughputHighToLow": 148, + "latencyLowToHigh": 250, "pricingLowToHigh": 269, "pricingHighToLow": 41 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", "slug": "vertex", "quantization": null, "context": 200000, @@ -3972,10 +3969,13 @@ export default { "tool_choice" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 56.618, + "latency": 1913 }, { "name": "Amazon Bedrock", + "icon": "https://openrouter.ai/images/icons/Bedrock.svg", "slug": "amazonBedrock", "quantization": null, "context": 200000, @@ -4000,10 +4000,13 @@ export default { "tool_choice" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 35.9425, + "latency": 1886.5 }, { "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", "slug": "anthropic", "quantization": null, "context": 200000, @@ -4028,10 +4031,13 @@ export default { "tool_choice" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 57.987, + "latency": 1626 }, { "name": "Google Vertex (Europe)", + "icon": "", "slug": "googleVertex (europe)", "quantization": null, "context": 200000, @@ -4056,7 +4062,497 @@ export default { "tool_choice" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 50.2515, + "latency": 2288.5 + } + ] + }, + { + "slug": "google/gemini-flash-1.5-8b", + "hfSlug": null, + "updatedAt": "2025-04-04T21:52:07.548792+00:00", + "createdAt": "2024-10-03T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Google: Gemini 1.5 Flash 8B", + "shortName": "Gemini 1.5 Flash 8B", + "author": "google", + "description": "Gemini Flash 1.5 8B is optimized for speed and efficiency, offering enhanced performance in small prompt tasks like chat, transcription, and translation. With reduced latency, it is highly effective for real-time and large-scale operations. This model focuses on cost-effective solutions while maintaining high-quality results.\n\n[Click here to learn more about this model](https://developers.googleblog.com/en/gemini-15-flash-8b-is-now-generally-available-for-use/).\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).", + "modelVersionGroupId": "3a412ab9-b077-48de-884c-90843f7abbf2", + "contextLength": 1000000, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Gemini", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "google/gemini-flash-1.5-8b", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "6f393253-cfcc-468e-a875-f5747a647953", + "name": "Google AI Studio | google/gemini-flash-1.5-8b", + "contextLength": 1000000, + "model": { + "slug": "google/gemini-flash-1.5-8b", + "hfSlug": null, + "updatedAt": "2025-04-04T21:52:07.548792+00:00", + "createdAt": "2024-10-03T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Google: Gemini 1.5 Flash 8B", + "shortName": "Gemini 1.5 Flash 8B", + "author": "google", + "description": "Gemini Flash 1.5 8B is optimized for speed and efficiency, offering enhanced performance in small prompt tasks like chat, transcription, and translation. With reduced latency, it is highly effective for real-time and large-scale operations. This model focuses on cost-effective solutions while maintaining high-quality results.\n\n[Click here to learn more about this model](https://developers.googleblog.com/en/gemini-15-flash-8b-is-now-generally-available-for-use/).\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).", + "modelVersionGroupId": "3a412ab9-b077-48de-884c-90843f7abbf2", + "contextLength": 1000000, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Gemini", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "google/gemini-flash-1.5-8b", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "google/gemini-flash-1.5-8b", + "modelVariantPermaslug": "google/gemini-flash-1.5-8b", + "providerName": "Google AI Studio", + "providerInfo": { + "name": "Google AI Studio", + "displayName": "Google AI Studio", + "slug": "google-ai-studio", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://cloud.google.com/terms/", + "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 55 + }, + "freeModels": { + "training": true, + "retainsPrompts": true, + "retentionDays": 55 + } + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": false, + "isAbortable": false, + "moderationRequired": false, + "group": "Google AI Studio", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/GoogleAIStudio.svg" + } + }, + "providerDisplayName": "Google AI Studio", + "providerModelId": "gemini-1.5-flash-8b-001", + "providerGroup": "Google AI Studio", + "quantization": "unknown", + "variant": "standard", + "isFree": false, + "canAbort": false, + "maxPromptTokens": null, + "maxCompletionTokens": 8192, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "tools", + "tool_choice", + "seed", + "response_format", + "structured_outputs" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://cloud.google.com/terms/", + "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 55 + }, + "freeModels": { + "training": true, + "retainsPrompts": true, + "retentionDays": 55 + }, + "training": false, + "retainsPrompts": true, + "retentionDays": 55 + }, + "pricing": { + "prompt": "0.0000000375", + "completion": "0.00000015", + "image": "0", + "request": "0", + "inputCacheRead": "0.00000001", + "inputCacheWrite": "0.0000000583", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [ + { + "type": "prompt-threshold", + "threshold": 128000, + "prompt": "0.000000075", + "completions": "0.0000003", + "inputCacheRead": "0.00000001", + "inputCacheWrite": "0.0000000583" + } + ], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": true, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": false, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 10, + "newest": 193, + "throughputHighToLow": 13, + "latencyLowToHigh": 9, + "pricingLowToHigh": 94, + "pricingHighToLow": 225 + }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", + "providers": [ + { + "name": "Google AI Studio", + "icon": "https://openrouter.ai/images/icons/GoogleAIStudio.svg", + "slug": "google", + "quantization": "unknown", + "context": 1000000, + "maxCompletionTokens": 8192, + "providerModelId": "gemini-1.5-flash-8b-001", + "pricing": { + "prompt": "0.0000000375", + "completion": "0.00000015", + "image": "0", + "request": "0", + "inputCacheRead": "0.00000001", + "inputCacheWrite": "0.0000000583", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "tools", + "tool_choice", + "seed", + "response_format", + "structured_outputs" + ], + "inputCost": 0.04, + "outputCost": 0.15, + "throughput": 201.072, + "latency": 209 + } + ] + }, + { + "slug": "google/gemini-2.5-flash-preview", + "hfSlug": "", + "updatedAt": "2025-04-23T18:36:05.411237+00:00", + "createdAt": "2025-04-17T18:31:07.815979+00:00", + "hfUpdatedAt": null, + "name": "Google: Gemini 2.5 Flash Preview (thinking)", + "shortName": "Gemini 2.5 Flash Preview (thinking)", + "author": "google", + "description": "Gemini 2.5 Flash is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in \"thinking\" capabilities, enabling it to provide responses with greater accuracy and nuanced context handling. \n\nNote: This model is available in two variants: thinking and non-thinking. The output pricing varies significantly depending on whether the thinking capability is active. If you select the standard variant (without the \":thinking\" suffix), the model will explicitly avoid generating thinking tokens. \n\nTo utilize the thinking capability and receive thinking tokens, you must choose the \":thinking\" variant, which will then incur the higher thinking-output pricing. \n\nAdditionally, Gemini 2.5 Flash is configurable through the \"max tokens for reasoning\" parameter, as described in the documentation (https://openrouter.ai/docs/use-cases/reasoning-tokens#max-tokens-for-reasoning).", + "modelVersionGroupId": null, + "contextLength": 1048576, + "inputModalities": [ + "image", + "text", + "file" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Gemini", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "google/gemini-2.5-flash-preview-04-17", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "a5215739-cb87-49cb-adcb-c0c61c1e653e", + "name": "Google | google/gemini-2.5-flash-preview-04-17:thinking", + "contextLength": 1048576, + "model": { + "slug": "google/gemini-2.5-flash-preview", + "hfSlug": "", + "updatedAt": "2025-04-23T18:36:05.411237+00:00", + "createdAt": "2025-04-17T18:31:07.815979+00:00", + "hfUpdatedAt": null, + "name": "Google: Gemini 2.5 Flash Preview", + "shortName": "Gemini 2.5 Flash Preview", + "author": "google", + "description": "Gemini 2.5 Flash is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in \"thinking\" capabilities, enabling it to provide responses with greater accuracy and nuanced context handling. \n\nNote: This model is available in two variants: thinking and non-thinking. The output pricing varies significantly depending on whether the thinking capability is active. If you select the standard variant (without the \":thinking\" suffix), the model will explicitly avoid generating thinking tokens. \n\nTo utilize the thinking capability and receive thinking tokens, you must choose the \":thinking\" variant, which will then incur the higher thinking-output pricing. \n\nAdditionally, Gemini 2.5 Flash is configurable through the \"max tokens for reasoning\" parameter, as described in the documentation (https://openrouter.ai/docs/use-cases/reasoning-tokens#max-tokens-for-reasoning).", + "modelVersionGroupId": null, + "contextLength": 1048576, + "inputModalities": [ + "image", + "text", + "file" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Gemini", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "google/gemini-2.5-flash-preview-04-17", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "google/gemini-2.5-flash-preview:thinking", + "modelVariantPermaslug": "google/gemini-2.5-flash-preview-04-17:thinking", + "providerName": "Google", + "providerInfo": { + "name": "Google", + "displayName": "Vertex Thinking", + "slug": "google-vertex", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://cloud.google.com/terms/", + "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", + "dataPolicyUrl": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/abuse-monitoring", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "freeModels": { + "training": true, + "retainsPrompts": true + } + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": false, + "isAbortable": false, + "moderationRequired": false, + "group": "Google", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": "https://status.cloud.google.com/products/sdXM79fz1FS6ekNpu37K/history", + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/GoogleVertex.svg" + } + }, + "providerDisplayName": "Vertex Thinking", + "providerModelId": "gemini-2.5-flash-preview-04-17", + "providerGroup": "Google", + "quantization": null, + "variant": "thinking", + "isFree": false, + "canAbort": false, + "maxPromptTokens": null, + "maxCompletionTokens": 65535, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "tools", + "tool_choice", + "stop", + "response_format", + "structured_outputs" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://cloud.google.com/terms/", + "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", + "dataPolicyUrl": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/abuse-monitoring", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "freeModels": { + "training": true, + "retainsPrompts": true + }, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "pricing": { + "prompt": "0.00000015", + "completion": "0.0000035", + "image": "0.0006192", + "request": "0", + "inputCacheRead": "0.0000000375", + "inputCacheWrite": "0.0000002333", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": true, + "supportsReasoning": true, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": false, + "hasChatCompletions": true, + "features": { + "supportedParameters": { + "responseFormat": true, + "structuredOutputs": true + }, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 3, + "newest": 41, + "throughputHighToLow": 52, + "latencyLowToHigh": 92, + "pricingLowToHigh": 141, + "pricingHighToLow": 144 + }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", + "providers": [ + { + "name": "Vertex Non-Thinking", + "icon": "", + "slug": "vertexNonThinking", + "quantization": null, + "context": 1048576, + "maxCompletionTokens": 65535, + "providerModelId": "gemini-2.5-flash-preview-04-17", + "pricing": { + "prompt": "0.00000015", + "completion": "0.0000006", + "image": "0.0006192", + "request": "0", + "inputCacheRead": "0.0000000375", + "inputCacheWrite": "0.0000002333", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "tools", + "tool_choice", + "stop", + "response_format", + "structured_outputs" + ], + "inputCost": 0.15, + "outputCost": 0.6, + "throughput": 100.101, + "latency": 593 + }, + { + "name": "AI Studio Non-Thinking", + "icon": "", + "slug": "aiStudioNonThinking", + "quantization": null, + "context": 1048576, + "maxCompletionTokens": 65535, + "providerModelId": "gemini-2.5-flash-preview-04-17", + "pricing": { + "prompt": "0.00000015", + "completion": "0.0000006", + "image": "0.0006192", + "request": "0", + "inputCacheRead": "0.0000000375", + "inputCacheWrite": "0.0000002333", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "tools", + "tool_choice", + "stop", + "response_format", + "structured_outputs" + ], + "inputCost": 0.15, + "outputCost": 0.6, + "throughput": 127.5345, + "latency": 624 } ] }, @@ -4229,16 +4725,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 11, + "topWeekly": 12, "newest": 143, - "throughputHighToLow": 215, - "latencyLowToHigh": 154, + "throughputHighToLow": 171, + "latencyLowToHigh": 111, "pricingLowToHigh": 53, "pricingHighToLow": 124 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", "context": 163840, @@ -4269,10 +4767,13 @@ export default { "min_p" ], "inputCost": 0.5, - "outputCost": 2.18 + "outputCost": 2.18, + "throughput": 41.738, + "latency": 785 }, { "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", "slug": "inferenceNet", "quantization": "fp8", "context": 128000, @@ -4305,10 +4806,13 @@ export default { "structured_outputs" ], "inputCost": 0.5, - "outputCost": 3 + "outputCost": 3, + "throughput": 19.955, + "latency": 1274.5 }, { "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", "slug": "lambda", "quantization": "fp8", "context": 163840, @@ -4341,10 +4845,13 @@ export default { "response_format" ], "inputCost": 0.54, - "outputCost": 2.18 + "outputCost": 2.18, + "throughput": 36.8895, + "latency": 1027 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": null, "context": 64000, @@ -4377,10 +4884,13 @@ export default { "logit_bias" ], "inputCost": 0.7, - "outputCost": 2.5 + "outputCost": 2.5, + "throughput": 28.961, + "latency": 1270.5 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 163840, @@ -4411,10 +4921,13 @@ export default { "top_logprobs" ], "inputCost": 0.8, - "outputCost": 2.4 + "outputCost": 2.4, + "throughput": 26.319, + "latency": 797 }, { "name": "DeepInfra Turbo", + "icon": "", "slug": "deepInfraTurbo", "quantization": "fp4", "context": 32768, @@ -4445,10 +4958,13 @@ export default { "min_p" ], "inputCost": 1, - "outputCost": 3 + "outputCost": 3, + "throughput": 104.019, + "latency": 522 }, { "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", "slug": "klusterAi", "quantization": "fp8", "context": 163840, @@ -4466,14 +4982,28 @@ export default { "supportedParameters": [ "max_tokens", "temperature", + "top_p", "reasoning", - "include_reasoning" + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", + "seed" ], "inputCost": 1.75, - "outputCost": 5 + "outputCost": 5, + "throughput": 29.2855, + "latency": 962 }, { "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", "slug": "parasail", "quantization": "fp8", "context": 131072, @@ -4500,10 +5030,13 @@ export default { "top_k" ], "inputCost": 1.95, - "outputCost": 5 + "outputCost": 5, + "throughput": 52.0385, + "latency": 557 }, { "name": "Nebius Fast", + "icon": "", "slug": "nebiusFast", "quantization": "fp8", "context": 163840, @@ -4534,10 +5067,13 @@ export default { "top_logprobs" ], "inputCost": 2, - "outputCost": 6 + "outputCost": 6, + "throughput": 51.42, + "latency": 829 }, { "name": "CentML", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://centml.ai/&size=256", "slug": "centMl", "quantization": "fp8", "context": 131072, @@ -4567,10 +5103,13 @@ export default { "repetition_penalty" ], "inputCost": 2.99, - "outputCost": 2.99 + "outputCost": 2.99, + "throughput": 58.293, + "latency": 964 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "fp8", "context": 163840, @@ -4601,10 +5140,13 @@ export default { "response_format" ], "inputCost": 3, - "outputCost": 7 + "outputCost": 7, + "throughput": 49.032, + "latency": 1279 }, { "name": "Friendli", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://friendli.ai/&size=256", "slug": "friendli", "quantization": "fp8", "context": 163840, @@ -4638,10 +5180,13 @@ export default { "structured_outputs" ], "inputCost": 3, - "outputCost": 7 + "outputCost": 7, + "throughput": 59.985, + "latency": 927 }, { "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", "slug": "fireworks", "quantization": "fp8", "context": 163840, @@ -4674,10 +5219,13 @@ export default { "top_logprobs" ], "inputCost": 3, - "outputCost": 8 + "outputCost": 8, + "throughput": 44.893, + "latency": 736 }, { "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", "slug": "sambaNova", "quantization": "fp8", "context": 32768, @@ -4702,10 +5250,13 @@ export default { "stop" ], "inputCost": 5, - "outputCost": 7 + "outputCost": 7, + "throughput": 114.5975, + "latency": 2452 }, { "name": "Minimax", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://minimaxi.com/&size=256", "slug": "minimax", "quantization": null, "context": 64000, @@ -4728,10 +5279,13 @@ export default { "include_reasoning" ], "inputCost": 0.55, - "outputCost": 2.19 + "outputCost": 2.19, + "throughput": 21.5065, + "latency": 2109 }, { "name": "DeepSeek", + "icon": "https://openrouter.ai/images/icons/DeepSeek.png", "slug": "deepSeek", "quantization": "fp8", "context": 64000, @@ -4752,10 +5306,13 @@ export default { "include_reasoning" ], "inputCost": 0.55, - "outputCost": 2.19 + "outputCost": 2.19, + "throughput": 21.5405, + "latency": 4052 }, { "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", "slug": "azure", "quantization": null, "context": 163840, @@ -4778,10 +5335,13 @@ export default { "include_reasoning" ], "inputCost": 1.48, - "outputCost": 5.94 + "outputCost": 5.94, + "throughput": 78.2155, + "latency": 1780 }, { "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", "slug": "featherless", "quantization": "fp8", "context": 32768, @@ -4811,188 +5371,17 @@ export default { "seed" ], "inputCost": 6.5, - "outputCost": 8 + "outputCost": 8, + "throughput": 13.743, + "latency": 2659 } ] }, { - "slug": "google/gemini-2.0-flash-exp", + "slug": "openai/gpt-4.1", "hfSlug": "", - "updatedAt": "2025-04-04T21:51:57.421583+00:00", - "createdAt": "2024-12-11T17:18:43.999311+00:00", - "hfUpdatedAt": null, - "name": "Google: Gemini 2.0 Flash Experimental (free)", - "shortName": "Gemini 2.0 Flash Experimental (free)", - "author": "google", - "description": "Gemini Flash 2.0 offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5). It introduces notable enhancements in multimodal understanding, coding capabilities, complex instruction following, and function calling. These advancements come together to deliver more seamless and robust agentic experiences.", - "modelVersionGroupId": "e993dfbf-2cbd-4680-b866-c05bbdcc8f4d", - "contextLength": 1048576, - "inputModalities": [ - "text", - "image" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Gemini", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "google/gemini-2.0-flash-exp", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "65df650a-3eae-46b0-b5b0-87546ca90cc3", - "name": "Google | google/gemini-2.0-flash-exp:free", - "contextLength": 1048576, - "model": { - "slug": "google/gemini-2.0-flash-exp", - "hfSlug": "", - "updatedAt": "2025-04-04T21:51:57.421583+00:00", - "createdAt": "2024-12-11T17:18:43.999311+00:00", - "hfUpdatedAt": null, - "name": "Google: Gemini 2.0 Flash Experimental", - "shortName": "Gemini 2.0 Flash Experimental", - "author": "google", - "description": "Gemini Flash 2.0 offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5). It introduces notable enhancements in multimodal understanding, coding capabilities, complex instruction following, and function calling. These advancements come together to deliver more seamless and robust agentic experiences.", - "modelVersionGroupId": "e993dfbf-2cbd-4680-b866-c05bbdcc8f4d", - "contextLength": 1048576, - "inputModalities": [ - "text", - "image" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Gemini", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "google/gemini-2.0-flash-exp", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "google/gemini-2.0-flash-exp:free", - "modelVariantPermaslug": "google/gemini-2.0-flash-exp:free", - "providerName": "Google", - "providerInfo": { - "name": "Google", - "displayName": "Google Vertex", - "slug": "google-vertex", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://cloud.google.com/terms/", - "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", - "dataPolicyUrl": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/abuse-monitoring", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "freeModels": { - "training": true, - "retainsPrompts": true - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, - "moderationRequired": false, - "group": "Google", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": "https://status.cloud.google.com/products/sdXM79fz1FS6ekNpu37K/history", - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/GoogleVertex.svg" - } - }, - "providerDisplayName": "Google Vertex", - "providerModelId": "gemini-2.0-flash-exp", - "providerGroup": "Google", - "quantization": null, - "variant": "free", - "isFree": true, - "canAbort": false, - "maxPromptTokens": null, - "maxCompletionTokens": 8192, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://cloud.google.com/terms/", - "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", - "dataPolicyUrl": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/abuse-monitoring", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "freeModels": { - "training": true, - "retainsPrompts": true - }, - "training": true, - "retainsPrompts": true - }, - "pricing": { - "prompt": "0", - "completion": "0", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": false, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 12, - "newest": 156, - "throughputHighToLow": 26, - "latencyLowToHigh": 145, - "pricingLowToHigh": 55, - "pricingHighToLow": 303 - }, - "providers": [] - }, - { - "slug": "openai/gpt-4.1", - "hfSlug": "", - "updatedAt": "2025-05-12T18:09:29.635063+00:00", - "createdAt": "2025-04-14T17:23:05+00:00", + "updatedAt": "2025-05-12T18:09:29.635063+00:00", + "createdAt": "2025-04-14T17:23:05+00:00", "hfUpdatedAt": null, "name": "OpenAI: GPT-4.1", "shortName": "GPT-4.1", @@ -5181,14 +5570,16 @@ export default { "sorting": { "topWeekly": 13, "newest": 48, - "throughputHighToLow": 147, - "latencyLowToHigh": 111, + "throughputHighToLow": 119, + "latencyLowToHigh": 104, "pricingLowToHigh": 247, "pricingHighToLow": 67 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", "slug": "openAi", "quantization": null, "context": 1047576, @@ -5222,26 +5613,27 @@ export default { "structured_outputs" ], "inputCost": 2, - "outputCost": 8 + "outputCost": 8, + "throughput": 64.0385, + "latency": 684 } ] }, { - "slug": "google/gemini-2.5-flash-preview", + "slug": "google/gemini-2.0-flash-exp", "hfSlug": "", - "updatedAt": "2025-04-23T18:36:05.411237+00:00", - "createdAt": "2025-04-17T18:31:07.815979+00:00", + "updatedAt": "2025-04-04T21:51:57.421583+00:00", + "createdAt": "2024-12-11T17:18:43.999311+00:00", "hfUpdatedAt": null, - "name": "Google: Gemini 2.5 Flash Preview (thinking)", - "shortName": "Gemini 2.5 Flash Preview (thinking)", + "name": "Google: Gemini 2.0 Flash Experimental (free)", + "shortName": "Gemini 2.0 Flash Experimental (free)", "author": "google", - "description": "Gemini 2.5 Flash is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in \"thinking\" capabilities, enabling it to provide responses with greater accuracy and nuanced context handling. \n\nNote: This model is available in two variants: thinking and non-thinking. The output pricing varies significantly depending on whether the thinking capability is active. If you select the standard variant (without the \":thinking\" suffix), the model will explicitly avoid generating thinking tokens. \n\nTo utilize the thinking capability and receive thinking tokens, you must choose the \":thinking\" variant, which will then incur the higher thinking-output pricing. \n\nAdditionally, Gemini 2.5 Flash is configurable through the \"max tokens for reasoning\" parameter, as described in the documentation (https://openrouter.ai/docs/use-cases/reasoning-tokens#max-tokens-for-reasoning).", - "modelVersionGroupId": null, + "description": "Gemini Flash 2.0 offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5). It introduces notable enhancements in multimodal understanding, coding capabilities, complex instruction following, and function calling. These advancements come together to deliver more seamless and robust agentic experiences.", + "modelVersionGroupId": "e993dfbf-2cbd-4680-b866-c05bbdcc8f4d", "contextLength": 1048576, "inputModalities": [ - "image", "text", - "file" + "image" ], "outputModalities": [ "text" @@ -5254,29 +5646,28 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "google/gemini-2.5-flash-preview-04-17", + "permaslug": "google/gemini-2.0-flash-exp", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "a5215739-cb87-49cb-adcb-c0c61c1e653e", - "name": "Google | google/gemini-2.5-flash-preview-04-17:thinking", + "id": "65df650a-3eae-46b0-b5b0-87546ca90cc3", + "name": "Google | google/gemini-2.0-flash-exp:free", "contextLength": 1048576, "model": { - "slug": "google/gemini-2.5-flash-preview", + "slug": "google/gemini-2.0-flash-exp", "hfSlug": "", - "updatedAt": "2025-04-23T18:36:05.411237+00:00", - "createdAt": "2025-04-17T18:31:07.815979+00:00", + "updatedAt": "2025-04-04T21:51:57.421583+00:00", + "createdAt": "2024-12-11T17:18:43.999311+00:00", "hfUpdatedAt": null, - "name": "Google: Gemini 2.5 Flash Preview", - "shortName": "Gemini 2.5 Flash Preview", + "name": "Google: Gemini 2.0 Flash Experimental", + "shortName": "Gemini 2.0 Flash Experimental", "author": "google", - "description": "Gemini 2.5 Flash is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in \"thinking\" capabilities, enabling it to provide responses with greater accuracy and nuanced context handling. \n\nNote: This model is available in two variants: thinking and non-thinking. The output pricing varies significantly depending on whether the thinking capability is active. If you select the standard variant (without the \":thinking\" suffix), the model will explicitly avoid generating thinking tokens. \n\nTo utilize the thinking capability and receive thinking tokens, you must choose the \":thinking\" variant, which will then incur the higher thinking-output pricing. \n\nAdditionally, Gemini 2.5 Flash is configurable through the \"max tokens for reasoning\" parameter, as described in the documentation (https://openrouter.ai/docs/use-cases/reasoning-tokens#max-tokens-for-reasoning).", - "modelVersionGroupId": null, + "description": "Gemini Flash 2.0 offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5). It introduces notable enhancements in multimodal understanding, coding capabilities, complex instruction following, and function calling. These advancements come together to deliver more seamless and robust agentic experiences.", + "modelVersionGroupId": "e993dfbf-2cbd-4680-b866-c05bbdcc8f4d", "contextLength": 1048576, "inputModalities": [ - "image", "text", - "file" + "image" ], "outputModalities": [ "text" @@ -5289,16 +5680,16 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "google/gemini-2.5-flash-preview-04-17", + "permaslug": "google/gemini-2.0-flash-exp", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "google/gemini-2.5-flash-preview:thinking", - "modelVariantPermaslug": "google/gemini-2.5-flash-preview-04-17:thinking", + "modelVariantSlug": "google/gemini-2.0-flash-exp:free", + "modelVariantPermaslug": "google/gemini-2.0-flash-exp:free", "providerName": "Google", "providerInfo": { "name": "Google", - "displayName": "Vertex Thinking", + "displayName": "Google Vertex", "slug": "google-vertex", "baseUrl": "url", "dataPolicy": { @@ -5331,26 +5722,22 @@ export default { "url": "/images/icons/GoogleVertex.svg" } }, - "providerDisplayName": "Vertex Thinking", - "providerModelId": "gemini-2.5-flash-preview-04-17", + "providerDisplayName": "Google Vertex", + "providerModelId": "gemini-2.0-flash-exp", "providerGroup": "Google", "quantization": null, - "variant": "thinking", - "isFree": false, + "variant": "free", + "isFree": true, "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 65535, + "maxCompletionTokens": 8192, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "tools", - "tool_choice", - "stop", - "response_format", - "structured_outputs" + "stop" ], "isByok": false, "moderationRequired": false, @@ -5367,17 +5754,14 @@ export default { "training": true, "retainsPrompts": true }, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.00000015", - "completion": "0.0000035", - "image": "0.0006192", + "prompt": "0", + "completion": "0", + "image": "0", "request": "0", - "inputCacheRead": "0.0000000375", - "inputCacheWrite": "0.0000002333", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -5387,93 +5771,27 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": true, - "supportsReasoning": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": false, "hasChatCompletions": true, "features": { - "supportedParameters": { - "responseFormat": true, - "structuredOutputs": true - }, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 3, - "newest": 41, - "throughputHighToLow": 65, - "latencyLowToHigh": 76, - "pricingLowToHigh": 144, - "pricingHighToLow": 144 + "topWeekly": 14, + "newest": 156, + "throughputHighToLow": 15, + "latencyLowToHigh": 162, + "pricingLowToHigh": 55, + "pricingHighToLow": 303 }, - "providers": [ - { - "name": "Vertex Non-Thinking", - "slug": "vertexNonThinking", - "quantization": null, - "context": 1048576, - "maxCompletionTokens": 65535, - "providerModelId": "gemini-2.5-flash-preview-04-17", - "pricing": { - "prompt": "0.00000015", - "completion": "0.0000006", - "image": "0.0006192", - "request": "0", - "inputCacheRead": "0.0000000375", - "inputCacheWrite": "0.0000002333", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "tools", - "tool_choice", - "stop", - "response_format", - "structured_outputs" - ], - "inputCost": 0.15, - "outputCost": 0.6 - }, - { - "name": "AI Studio Non-Thinking", - "slug": "aiStudioNonThinking", - "quantization": null, - "context": 1048576, - "maxCompletionTokens": 65535, - "providerModelId": "gemini-2.5-flash-preview-04-17", - "pricing": { - "prompt": "0.00000015", - "completion": "0.0000006", - "image": "0.0006192", - "request": "0", - "inputCacheRead": "0.0000000375", - "inputCacheWrite": "0.0000002333", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "tools", - "tool_choice", - "stop", - "response_format", - "structured_outputs" - ], - "inputCost": 0.15, - "outputCost": 0.6 - } - ] + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", + "providers": [] }, { "slug": "mistralai/mistral-nemo", @@ -5486,7 +5804,7 @@ export default { "author": "mistralai", "description": "A 12B parameter model with a 128k token context length built by Mistral in collaboration with NVIDIA.\n\nThe model is multilingual, supporting English, French, German, Spanish, Italian, Portuguese, Chinese, Japanese, Korean, Arabic, and Hindi.\n\nIt supports function calling and is released under the Apache 2.0 license.", "modelVersionGroupId": null, - "contextLength": 98304, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -5508,9 +5826,9 @@ export default { "reasoningConfig": null, "features": {}, "endpoint": { - "id": "32bbf965-7c67-4f7c-b8a9-eccfc49bc906", - "name": "Enfer | mistralai/mistral-nemo", - "contextLength": 98304, + "id": "cf12e7bb-214e-4054-967f-d55aafda8a6f", + "name": "Kluster | mistralai/mistral-nemo", + "contextLength": 131072, "model": { "slug": "mistralai/mistral-nemo", "hfSlug": "mistralai/Mistral-Nemo-Instruct-2407", @@ -5546,24 +5864,26 @@ export default { }, "modelVariantSlug": "mistralai/mistral-nemo", "modelVariantPermaslug": "mistralai/mistral-nemo", - "providerName": "Enfer", + "providerName": "Kluster", "providerInfo": { - "name": "Enfer", - "displayName": "Enfer", - "slug": "enfer", + "name": "Kluster", + "displayName": "kluster.ai", + "slug": "klusterai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://enfer.ai/privacy-policy", - "privacyPolicyUrl": "https://enfer.ai/privacy-policy", + "termsOfServiceUrl": "https://www.kluster.ai/terms-of-use", + "privacyPolicyUrl": "https://www.kluster.ai/privacy-policy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false } }, + "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, + "hasCompletions": false, "isAbortable": true, "moderationRequired": false, - "group": "Enfer", + "group": "Kluster", "editors": [], "owners": [], "isMultipartSupported": true, @@ -5571,44 +5891,49 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://enfer.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256" } }, - "providerDisplayName": "Enfer", - "providerModelId": "mistralai/mistral-nemo", - "providerGroup": "Enfer", + "providerDisplayName": "kluster.ai", + "providerModelId": "mistralai/Mistral-Nemo-Instruct-2407", + "providerGroup": "Kluster", "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 49152, + "maxCompletionTokens": 131072, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", + "top_k", + "repetition_penalty", "logit_bias", - "logprobs" + "logprobs", + "top_logprobs", + "min_p", + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://enfer.ai/privacy-policy", - "privacyPolicyUrl": "https://enfer.ai/privacy-policy", + "termsOfServiceUrl": "https://www.kluster.ai/terms-of-use", + "privacyPolicyUrl": "https://www.kluster.ai/privacy-policy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false }, - "training": false + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0.00000003", + "prompt": "0.000000025", "completion": "0.00000007", "image": "0", "request": "0", @@ -5620,12 +5945,12 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { "supportedParameters": {}, @@ -5636,14 +5961,53 @@ export default { "sorting": { "topWeekly": 15, "newest": 234, - "throughputHighToLow": 155, - "latencyLowToHigh": 290, + "throughputHighToLow": 57, + "latencyLowToHigh": 61, "pricingLowToHigh": 68, - "pricingHighToLow": 233 + "pricingHighToLow": 236 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ + { + "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", + "slug": "klusterAi", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "mistralai/Mistral-Nemo-Instruct-2407", + "pricing": { + "prompt": "0.000000025", + "completion": "0.00000007", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", + "seed" + ], + "inputCost": 0.02, + "outputCost": 0.07, + "throughput": 119.605, + "latency": 470 + }, { "name": "Enfer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://enfer.ai/&size=256", "slug": "enfer", "quantization": null, "context": 131072, @@ -5671,10 +6035,45 @@ export default { "logprobs" ], "inputCost": 0.03, - "outputCost": 0.07 + "outputCost": 0.07, + "throughput": 53.943, + "latency": 6449.5 + }, + { + "name": "NextBit", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nextbit256.com/&size=256", + "slug": "nextBit", + "quantization": "bf16", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "mistral:nemo", + "pricing": { + "prompt": "0.000000033", + "completion": "0.00000007", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "structured_outputs" + ], + "inputCost": 0.03, + "outputCost": 0.07, + "throughput": 51.1245, + "latency": 1610 }, { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "bf16", "context": 131072, @@ -5703,10 +6102,13 @@ export default { "min_p" ], "inputCost": 0.04, - "outputCost": 0.08 + "outputCost": 0.08, + "throughput": 55.12, + "latency": 453 }, { "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", "slug": "inferenceNet", "quantization": "fp8", "context": 16384, @@ -5735,10 +6137,13 @@ export default { "top_logprobs" ], "inputCost": 0.04, - "outputCost": 0.1 + "outputCost": 0.1, + "throughput": 61.076, + "latency": 1211 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 128000, @@ -5767,10 +6172,13 @@ export default { "top_logprobs" ], "inputCost": 0.04, - "outputCost": 0.12 + "outputCost": 0.12, + "throughput": 34.408, + "latency": 657 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "unknown", "context": 131072, @@ -5801,39 +6209,13 @@ export default { "logit_bias" ], "inputCost": 0.04, - "outputCost": 0.17 - }, - { - "name": "NextBit", - "slug": "nextBit", - "quantization": "bf16", - "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "mistral:nemo", - "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "structured_outputs" - ], - "inputCost": 0.1, - "outputCost": 0.1 + "outputCost": 0.17, + "throughput": 36.1535, + "latency": 1354 }, { "name": "Atoma", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://atoma.network/&size=256", "slug": "atoma", "quantization": null, "context": 128000, @@ -5854,10 +6236,13 @@ export default { "top_p" ], "inputCost": 0.1, - "outputCost": 0.1 + "outputCost": 0.1, + "throughput": 52.1375, + "latency": 636 }, { "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", "slug": "parasail", "quantization": "fp8", "context": 131072, @@ -5882,10 +6267,13 @@ export default { "top_k" ], "inputCost": 0.11, - "outputCost": 0.11 + "outputCost": 0.11, + "throughput": 142.857, + "latency": 523 }, { "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", "slug": "mistral", "quantization": "unknown", "context": 131072, @@ -5914,10 +6302,13 @@ export default { "seed" ], "inputCost": 0.15, - "outputCost": 0.15 + "outputCost": 0.15, + "throughput": 143.466, + "latency": 212 }, { "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", "slug": "azure", "quantization": "unknown", "context": 128000, @@ -5943,65 +6334,309 @@ export default { "seed" ], "inputCost": 0.3, - "outputCost": 0.3 + "outputCost": 0.3, + "throughput": 98.261, + "latency": 1152 } ] }, { - "slug": "anthropic/claude-3.7-sonnet", + "slug": "openai/gpt-4.1-mini", "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-24T18:35:10.00008+00:00", + "updatedAt": "2025-05-12T18:45:53.698316+00:00", + "createdAt": "2025-04-14T17:23:01+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3.7 Sonnet (self-moderated)", - "shortName": "Claude 3.7 Sonnet (self-moderated)", - "author": "anthropic", - "description": "Claude 3.7 Sonnet is an advanced large language model with improved reasoning, coding, and problem-solving capabilities. It introduces a hybrid reasoning approach, allowing users to choose between rapid responses and extended, step-by-step processing for complex tasks. The model demonstrates notable improvements in coding, particularly in front-end development and full-stack updates, and excels in agentic workflows, where it can autonomously navigate multi-step processes. \n\nClaude 3.7 Sonnet maintains performance parity with its predecessor in standard mode while offering an extended reasoning mode for enhanced accuracy in math, coding, and instruction-following tasks.\n\nRead more at the [blog post here](https://www.anthropic.com/news/claude-3-7-sonnet)", - "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", - "contextLength": 200000, + "name": "OpenAI: GPT-4.1 Mini", + "shortName": "GPT-4.1 Mini", + "author": "openai", + "description": "GPT-4.1 Mini is a mid-sized model delivering performance competitive with GPT-4o at substantially lower latency and cost. It retains a 1 million token context window and scores 45.1% on hard instruction evals, 35.8% on MultiChallenge, and 84.1% on IFEval. Mini also shows strong coding ability (e.g., 31.6% on Aider’s polyglot diff benchmark) and vision understanding, making it suitable for interactive applications with tight performance constraints.", + "modelVersionGroupId": null, + "contextLength": 1047576, "inputModalities": [ + "image", "text", - "image" + "file" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-7-sonnet-20250219", + "permaslug": "openai/gpt-4.1-mini-2025-04-14", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "9770425a-9db8-46be-87fa-d66ad537aafc", - "name": "Anthropic | anthropic/claude-3-7-sonnet-20250219:beta", - "contextLength": 200000, + "id": "872eccb7-9c85-45fc-974a-ff7c8e2407e6", + "name": "OpenAI | openai/gpt-4.1-mini-2025-04-14", + "contextLength": 1047576, "model": { - "slug": "anthropic/claude-3.7-sonnet", + "slug": "openai/gpt-4.1-mini", "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-24T18:35:10.00008+00:00", + "updatedAt": "2025-05-12T18:45:53.698316+00:00", + "createdAt": "2025-04-14T17:23:01+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3.7 Sonnet", - "shortName": "Claude 3.7 Sonnet", - "author": "anthropic", - "description": "Claude 3.7 Sonnet is an advanced large language model with improved reasoning, coding, and problem-solving capabilities. It introduces a hybrid reasoning approach, allowing users to choose between rapid responses and extended, step-by-step processing for complex tasks. The model demonstrates notable improvements in coding, particularly in front-end development and full-stack updates, and excels in agentic workflows, where it can autonomously navigate multi-step processes. \n\nClaude 3.7 Sonnet maintains performance parity with its predecessor in standard mode while offering an extended reasoning mode for enhanced accuracy in math, coding, and instruction-following tasks.\n\nRead more at the [blog post here](https://www.anthropic.com/news/claude-3-7-sonnet)", - "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", - "contextLength": 200000, + "name": "OpenAI: GPT-4.1 Mini", + "shortName": "GPT-4.1 Mini", + "author": "openai", + "description": "GPT-4.1 Mini is a mid-sized model delivering performance competitive with GPT-4o at substantially lower latency and cost. It retains a 1 million token context window and scores 45.1% on hard instruction evals, 35.8% on MultiChallenge, and 84.1% on IFEval. Mini also shows strong coding ability (e.g., 31.6% on Aider’s polyglot diff benchmark) and vision understanding, making it suitable for interactive applications with tight performance constraints.", + "modelVersionGroupId": null, + "contextLength": 1047576, "inputModalities": [ + "image", "text", - "image" + "file" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", + "group": "GPT", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "openai/gpt-4.1-mini-2025-04-14", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "openai/gpt-4.1-mini", + "modelVariantPermaslug": "openai/gpt-4.1-mini-2025-04-14", + "providerName": "OpenAI", + "providerInfo": { + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": true, + "group": "OpenAI", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": "https://status.openai.com/", + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/OpenAI.svg", + "invertRequired": true + } + }, + "providerDisplayName": "OpenAI", + "providerModelId": "gpt-4.1-mini-2025-04-14", + "providerGroup": "OpenAI", + "quantization": null, + "variant": "standard", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": 32768, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "web_search_options", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" + ], + "isByok": false, + "moderationRequired": true, + "dataPolicy": { + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "pricing": { + "prompt": "0.0000004", + "completion": "0.0000016", + "image": "0", + "request": "0", + "inputCacheRead": "0.0000001", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [ + { + "type": "search-threshold", + "threshold": "high", + "request": "0.03" + }, + { + "type": "search-threshold", + "threshold": "medium", + "request": "0.0275" + }, + { + "type": "search-threshold", + "threshold": "low", + "request": "0.025" + } + ], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": true, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 16, + "newest": 49, + "throughputHighToLow": 122, + "latencyLowToHigh": 116, + "pricingLowToHigh": 180, + "pricingHighToLow": 138 + }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", + "providers": [ + { + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": null, + "context": 1047576, + "maxCompletionTokens": 32768, + "providerModelId": "gpt-4.1-mini-2025-04-14", + "pricing": { + "prompt": "0.0000004", + "completion": "0.0000016", + "image": "0", + "request": "0", + "inputCacheRead": "0.0000001", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "web_search_options", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" + ], + "inputCost": 0.4, + "outputCost": 1.6, + "throughput": 64.683, + "latency": 718 + } + ] + }, + { + "slug": "anthropic/claude-3.7-sonnet", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-24T18:35:10.00008+00:00", + "hfUpdatedAt": null, + "name": "Anthropic: Claude 3.7 Sonnet (self-moderated)", + "shortName": "Claude 3.7 Sonnet (self-moderated)", + "author": "anthropic", + "description": "Claude 3.7 Sonnet is an advanced large language model with improved reasoning, coding, and problem-solving capabilities. It introduces a hybrid reasoning approach, allowing users to choose between rapid responses and extended, step-by-step processing for complex tasks. The model demonstrates notable improvements in coding, particularly in front-end development and full-stack updates, and excels in agentic workflows, where it can autonomously navigate multi-step processes. \n\nClaude 3.7 Sonnet maintains performance parity with its predecessor in standard mode while offering an extended reasoning mode for enhanced accuracy in math, coding, and instruction-following tasks.\n\nRead more at the [blog post here](https://www.anthropic.com/news/claude-3-7-sonnet)", + "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", + "contextLength": 200000, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Claude", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "anthropic/claude-3-7-sonnet-20250219", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "9770425a-9db8-46be-87fa-d66ad537aafc", + "name": "Anthropic | anthropic/claude-3-7-sonnet-20250219:beta", + "contextLength": 200000, + "model": { + "slug": "anthropic/claude-3.7-sonnet", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-24T18:35:10.00008+00:00", + "hfUpdatedAt": null, + "name": "Anthropic: Claude 3.7 Sonnet", + "shortName": "Claude 3.7 Sonnet", + "author": "anthropic", + "description": "Claude 3.7 Sonnet is an advanced large language model with improved reasoning, coding, and problem-solving capabilities. It introduces a hybrid reasoning approach, allowing users to choose between rapid responses and extended, step-by-step processing for complex tasks. The model demonstrates notable improvements in coding, particularly in front-end development and full-stack updates, and excels in agentic workflows, where it can autonomously navigate multi-step processes. \n\nClaude 3.7 Sonnet maintains performance parity with its predecessor in standard mode while offering an extended reasoning mode for enhanced accuracy in math, coding, and instruction-following tasks.\n\nRead more at the [blog post here](https://www.anthropic.com/news/claude-3-7-sonnet)", + "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", + "contextLength": 200000, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Claude", "instructType": null, "defaultSystem": null, "defaultStops": [], @@ -6112,14 +6747,16 @@ export default { "sorting": { "topWeekly": 1, "newest": 108, - "throughputHighToLow": 150, - "latencyLowToHigh": 204, + "throughputHighToLow": 148, + "latencyLowToHigh": 250, "pricingLowToHigh": 269, "pricingHighToLow": 41 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", "slug": "vertex", "quantization": null, "context": 200000, @@ -6144,10 +6781,13 @@ export default { "tool_choice" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 56.618, + "latency": 1913 }, { "name": "Amazon Bedrock", + "icon": "https://openrouter.ai/images/icons/Bedrock.svg", "slug": "amazonBedrock", "quantization": null, "context": 200000, @@ -6172,10 +6812,13 @@ export default { "tool_choice" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 35.9425, + "latency": 1886.5 }, { "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", "slug": "anthropic", "quantization": null, "context": 200000, @@ -6200,10 +6843,13 @@ export default { "tool_choice" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 57.987, + "latency": 1626 }, { "name": "Google Vertex (Europe)", + "icon": "", "slug": "googleVertex (europe)", "quantization": null, "context": 200000, @@ -6228,7 +6874,9 @@ export default { "tool_choice" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 50.2515, + "latency": 2288.5 } ] }, @@ -6406,16 +7054,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 17, + "topWeekly": 18, "newest": 107, - "throughputHighToLow": 28, - "latencyLowToHigh": 48, - "pricingLowToHigh": 110, + "throughputHighToLow": 30, + "latencyLowToHigh": 59, + "pricingLowToHigh": 111, "pricingHighToLow": 207 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [ { "name": "Google AI Studio", + "icon": "https://openrouter.ai/images/icons/GoogleAIStudio.svg", "slug": "google", "quantization": null, "context": 1048576, @@ -6444,10 +7094,13 @@ export default { "structured_outputs" ], "inputCost": 0.07, - "outputCost": 0.3 + "outputCost": 0.3, + "throughput": 158.8315, + "latency": 442 }, { "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", "slug": "vertex", "quantization": null, "context": 1048576, @@ -6476,7 +7129,9 @@ export default { "structured_outputs" ], "inputCost": 0.07, - "outputCost": 0.3 + "outputCost": 0.3, + "throughput": 185.1125, + "latency": 1111 } ] }, @@ -6642,16 +7297,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 18, + "topWeekly": 19, "newest": 183, - "throughputHighToLow": 153, - "latencyLowToHigh": 250, + "throughputHighToLow": 137, + "latencyLowToHigh": 210, "pricingLowToHigh": 272, "pricingHighToLow": 44 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", "slug": "anthropic", "quantization": "unknown", "context": 200000, @@ -6678,10 +7335,13 @@ export default { "stop" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 57.5795, + "latency": 1371 }, { "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", "slug": "vertex", "quantization": "unknown", "context": 200000, @@ -6708,10 +7368,13 @@ export default { "stop" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 61.168, + "latency": 1377 }, { "name": "Amazon Bedrock", + "icon": "https://openrouter.ai/images/icons/Bedrock.svg", "slug": "amazonBedrock", "quantization": "unknown", "context": 200000, @@ -6736,10 +7399,13 @@ export default { "stop" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 35.7255, + "latency": 1781.5 }, { "name": "Amazon Bedrock", + "icon": "https://openrouter.ai/images/icons/Bedrock.svg", "slug": "amazonBedrock", "quantization": null, "context": 200000, @@ -6766,10 +7432,13 @@ export default { "stop" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 37.866, + "latency": 1632 }, { "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", "slug": "vertex", "quantization": "unknown", "context": 200000, @@ -6796,245 +7465,9 @@ export default { "stop" ], "inputCost": 3, - "outputCost": 15 - } - ] - }, - { - "slug": "openai/gpt-4.1-mini", - "hfSlug": "", - "updatedAt": "2025-05-12T18:45:53.698316+00:00", - "createdAt": "2025-04-14T17:23:01+00:00", - "hfUpdatedAt": null, - "name": "OpenAI: GPT-4.1 Mini", - "shortName": "GPT-4.1 Mini", - "author": "openai", - "description": "GPT-4.1 Mini is a mid-sized model delivering performance competitive with GPT-4o at substantially lower latency and cost. It retains a 1 million token context window and scores 45.1% on hard instruction evals, 35.8% on MultiChallenge, and 84.1% on IFEval. Mini also shows strong coding ability (e.g., 31.6% on Aider’s polyglot diff benchmark) and vision understanding, making it suitable for interactive applications with tight performance constraints.", - "modelVersionGroupId": null, - "contextLength": 1047576, - "inputModalities": [ - "image", - "text", - "file" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "GPT", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "openai/gpt-4.1-mini-2025-04-14", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "872eccb7-9c85-45fc-974a-ff7c8e2407e6", - "name": "OpenAI | openai/gpt-4.1-mini-2025-04-14", - "contextLength": 1047576, - "model": { - "slug": "openai/gpt-4.1-mini", - "hfSlug": "", - "updatedAt": "2025-05-12T18:45:53.698316+00:00", - "createdAt": "2025-04-14T17:23:01+00:00", - "hfUpdatedAt": null, - "name": "OpenAI: GPT-4.1 Mini", - "shortName": "GPT-4.1 Mini", - "author": "openai", - "description": "GPT-4.1 Mini is a mid-sized model delivering performance competitive with GPT-4o at substantially lower latency and cost. It retains a 1 million token context window and scores 45.1% on hard instruction evals, 35.8% on MultiChallenge, and 84.1% on IFEval. Mini also shows strong coding ability (e.g., 31.6% on Aider’s polyglot diff benchmark) and vision understanding, making it suitable for interactive applications with tight performance constraints.", - "modelVersionGroupId": null, - "contextLength": 1047576, - "inputModalities": [ - "image", - "text", - "file" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "GPT", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "openai/gpt-4.1-mini-2025-04-14", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "openai/gpt-4.1-mini", - "modelVariantPermaslug": "openai/gpt-4.1-mini-2025-04-14", - "providerName": "OpenAI", - "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true - } - }, - "providerDisplayName": "OpenAI", - "providerModelId": "gpt-4.1-mini-2025-04-14", - "providerGroup": "OpenAI", - "quantization": null, - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 32768, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "web_search_options", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" - ], - "isByok": false, - "moderationRequired": true, - "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "pricing": { - "prompt": "0.0000004", - "completion": "0.0000016", - "image": "0", - "request": "0", - "inputCacheRead": "0.0000001", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [ - { - "type": "search-threshold", - "threshold": "high", - "request": "0.03" - }, - { - "type": "search-threshold", - "threshold": "medium", - "request": "0.0275" - }, - { - "type": "search-threshold", - "threshold": "low", - "request": "0.025" - } - ], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 19, - "newest": 49, - "throughputHighToLow": 127, - "latencyLowToHigh": 117, - "pricingLowToHigh": 180, - "pricingHighToLow": 138 - }, - "providers": [ - { - "name": "OpenAI", - "slug": "openAi", - "quantization": null, - "context": 1047576, - "maxCompletionTokens": 32768, - "providerModelId": "gpt-4.1-mini-2025-04-14", - "pricing": { - "prompt": "0.0000004", - "completion": "0.0000016", - "image": "0", - "request": "0", - "inputCacheRead": "0.0000001", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "web_search_options", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" - ], - "inputCost": 0.4, - "outputCost": 1.6 + "outputCost": 15, + "throughput": 59.922, + "latency": 1022 } ] }, @@ -7224,14 +7657,16 @@ export default { "sorting": { "topWeekly": 20, "newest": 257, - "throughputHighToLow": 34, - "latencyLowToHigh": 40, + "throughputHighToLow": 37, + "latencyLowToHigh": 37, "pricingLowToHigh": 112, - "pricingHighToLow": 209 + "pricingHighToLow": 208 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [ { "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", "slug": "vertex", "quantization": "unknown", "context": 1000000, @@ -7262,10 +7697,13 @@ export default { "structured_outputs" ], "inputCost": 0.07, - "outputCost": 0.3 + "outputCost": 0.3, + "throughput": 148.56, + "latency": 355 }, { "name": "Google AI Studio", + "icon": "https://openrouter.ai/images/icons/GoogleAIStudio.svg", "slug": "google", "quantization": "unknown", "context": 1000000, @@ -7296,7 +7734,9 @@ export default { "structured_outputs" ], "inputCost": 0.07, - "outputCost": 0.3 + "outputCost": 0.3, + "throughput": 150.492, + "latency": 395 } ] }, @@ -7485,14 +7925,16 @@ export default { "sorting": { "topWeekly": 21, "newest": 141, - "throughputHighToLow": 170, - "latencyLowToHigh": 90, + "throughputHighToLow": 162, + "latencyLowToHigh": 39, "pricingLowToHigh": 52, "pricingHighToLow": 191 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", "context": 131072, @@ -7523,10 +7965,13 @@ export default { "min_p" ], "inputCost": 0.1, - "outputCost": 0.4 + "outputCost": 0.4, + "throughput": 34.986, + "latency": 459.5 }, { "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", "slug": "inferenceNet", "quantization": null, "context": 128000, @@ -7559,10 +8004,13 @@ export default { "structured_outputs" ], "inputCost": 0.1, - "outputCost": 0.4 + "outputCost": 0.4, + "throughput": 37.947, + "latency": 754 }, { "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", "slug": "lambda", "quantization": "fp8", "context": 131072, @@ -7593,10 +8041,13 @@ export default { "response_format" ], "inputCost": 0.2, - "outputCost": 0.6 + "outputCost": 0.6, + "throughput": 67.7255, + "latency": 444 }, { "name": "Phala", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://phala.network/&size=256", "slug": "phala", "quantization": null, "context": 131072, @@ -7626,10 +8077,13 @@ export default { "repetition_penalty" ], "inputCost": 0.2, - "outputCost": 0.7 + "outputCost": 0.7, + "throughput": 36.5945, + "latency": 608 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 131072, @@ -7660,10 +8114,13 @@ export default { "top_logprobs" ], "inputCost": 0.25, - "outputCost": 0.75 + "outputCost": 0.75, + "throughput": 59.612, + "latency": 484 }, { "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", "slug": "sambaNova", "quantization": "bf16", "context": 131072, @@ -7688,10 +8145,13 @@ export default { "stop" ], "inputCost": 0.7, - "outputCost": 1.4 + "outputCost": 1.4, + "throughput": 232.5305, + "latency": 2045.5 }, { "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", "slug": "groq", "quantization": null, "context": 131072, @@ -7724,10 +8184,13 @@ export default { "seed" ], "inputCost": 0.75, - "outputCost": 0.99 + "outputCost": 0.99, + "throughput": 293.145, + "latency": 881 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": null, "context": 32000, @@ -7758,10 +8221,13 @@ export default { "logit_bias" ], "inputCost": 0.8, - "outputCost": 0.8 + "outputCost": 0.8, + "throughput": 71.3745, + "latency": 891 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": null, "context": 131072, @@ -7792,10 +8258,13 @@ export default { "response_format" ], "inputCost": 2, - "outputCost": 2 + "outputCost": 2, + "throughput": 101.26, + "latency": 690.5 }, { "name": "Cerebras", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cerebras.ai/&size=256", "slug": "cerebras", "quantization": null, "context": 32000, @@ -7826,22 +8295,24 @@ export default { "top_logprobs" ], "inputCost": 2.2, - "outputCost": 2.5 + "outputCost": 2.5, + "throughput": 2122.065, + "latency": 865 } ] }, { - "slug": "qwen/qwen3-235b-a22b", - "hfSlug": "Qwen/Qwen3-235B-A22B", - "updatedAt": "2025-05-11T22:44:42.939095+00:00", - "createdAt": "2025-04-28T21:29:17.25671+00:00", + "slug": "qwen/qwen-2.5-7b-instruct", + "hfSlug": "Qwen/Qwen2.5-7B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-10-16T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 235B A22B", - "shortName": "Qwen3 235B A22B", + "name": "Qwen2.5 7B Instruct", + "shortName": "Qwen2.5 7B Instruct", "author": "qwen", - "description": "Qwen3-235B-A22B is a 235B parameter mixture-of-experts (MoE) model developed by Qwen, activating 22B parameters per forward pass. It supports seamless switching between a \"thinking\" mode for complex reasoning, math, and code tasks, and a \"non-thinking\" mode for general conversational efficiency. The model demonstrates strong reasoning ability, multilingual support (100+ languages and dialects), advanced instruction-following, and agent tool-calling capabilities. It natively handles a 32K token context window and extends up to 131K tokens using YaRN-based scaling.", + "description": "Qwen2.5 7B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", "modelVersionGroupId": null, - "contextLength": 40960, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -7849,41 +8320,34 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Qwen", + "instructType": "chatml", "defaultSystem": null, "defaultStops": [ "<|im_start|>", - "<|im_end|>" + "<|im_end|>", + "<|endoftext|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen3-235b-a22b-04-28", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "permaslug": "qwen/qwen-2.5-7b-instruct", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "5f108245-8fec-4702-859d-56c5b54a71d7", - "name": "DeepInfra | qwen/qwen3-235b-a22b-04-28", - "contextLength": 40960, + "id": "7b70cad4-5c0f-4bfc-8317-17e03eb17c8b", + "name": "NextBit | qwen/qwen-2.5-7b-instruct", + "contextLength": 32768, "model": { - "slug": "qwen/qwen3-235b-a22b", - "hfSlug": "Qwen/Qwen3-235B-A22B", - "updatedAt": "2025-05-11T22:44:42.939095+00:00", - "createdAt": "2025-04-28T21:29:17.25671+00:00", + "slug": "qwen/qwen-2.5-7b-instruct", + "hfSlug": "Qwen/Qwen2.5-7B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-10-16T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 235B A22B", - "shortName": "Qwen3 235B A22B", + "name": "Qwen2.5 7B Instruct", + "shortName": "Qwen2.5 7B Instruct", "author": "qwen", - "description": "Qwen3-235B-A22B is a 235B parameter mixture-of-experts (MoE) model developed by Qwen, activating 22B parameters per forward pass. It supports seamless switching between a \"thinking\" mode for complex reasoning, math, and code tasks, and a \"non-thinking\" mode for general conversational efficiency. The model demonstrates strong reasoning ability, multilingual support (100+ languages and dialects), advanced instruction-following, and agent tool-calling capabilities. It natively handles a 32K token context window and extends up to 131K tokens using YaRN-based scaling.", + "description": "Qwen2.5 7B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ @@ -7893,30 +8357,326 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Qwen", + "instructType": "chatml", "defaultSystem": null, "defaultStops": [ "<|im_start|>", - "<|im_end|>" + "<|im_end|>", + "<|endoftext|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen3-235b-a22b-04-28", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" + "permaslug": "qwen/qwen-2.5-7b-instruct", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "qwen/qwen-2.5-7b-instruct", + "modelVariantPermaslug": "qwen/qwen-2.5-7b-instruct", + "providerName": "NextBit", + "providerInfo": { + "name": "NextBit", + "displayName": "NextBit", + "slug": "nextbit", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://www.nextbit256.com/docs/terms-of-service", + "privacyPolicyUrl": "https://www.nextbit256.com/docs/privacy-policy", + "paidModels": { + "training": false, + "retainsPrompts": false } + }, + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": false, + "moderationRequired": false, + "group": "NextBit", + "editors": [], + "owners": [], + "isMultipartSupported": false, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nextbit256.com/&size=256" } }, - "modelVariantSlug": "qwen/qwen3-235b-a22b", - "modelVariantPermaslug": "qwen/qwen3-235b-a22b-04-28", + "providerDisplayName": "NextBit", + "providerModelId": "qwen:2-5-7b", + "providerGroup": "NextBit", + "quantization": "bf16", + "variant": "standard", + "isFree": false, + "canAbort": false, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "structured_outputs" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://www.nextbit256.com/docs/terms-of-service", + "privacyPolicyUrl": "https://www.nextbit256.com/docs/privacy-policy", + "paidModels": { + "training": false, + "retainsPrompts": false + }, + "training": false, + "retainsPrompts": false + }, + "pricing": { + "prompt": "0.00000004", + "completion": "0.0000001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": false, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 22, + "newest": 188, + "throughputHighToLow": 5, + "latencyLowToHigh": 89, + "pricingLowToHigh": 60, + "pricingHighToLow": 227 + }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", + "providers": [ + { + "name": "NextBit", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nextbit256.com/&size=256", + "slug": "nextBit", + "quantization": "bf16", + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "qwen:2-5-7b", + "pricing": { + "prompt": "0.00000004", + "completion": "0.0000001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "structured_outputs" + ], + "inputCost": 0.04, + "outputCost": 0.1, + "throughput": 7.6005, + "latency": 1819 + }, + { + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 32768, + "maxCompletionTokens": 16384, + "providerModelId": "Qwen/Qwen2.5-7B-Instruct", + "pricing": { + "prompt": "0.00000005", + "completion": "0.0000001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" + ], + "inputCost": 0.05, + "outputCost": 0.1, + "throughput": 52.994, + "latency": 241 + }, + { + "name": "nCompass", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://console.ncompass.tech/&size=256", + "slug": "nCompass", + "quantization": "bf16", + "context": 32768, + "maxCompletionTokens": 32768, + "providerModelId": "Qwen/Qwen2.5-7B-Instruct", + "pricing": { + "prompt": "0.0000002", + "completion": "0.0000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty" + ], + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 145.8585, + "latency": 114 + }, + { + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": "fp8", + "context": 32768, + "maxCompletionTokens": 2048, + "providerModelId": "Qwen/Qwen2.5-7B-Instruct-Turbo", + "pricing": { + "prompt": "0.0000003", + "completion": "0.0000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" + ], + "inputCost": 0.3, + "outputCost": 0.3, + "throughput": 179.537, + "latency": 203 + } + ] + }, + { + "slug": "meta-llama/llama-4-maverick", + "hfSlug": "meta-llama/Llama-4-Maverick-17B-128E-Instruct", + "updatedAt": "2025-04-05T19:58:55.362967+00:00", + "createdAt": "2025-04-05T19:37:02.129674+00:00", + "hfUpdatedAt": null, + "name": "Meta: Llama 4 Maverick", + "shortName": "Llama 4 Maverick", + "author": "meta-llama", + "description": "Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward pass (400B total). It supports multilingual text and image input, and produces multilingual text and code output across 12 supported languages. Optimized for vision-language tasks, Maverick is instruction-tuned for assistant-like behavior, image reasoning, and general-purpose multimodal interaction.\n\nMaverick features early fusion for native multimodality and a 1 million token context window. It was trained on a curated mixture of public, licensed, and Meta-platform data, covering ~22 trillion tokens, with a knowledge cutoff in August 2024. Released on April 5, 2025 under the Llama 4 Community License, Maverick is suited for research and commercial applications requiring advanced multimodal understanding and high model throughput.", + "modelVersionGroupId": null, + "contextLength": 1048576, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Other", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "69a5d06e-1935-4aa5-903f-71058e64399f", + "name": "DeepInfra | meta-llama/llama-4-maverick-17b-128e-instruct", + "contextLength": 1048576, + "model": { + "slug": "meta-llama/llama-4-maverick", + "hfSlug": "meta-llama/Llama-4-Maverick-17B-128E-Instruct", + "updatedAt": "2025-04-05T19:58:55.362967+00:00", + "createdAt": "2025-04-05T19:37:02.129674+00:00", + "hfUpdatedAt": null, + "name": "Meta: Llama 4 Maverick", + "shortName": "Llama 4 Maverick", + "author": "meta-llama", + "description": "Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward pass (400B total). It supports multilingual text and image input, and produces multilingual text and code output across 12 supported languages. Optimized for vision-language tasks, Maverick is instruction-tuned for assistant-like behavior, image reasoning, and general-purpose multimodal interaction.\n\nMaverick features early fusion for native multimodality and a 1 million token context window. It was trained on a curated mixture of public, licensed, and Meta-platform data, covering ~22 trillion tokens, with a knowledge cutoff in August 2024. Released on April 5, 2025 under the Llama 4 Community License, Maverick is suited for research and commercial applications requiring advanced multimodal understanding and high model throughput.", + "modelVersionGroupId": null, + "contextLength": 1048576, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Other", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "meta-llama/llama-4-maverick", + "modelVariantPermaslug": "meta-llama/llama-4-maverick-17b-128e-instruct", "providerName": "DeepInfra", "providerInfo": { "name": "DeepInfra", @@ -7949,22 +8709,20 @@ export default { } }, "providerDisplayName": "DeepInfra", - "providerModelId": "Qwen/Qwen3-235B-A22B", + "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", "providerGroup": "DeepInfra", "quantization": "fp8", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 40960, - "maxPromptImages": null, - "maxTokensPerImage": null, + "maxCompletionTokens": 16384, + "maxPromptImages": 1, + "maxTokensPerImage": 3342, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -7988,9 +8746,9 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000014", + "prompt": "0.00000016", "completion": "0.0000006", - "image": "0", + "image": "0.0006684", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -8001,7 +8759,7 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, @@ -8014,25 +8772,27 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 22, - "newest": 30, - "throughputHighToLow": 238, - "latencyLowToHigh": 192, - "pricingLowToHigh": 13, - "pricingHighToLow": 175 + "topWeekly": 23, + "newest": 61, + "throughputHighToLow": 84, + "latencyLowToHigh": 60, + "pricingLowToHigh": 26, + "pricingHighToLow": 169 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": 40960, - "providerModelId": "Qwen/Qwen3-235B-A22B", + "context": 1048576, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", "pricing": { - "prompt": "0.00000014", + "prompt": "0.00000016", "completion": "0.0000006", - "image": "0", + "image": "0.0006684", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -8042,8 +8802,6 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -8053,19 +8811,94 @@ export default { "seed", "min_p" ], - "inputCost": 0.14, - "outputCost": 0.6 + "inputCost": 0.16, + "outputCost": 0.6, + "throughput": 74.5405, + "latency": 424 }, { "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", "slug": "klusterAi", "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": 40960, - "providerModelId": "Qwen/Qwen3-235B-A22B-FP8", + "context": 1048576, + "maxCompletionTokens": 1048576, + "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", "pricing": { - "prompt": "0.00000014", - "completion": "0.000002", + "prompt": "0.00000016", + "completion": "0.0000008", + "image": "0.0008355", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", + "seed" + ], + "inputCost": 0.16, + "outputCost": 0.8, + "throughput": 127.202, + "latency": 513 + }, + { + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": "fp8", + "context": 1048576, + "maxCompletionTokens": 1048576, + "providerModelId": "meta-llama/llama-4-maverick-17b-128e-instruct-fp8", + "pricing": { + "prompt": "0.00000017", + "completion": "0.00000085", + "image": "0.0006684", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" + ], + "inputCost": 0.17, + "outputCost": 0.85, + "throughput": 58.138, + "latency": 708 + }, + { + "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", + "slug": "lambda", + "quantization": "fp8", + "context": 1048576, + "maxCompletionTokens": 1048576, + "providerModelId": "llama-4-maverick-17b-128e-instruct-fp8", + "pricing": { + "prompt": "0.00000018", + "completion": "0.0000006", "image": "0", "request": "0", "webSearch": "0", @@ -8075,23 +8908,33 @@ export default { "supportedParameters": [ "max_tokens", "temperature", - "reasoning", - "include_reasoning" + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" ], - "inputCost": 0.14, - "outputCost": 2 + "inputCost": 0.18, + "outputCost": 0.6, + "throughput": 149.191, + "latency": 681 }, { "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", "slug": "parasail", "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": 40960, - "providerModelId": "parasail-qwen3-235b-a22b", + "context": 1048576, + "maxCompletionTokens": 1048576, + "providerModelId": "parasail-llama-4-maverick-instruct-fp8", "pricing": { - "prompt": "0.00000018", + "prompt": "0.00000019", "completion": "0.00000085", - "image": "0", + "image": "0.00070182", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -8101,26 +8944,27 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "presence_penalty", "frequency_penalty", "repetition_penalty", "top_k" ], - "inputCost": 0.18, - "outputCost": 0.85 + "inputCost": 0.19, + "outputCost": 0.85, + "throughput": 151.3725, + "latency": 492 }, { - "name": "Together", - "slug": "together", + "name": "CentML", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://centml.ai/&size=256", + "slug": "centMl", "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen3-235B-A22B-fp8-tput", + "context": 1048576, + "maxCompletionTokens": 1048576, + "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", "pricing": { "prompt": "0.0000002", - "completion": "0.0000006", + "completion": "0.0000002", "image": "0", "request": "0", "webSearch": "0", @@ -8131,27 +8975,27 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", + "seed", "top_k", - "repetition_penalty", - "logit_bias", "min_p", - "response_format" + "repetition_penalty" ], "inputCost": 0.2, - "outputCost": 0.6 + "outputCost": 0.2, + "throughput": 79.6685, + "latency": 407 }, { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen3-235B-A22B", + "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", + "slug": "groq", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 8192, + "providerModelId": "meta-llama/llama-4-maverick-17b-128e-instruct", "pricing": { "prompt": "0.0000002", "completion": "0.0000006", @@ -8162,33 +9006,36 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "logit_bias", + "response_format", + "top_logprobs", "logprobs", - "top_logprobs" + "logit_bias", + "seed" ], "inputCost": 0.2, - "outputCost": 0.6 + "outputCost": 0.6, + "throughput": 265.193, + "latency": 474 }, { - "name": "NovitaAI", - "slug": "novitaAi", + "name": "nCompass", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://console.ncompass.tech/&size=256", + "slug": "nCompass", "quantization": "fp8", - "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "qwen/qwen3-235b-a22b-fp8", + "context": 400000, + "maxCompletionTokens": 400000, + "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", "pricing": { "prompt": "0.0000002", - "completion": "0.0000008", + "completion": "0.0000007", "image": "0", "request": "0", "webSearch": "0", @@ -8196,32 +9043,30 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", "seed", "top_k", "min_p", - "repetition_penalty", - "logit_bias" + "repetition_penalty" ], "inputCost": 0.2, - "outputCost": 0.8 + "outputCost": 0.7, + "throughput": 133.7025, + "latency": 439 }, { "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", "slug": "fireworks", "quantization": null, - "context": 128000, + "context": 1048576, "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/qwen3-235b-a22b", + "providerModelId": "accounts/fireworks/models/llama4-maverick-instruct-basic", "pricing": { "prompt": "0.00000022", "completion": "0.00000088", @@ -8237,8 +9082,6 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -8251,18 +9094,21 @@ export default { "top_logprobs" ], "inputCost": 0.22, - "outputCost": 0.88 + "outputCost": 0.88, + "throughput": 83.029, + "latency": 634 }, { "name": "GMICloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://gmicloud.ai/&size=256", "slug": "gmiCloud", "quantization": "fp8", - "context": 32768, + "context": 1048576, "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen3-235B-A22B-FP8", + "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct", "pricing": { "prompt": "0.00000025", - "completion": "0.00000109", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -8273,27 +9119,93 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "seed" ], "inputCost": 0.25, - "outputCost": 1.09 + "outputCost": 0.8, + "throughput": 93.446, + "latency": 716 + }, + { + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": "fp8", + "context": 1048576, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", + "pricing": { + "prompt": "0.00000027", + "completion": "0.00000085", + "image": "0.00090234", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" + ], + "inputCost": 0.27, + "outputCost": 0.85, + "throughput": 98.9445, + "latency": 446 + }, + { + "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", + "slug": "sambaNova", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 4096, + "providerModelId": "Llama-4-Maverick-17B-128E-Instruct", + "pricing": { + "prompt": "0.00000063", + "completion": "0.0000018", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 0.63, + "outputCost": 1.8, + "throughput": 647.386, + "latency": 804 } ] }, { - "slug": "qwen/qwen-2.5-7b-instruct", - "hfSlug": "Qwen/Qwen2.5-7B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-10-16T00:00:00+00:00", + "slug": "qwen/qwen3-235b-a22b", + "hfSlug": "Qwen/Qwen3-235B-A22B", + "updatedAt": "2025-05-11T22:44:42.939095+00:00", + "createdAt": "2025-04-28T21:29:17.25671+00:00", "hfUpdatedAt": null, - "name": "Qwen2.5 7B Instruct", - "shortName": "Qwen2.5 7B Instruct", + "name": "Qwen: Qwen3 235B A22B", + "shortName": "Qwen3 235B A22B", "author": "qwen", - "description": "Qwen2.5 7B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "description": "Qwen3-235B-A22B is a 235B parameter mixture-of-experts (MoE) model developed by Qwen, activating 22B parameters per forward pass. It supports seamless switching between a \"thinking\" mode for complex reasoning, math, and code tasks, and a \"non-thinking\" mode for general conversational efficiency. The model demonstrates strong reasoning ability, multilingual support (100+ languages and dialects), advanced instruction-following, and agent tool-calling capabilities. It natively handles a 32K token context window and extends up to 131K tokens using YaRN-based scaling.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 40960, "inputModalities": [ "text" ], @@ -8301,34 +9213,41 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, "defaultStops": [ "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" + "<|im_end|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-2.5-7b-instruct", - "reasoningConfig": null, - "features": {}, + "permaslug": "qwen/qwen3-235b-a22b-04-28", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "928bf8a4-dddf-4c09-8d42-488e6907b11c", - "name": "DeepInfra | qwen/qwen-2.5-7b-instruct", - "contextLength": 32768, + "id": "5f108245-8fec-4702-859d-56c5b54a71d7", + "name": "DeepInfra | qwen/qwen3-235b-a22b-04-28", + "contextLength": 40960, "model": { - "slug": "qwen/qwen-2.5-7b-instruct", - "hfSlug": "Qwen/Qwen2.5-7B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-10-16T00:00:00+00:00", + "slug": "qwen/qwen3-235b-a22b", + "hfSlug": "Qwen/Qwen3-235B-A22B", + "updatedAt": "2025-05-11T22:44:42.939095+00:00", + "createdAt": "2025-04-28T21:29:17.25671+00:00", "hfUpdatedAt": null, - "name": "Qwen2.5 7B Instruct", - "shortName": "Qwen2.5 7B Instruct", + "name": "Qwen: Qwen3 235B A22B", + "shortName": "Qwen3 235B A22B", "author": "qwen", - "description": "Qwen2.5 7B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "description": "Qwen3-235B-A22B is a 235B parameter mixture-of-experts (MoE) model developed by Qwen, activating 22B parameters per forward pass. It supports seamless switching between a \"thinking\" mode for complex reasoning, math, and code tasks, and a \"non-thinking\" mode for general conversational efficiency. The model demonstrates strong reasoning ability, multilingual support (100+ languages and dialects), advanced instruction-following, and agent tool-calling capabilities. It natively handles a 32K token context window and extends up to 131K tokens using YaRN-based scaling.", "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ @@ -8338,23 +9257,30 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, "defaultStops": [ "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" + "<|im_end|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-2.5-7b-instruct", - "reasoningConfig": null, - "features": {} + "permaslug": "qwen/qwen3-235b-a22b-04-28", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "qwen/qwen-2.5-7b-instruct", - "modelVariantPermaslug": "qwen/qwen-2.5-7b-instruct", + "modelVariantSlug": "qwen/qwen3-235b-a22b", + "modelVariantPermaslug": "qwen/qwen3-235b-a22b-04-28", "providerName": "DeepInfra", "providerInfo": { "name": "DeepInfra", @@ -8387,20 +9313,22 @@ export default { } }, "providerDisplayName": "DeepInfra", - "providerModelId": "Qwen/Qwen2.5-7B-Instruct", + "providerModelId": "Qwen/Qwen3-235B-A22B", "providerGroup": "DeepInfra", - "quantization": "bf16", + "quantization": "fp8", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 16384, + "maxCompletionTokens": 40960, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -8424,8 +9352,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000005", - "completion": "0.0000001", + "prompt": "0.00000014", + "completion": "0.0000006", "image": "0", "request": "0", "webSearch": "0", @@ -8437,36 +9365,39 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 23, - "newest": 188, - "throughputHighToLow": 11, - "latencyLowToHigh": 18, - "pricingLowToHigh": 60, - "pricingHighToLow": 222 + "topWeekly": 24, + "newest": 30, + "throughputHighToLow": 244, + "latencyLowToHigh": 180, + "pricingLowToHigh": 13, + "pricingHighToLow": 178 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", - "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": 16384, - "providerModelId": "Qwen/Qwen2.5-7B-Instruct", + "quantization": "fp8", + "context": 40960, + "maxCompletionTokens": 40960, + "providerModelId": "Qwen/Qwen3-235B-A22B", "pricing": { - "prompt": "0.00000005", - "completion": "0.0000001", + "prompt": "0.00000014", + "completion": "0.0000006", "image": "0", "request": "0", "webSearch": "0", @@ -8477,6 +9408,8 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -8486,19 +9419,22 @@ export default { "seed", "min_p" ], - "inputCost": 0.05, - "outputCost": 0.1 + "inputCost": 0.14, + "outputCost": 0.6, + "throughput": 23.8015, + "latency": 1157 }, { - "name": "nCompass", - "slug": "nCompass", - "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": 32768, - "providerModelId": "Qwen/Qwen2.5-7B-Instruct", + "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", + "slug": "klusterAi", + "quantization": "fp8", + "context": 40960, + "maxCompletionTokens": 40960, + "providerModelId": "Qwen/Qwen3-235B-A22B-FP8", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", + "prompt": "0.00000014", + "completion": "0.000002", "image": "0", "request": "0", "webSearch": "0", @@ -8506,30 +9442,73 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "seed", "top_k", + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", "min_p", - "repetition_penalty" + "seed" ], - "inputCost": 0.2, - "outputCost": 0.2 + "inputCost": 0.14, + "outputCost": 2, + "throughput": 40.982, + "latency": 797.5 + }, + { + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", + "quantization": "fp8", + "context": 40960, + "maxCompletionTokens": 40960, + "providerModelId": "parasail-qwen3-235b-a22b", + "pricing": { + "prompt": "0.00000018", + "completion": "0.00000085", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "presence_penalty", + "frequency_penalty", + "repetition_penalty", + "top_k" + ], + "inputCost": 0.18, + "outputCost": 0.85, + "throughput": 55.8185, + "latency": 968 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "fp8", - "context": 32768, - "maxCompletionTokens": 2048, - "providerModelId": "Qwen/Qwen2.5-7B-Instruct-Turbo", + "context": 40960, + "maxCompletionTokens": null, + "providerModelId": "Qwen/Qwen3-235B-A22B-fp8-tput", "pricing": { - "prompt": "0.0000003", - "completion": "0.0000003", + "prompt": "0.0000002", + "completion": "0.0000006", "image": "0", "request": "0", "webSearch": "0", @@ -8540,6 +9519,8 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -8549,277 +9530,21 @@ export default { "min_p", "response_format" ], - "inputCost": 0.3, - "outputCost": 0.3 - } - ] - }, - { - "slug": "meta-llama/llama-4-maverick", - "hfSlug": "meta-llama/Llama-4-Maverick-17B-128E-Instruct", - "updatedAt": "2025-04-05T19:58:55.362967+00:00", - "createdAt": "2025-04-05T19:37:02.129674+00:00", - "hfUpdatedAt": null, - "name": "Meta: Llama 4 Maverick", - "shortName": "Llama 4 Maverick", - "author": "meta-llama", - "description": "Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward pass (400B total). It supports multilingual text and image input, and produces multilingual text and code output across 12 supported languages. Optimized for vision-language tasks, Maverick is instruction-tuned for assistant-like behavior, image reasoning, and general-purpose multimodal interaction.\n\nMaverick features early fusion for native multimodality and a 1 million token context window. It was trained on a curated mixture of public, licensed, and Meta-platform data, covering ~22 trillion tokens, with a knowledge cutoff in August 2024. Released on April 5, 2025 under the Llama 4 Community License, Maverick is suited for research and commercial applications requiring advanced multimodal understanding and high model throughput.", - "modelVersionGroupId": null, - "contextLength": 1048576, - "inputModalities": [ - "text", - "image" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "69a5d06e-1935-4aa5-903f-71058e64399f", - "name": "DeepInfra | meta-llama/llama-4-maverick-17b-128e-instruct", - "contextLength": 1048576, - "model": { - "slug": "meta-llama/llama-4-maverick", - "hfSlug": "meta-llama/Llama-4-Maverick-17B-128E-Instruct", - "updatedAt": "2025-04-05T19:58:55.362967+00:00", - "createdAt": "2025-04-05T19:37:02.129674+00:00", - "hfUpdatedAt": null, - "name": "Meta: Llama 4 Maverick", - "shortName": "Llama 4 Maverick", - "author": "meta-llama", - "description": "Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward pass (400B total). It supports multilingual text and image input, and produces multilingual text and code output across 12 supported languages. Optimized for vision-language tasks, Maverick is instruction-tuned for assistant-like behavior, image reasoning, and general-purpose multimodal interaction.\n\nMaverick features early fusion for native multimodality and a 1 million token context window. It was trained on a curated mixture of public, licensed, and Meta-platform data, covering ~22 trillion tokens, with a knowledge cutoff in August 2024. Released on April 5, 2025 under the Llama 4 Community License, Maverick is suited for research and commercial applications requiring advanced multimodal understanding and high model throughput.", - "modelVersionGroupId": null, - "contextLength": 1048576, - "inputModalities": [ - "text", - "image" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "meta-llama/llama-4-maverick", - "modelVariantPermaslug": "meta-llama/llama-4-maverick-17b-128e-instruct", - "providerName": "DeepInfra", - "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", - "baseUrl": "url", - "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", - "paidModels": { - "training": false, - "retainsPrompts": false - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "DeepInfra", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/DeepInfra.webp" - } - }, - "providerDisplayName": "DeepInfra", - "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", - "providerGroup": "DeepInfra", - "quantization": "fp8", - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 16384, - "maxPromptImages": 1, - "maxTokensPerImage": 3342, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", - "paidModels": { - "training": false, - "retainsPrompts": false - }, - "training": false, - "retainsPrompts": false - }, - "pricing": { - "prompt": "0.00000016", - "completion": "0.0000006", - "image": "0.0006684", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 24, - "newest": 61, - "throughputHighToLow": 93, - "latencyLowToHigh": 23, - "pricingLowToHigh": 26, - "pricingHighToLow": 166 - }, - "providers": [ - { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "fp8", - "context": 1048576, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", - "pricing": { - "prompt": "0.00000016", - "completion": "0.0000006", - "image": "0.0006684", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "inputCost": 0.16, - "outputCost": 0.6 - }, - { - "name": "kluster.ai", - "slug": "klusterAi", - "quantization": "fp8", - "context": 1048576, - "maxCompletionTokens": 1048576, - "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", - "pricing": { - "prompt": "0.00000016", - "completion": "0.0000008", - "image": "0.0008355", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature" - ], - "inputCost": 0.16, - "outputCost": 0.8 - }, - { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": "fp8", - "context": 1048576, - "maxCompletionTokens": 1048576, - "providerModelId": "meta-llama/llama-4-maverick-17b-128e-instruct-fp8", - "pricing": { - "prompt": "0.00000017", - "completion": "0.00000085", - "image": "0.0006684", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" - ], - "inputCost": 0.17, - "outputCost": 0.85 + "inputCost": 0.2, + "outputCost": 0.6, + "throughput": 30.284, + "latency": 833 }, { - "name": "Lambda", - "slug": "lambda", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", "quantization": "fp8", - "context": 1048576, - "maxCompletionTokens": 1048576, - "providerModelId": "llama-4-maverick-17b-128e-instruct-fp8", + "context": 40960, + "maxCompletionTokens": null, + "providerModelId": "Qwen/Qwen3-235B-A22B", "pricing": { - "prompt": "0.00000018", + "prompt": "0.0000002", "completion": "0.0000006", "image": "0", "request": "0", @@ -8831,87 +9556,33 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", "seed", + "top_k", "logit_bias", "logprobs", - "top_logprobs", - "response_format" - ], - "inputCost": 0.18, - "outputCost": 0.6 - }, - { - "name": "Parasail", - "slug": "parasail", - "quantization": "fp8", - "context": 1048576, - "maxCompletionTokens": 1048576, - "providerModelId": "parasail-llama-4-maverick-instruct-fp8", - "pricing": { - "prompt": "0.00000019", - "completion": "0.00000085", - "image": "0.00070182", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "presence_penalty", - "frequency_penalty", - "repetition_penalty", - "top_k" - ], - "inputCost": 0.19, - "outputCost": 0.85 - }, - { - "name": "CentML", - "slug": "centMl", - "quantization": "fp8", - "context": 1048576, - "maxCompletionTokens": 1048576, - "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", - "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty" + "top_logprobs" ], "inputCost": 0.2, - "outputCost": 0.2 + "outputCost": 0.6, + "throughput": 32.01, + "latency": 699 }, { - "name": "Groq", - "slug": "groq", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 8192, - "providerModelId": "meta-llama/llama-4-maverick-17b-128e-instruct", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": "fp8", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "qwen/qwen3-235b-a22b-fp8", "pricing": { "prompt": "0.0000002", - "completion": "0.0000006", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -8924,56 +9595,30 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "top_logprobs", - "logprobs", - "logit_bias", - "seed" - ], - "inputCost": 0.2, - "outputCost": 0.6 - }, - { - "name": "nCompass", - "slug": "nCompass", - "quantization": "fp8", - "context": 400000, - "maxCompletionTokens": 400000, - "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", - "pricing": { - "prompt": "0.0000002", - "completion": "0.0000007", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", "seed", "top_k", "min_p", - "repetition_penalty" + "repetition_penalty", + "logit_bias" ], "inputCost": 0.2, - "outputCost": 0.7 + "outputCost": 0.8, + "throughput": 55.828, + "latency": 1290 }, { "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", "slug": "fireworks", "quantization": null, - "context": 1048576, + "context": 128000, "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/llama4-maverick-instruct-basic", + "providerModelId": "accounts/fireworks/models/qwen3-235b-a22b", "pricing": { "prompt": "0.00000022", "completion": "0.00000088", @@ -8989,6 +9634,8 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -9001,18 +9648,21 @@ export default { "top_logprobs" ], "inputCost": 0.22, - "outputCost": 0.88 + "outputCost": 0.88, + "throughput": 53.7815, + "latency": 857.5 }, { "name": "GMICloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://gmicloud.ai/&size=256", "slug": "gmiCloud", "quantization": "fp8", - "context": 1048576, + "context": 32768, "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct", + "providerModelId": "Qwen/Qwen3-235B-A22B-FP8", "pricing": { "prompt": "0.00000025", - "completion": "0.0000008", + "completion": "0.00000109", "image": "0", "request": "0", "webSearch": "0", @@ -9023,70 +9673,14 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "seed" ], "inputCost": 0.25, - "outputCost": 0.8 - }, - { - "name": "Together", - "slug": "together", - "quantization": "fp8", - "context": 1048576, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", - "pricing": { - "prompt": "0.00000027", - "completion": "0.00000085", - "image": "0.00090234", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" - ], - "inputCost": 0.27, - "outputCost": 0.85 - }, - { - "name": "SambaNova", - "slug": "sambaNova", - "quantization": null, - "context": 65536, - "maxCompletionTokens": 4096, - "providerModelId": "Llama-4-Maverick-17B-128E-Instruct", - "pricing": { - "prompt": "0.00000063", - "completion": "0.0000018", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "top_k", - "stop" - ], - "inputCost": 0.63, - "outputCost": 1.8 + "outputCost": 1.09, + "throughput": 48.529, + "latency": 934 } ] }, @@ -9258,15 +9852,17 @@ export default { }, "sorting": { "topWeekly": 25, - "newest": 231, - "throughputHighToLow": 223, - "latencyLowToHigh": 32, + "newest": 232, + "throughputHighToLow": 219, + "latencyLowToHigh": 27, "pricingLowToHigh": 120, "pricingHighToLow": 198 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", "context": 131072, @@ -9297,10 +9893,13 @@ export default { "min_p" ], "inputCost": 0.1, - "outputCost": 0.28 + "outputCost": 0.28, + "throughput": 37.259, + "latency": 306.5 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "bf16", "context": 32768, @@ -9329,10 +9928,13 @@ export default { "logit_bias" ], "inputCost": 0.12, - "outputCost": 0.39 + "outputCost": 0.39, + "throughput": 30.581, + "latency": 1545 }, { "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", "slug": "lambda", "quantization": "fp8", "context": 131072, @@ -9361,10 +9963,13 @@ export default { "response_format" ], "inputCost": 0.12, - "outputCost": 0.3 + "outputCost": 0.3, + "throughput": 53.078, + "latency": 510 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 131072, @@ -9393,10 +9998,13 @@ export default { "top_logprobs" ], "inputCost": 0.13, - "outputCost": 0.4 + "outputCost": 0.4, + "throughput": 35.3465, + "latency": 775 }, { "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", "slug": "inferenceNet", "quantization": "fp16", "context": 16384, @@ -9429,6 +10037,7 @@ export default { }, { "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", "slug": "hyperbolic", "quantization": "fp8", "context": 131072, @@ -9459,10 +10068,13 @@ export default { "repetition_penalty" ], "inputCost": 0.4, - "outputCost": 0.4 + "outputCost": 0.4, + "throughput": 113.116, + "latency": 938.5 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "fp8", "context": 131072, @@ -9493,10 +10105,13 @@ export default { "response_format" ], "inputCost": 0.88, - "outputCost": 0.88 + "outputCost": 0.88, + "throughput": 95.502, + "latency": 626 }, { "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", "slug": "fireworks", "quantization": "unknown", "context": 131072, @@ -9529,95 +10144,102 @@ export default { "top_logprobs" ], "inputCost": 0.9, - "outputCost": 0.9 + "outputCost": 0.9, + "throughput": 125.7935, + "latency": 379 } ] }, { - "slug": "meta-llama/llama-4-maverick", - "hfSlug": "meta-llama/Llama-4-Maverick-17B-128E-Instruct", - "updatedAt": "2025-04-05T19:58:55.362967+00:00", - "createdAt": "2025-04-05T19:37:02.129674+00:00", + "slug": "meta-llama/llama-3.1-8b-instruct", + "hfSlug": "meta-llama/Meta-Llama-3.1-8B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-07-23T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 4 Maverick (free)", - "shortName": "Llama 4 Maverick (free)", + "name": "Meta: Llama 3.1 8B Instruct", + "shortName": "Llama 3.1 8B Instruct", "author": "meta-llama", - "description": "Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward pass (400B total). It supports multilingual text and image input, and produces multilingual text and code output across 12 supported languages. Optimized for vision-language tasks, Maverick is instruction-tuned for assistant-like behavior, image reasoning, and general-purpose multimodal interaction.\n\nMaverick features early fusion for native multimodality and a 1 million token context window. It was trained on a curated mixture of public, licensed, and Meta-platform data, covering ~22 trillion tokens, with a knowledge cutoff in August 2024. Released on April 5, 2025 under the Llama 4 Community License, Maverick is suited for research and commercial applications requiring advanced multimodal understanding and high model throughput.", - "modelVersionGroupId": null, - "contextLength": 256000, + "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 8B instruct-tuned version is fast and efficient.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "modelVersionGroupId": "803c32ed-9861-4abf-b5da-7d9c9e6dcf04", + "contextLength": 16384, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct", + "permaslug": "meta-llama/llama-3.1-8b-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "2ff5f508-d666-4051-9625-0c795e746c32", - "name": "Chutes | meta-llama/llama-4-maverick-17b-128e-instruct:free", - "contextLength": 256000, + "id": "612663e2-9ede-4c5d-b993-c55cd12e3055", + "name": "InferenceNet | meta-llama/llama-3.1-8b-instruct", + "contextLength": 16384, "model": { - "slug": "meta-llama/llama-4-maverick", - "hfSlug": "meta-llama/Llama-4-Maverick-17B-128E-Instruct", - "updatedAt": "2025-04-05T19:58:55.362967+00:00", - "createdAt": "2025-04-05T19:37:02.129674+00:00", + "slug": "meta-llama/llama-3.1-8b-instruct", + "hfSlug": "meta-llama/Meta-Llama-3.1-8B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-07-23T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 4 Maverick", - "shortName": "Llama 4 Maverick", + "name": "Meta: Llama 3.1 8B Instruct", + "shortName": "Llama 3.1 8B Instruct", "author": "meta-llama", - "description": "Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward pass (400B total). It supports multilingual text and image input, and produces multilingual text and code output across 12 supported languages. Optimized for vision-language tasks, Maverick is instruction-tuned for assistant-like behavior, image reasoning, and general-purpose multimodal interaction.\n\nMaverick features early fusion for native multimodality and a 1 million token context window. It was trained on a curated mixture of public, licensed, and Meta-platform data, covering ~22 trillion tokens, with a knowledge cutoff in August 2024. Released on April 5, 2025 under the Llama 4 Community License, Maverick is suited for research and commercial applications requiring advanced multimodal understanding and high model throughput.", - "modelVersionGroupId": null, - "contextLength": 1048576, + "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 8B instruct-tuned version is fast and efficient.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "modelVersionGroupId": "803c32ed-9861-4abf-b5da-7d9c9e6dcf04", + "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct", + "permaslug": "meta-llama/llama-3.1-8b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "meta-llama/llama-4-maverick:free", - "modelVariantPermaslug": "meta-llama/llama-4-maverick-17b-128e-instruct:free", - "providerName": "Chutes", + "modelVariantSlug": "meta-llama/llama-3.1-8b-instruct", + "modelVariantPermaslug": "meta-llama/llama-3.1-8b-instruct", + "providerName": "InferenceNet", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "InferenceNet", + "displayName": "inference.net", + "slug": "inference-net", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "privacyPolicyUrl": "https://inference.net/privacy-policy", + "termsOfServiceUrl": "https://inference.net/terms-of-service", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false } }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": true, + "isAbortable": false, "moderationRequired": false, - "group": "Chutes", + "group": "InferenceNet", "editors": [], "owners": [], "isMultipartSupported": true, @@ -9625,51 +10247,47 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", + "invertRequired": true } }, - "providerDisplayName": "Chutes", - "providerModelId": "chutesai/Llama-4-Maverick-17B-128E-Instruct-FP8", - "providerGroup": "Chutes", - "quantization": "fp8", - "variant": "free", - "isFree": true, - "canAbort": true, + "providerDisplayName": "inference.net", + "providerModelId": "meta-llama/llama-3.1-8b-instruct/fp-16", + "providerGroup": "InferenceNet", + "quantization": "fp16", + "variant": "standard", + "isFree": false, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": 8, - "maxTokensPerImage": 3342, + "maxCompletionTokens": 16384, + "maxPromptImages": null, + "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "structured_outputs", - "response_format", "stop", "frequency_penalty", "presence_penalty", "seed", "top_k", "min_p", - "repetition_penalty", - "logprobs", "logit_bias", "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "privacyPolicyUrl": "https://inference.net/privacy-policy", + "termsOfServiceUrl": "https://inference.net/terms-of-service", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false }, - "training": true, - "retainsPrompts": true + "training": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.00000002", + "completion": "0.00000003", "image": "0", "request": "0", "webSearch": "0", @@ -9688,34 +10306,32 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": { - "responseFormat": true, - "structuredOutputs": true - }, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 24, - "newest": 61, - "throughputHighToLow": 93, - "latencyLowToHigh": 23, - "pricingLowToHigh": 26, - "pricingHighToLow": 166 + "topWeekly": 26, + "newest": 229, + "throughputHighToLow": 147, + "latencyLowToHigh": 172, + "pricingLowToHigh": 67, + "pricingHighToLow": 242 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "fp8", - "context": 1048576, + "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", + "slug": "inferenceNet", + "quantization": "fp16", + "context": 16384, "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", + "providerModelId": "meta-llama/llama-3.1-8b-instruct/fp-16", "pricing": { - "prompt": "0.00000016", - "completion": "0.0000006", - "image": "0.0006684", + "prompt": "0.00000002", + "completion": "0.00000003", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -9728,49 +10344,66 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", "seed", - "min_p" + "top_k", + "min_p", + "logit_bias", + "top_logprobs" ], - "inputCost": 0.16, - "outputCost": 0.6 + "inputCost": 0.02, + "outputCost": 0.03, + "throughput": 51.254, + "latency": 1218 }, { - "name": "kluster.ai", - "slug": "klusterAi", + "name": "DeepInfra (Turbo)", + "icon": "", + "slug": "deepInfra (turbo)", "quantization": "fp8", - "context": 1048576, - "maxCompletionTokens": 1048576, - "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", + "context": 131072, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo", "pricing": { - "prompt": "0.00000016", - "completion": "0.0000008", - "image": "0.0008355", + "prompt": "0.00000002", + "completion": "0.00000003", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", - "temperature" + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" ], - "inputCost": 0.16, - "outputCost": 0.8 + "inputCost": 0.02, + "outputCost": 0.03, + "throughput": 101.6575, + "latency": 250 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "fp8", - "context": 1048576, - "maxCompletionTokens": 1048576, - "providerModelId": "meta-llama/llama-4-maverick-17b-128e-instruct-fp8", + "context": 16384, + "maxCompletionTokens": 8192, + "providerModelId": "meta-llama/llama-3.1-8b-instruct", "pricing": { - "prompt": "0.00000017", - "completion": "0.00000085", - "image": "0.0006684", + "prompt": "0.00000002", + "completion": "0.00000005", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -9789,19 +10422,57 @@ export default { "repetition_penalty", "logit_bias" ], - "inputCost": 0.17, - "outputCost": 0.85 + "inputCost": 0.02, + "outputCost": 0.05, + "throughput": 78.6405, + "latency": 720 + }, + { + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Meta-Llama-3.1-8B-Instruct", + "pricing": { + "prompt": "0.00000002", + "completion": "0.00000006", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.02, + "outputCost": 0.06, + "throughput": 54.269, + "latency": 482 }, { "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", "slug": "lambda", - "quantization": "fp8", - "context": 1048576, - "maxCompletionTokens": 1048576, - "providerModelId": "llama-4-maverick-17b-128e-instruct-fp8", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "llama3.1-8b-instruct", "pricing": { - "prompt": "0.00000018", - "completion": "0.0000006", + "prompt": "0.000000025", + "completion": "0.00000004", "image": "0", "request": "0", "webSearch": "0", @@ -9821,47 +10492,60 @@ export default { "top_logprobs", "response_format" ], - "inputCost": 0.18, - "outputCost": 0.6 + "inputCost": 0.02, + "outputCost": 0.04, + "throughput": 162.1145, + "latency": 563 }, { - "name": "Parasail", - "slug": "parasail", - "quantization": "fp8", - "context": 1048576, - "maxCompletionTokens": 1048576, - "providerModelId": "parasail-llama-4-maverick-instruct-fp8", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/Meta-Llama-3.1-8B-Instruct", "pricing": { - "prompt": "0.00000019", - "completion": "0.00000085", - "image": "0.00070182", + "prompt": "0.00000003", + "completion": "0.00000005", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "presence_penalty", + "structured_outputs", + "response_format", + "stop", "frequency_penalty", + "presence_penalty", "repetition_penalty", - "top_k" + "top_k", + "seed", + "min_p" ], - "inputCost": 0.19, - "outputCost": 0.85 + "inputCost": 0.03, + "outputCost": 0.05, + "throughput": 51.918, + "latency": 338 }, { - "name": "CentML", - "slug": "centMl", + "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", + "slug": "cloudflare", "quantization": "fp8", - "context": 1048576, - "maxCompletionTokens": 1048576, - "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", + "context": 32000, + "maxCompletionTokens": null, + "providerModelId": "@cf/meta/llama-3.1-8b-instruct-fp8", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", + "prompt": "0.000000045", + "completion": "0.000000384", "image": "0", "request": "0", "webSearch": "0", @@ -9869,30 +10553,33 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", "top_k", - "min_p", - "repetition_penalty" + "seed", + "repetition_penalty", + "frequency_penalty", + "presence_penalty" ], - "inputCost": 0.2, - "outputCost": 0.2 + "inputCost": 0.04, + "outputCost": 0.38, + "throughput": 23.5165, + "latency": 784 }, { "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", "slug": "groq", "quantization": null, "context": 131072, - "maxCompletionTokens": 8192, - "providerModelId": "meta-llama/llama-4-maverick-17b-128e-instruct", + "maxCompletionTokens": 131072, + "providerModelId": "llama-3.1-8b-instant", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000006", + "prompt": "0.00000005", + "completion": "0.00000008", "image": "0", "request": "0", "webSearch": "0", @@ -9914,19 +10601,54 @@ export default { "logit_bias", "seed" ], - "inputCost": 0.2, - "outputCost": 0.6 + "inputCost": 0.05, + "outputCost": 0.08, + "throughput": 1512.5825, + "latency": 1437 }, { - "name": "nCompass", - "slug": "nCompass", + "name": "NextBit", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nextbit256.com/&size=256", + "slug": "nextBit", "quantization": "fp8", - "context": 400000, - "maxCompletionTokens": 400000, - "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "llama3.1:8b", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000007", + "prompt": "0.00000006", + "completion": "0.0000001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "structured_outputs" + ], + "inputCost": 0.06, + "outputCost": 0.1, + "throughput": 114.779, + "latency": 1597 + }, + { + "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", + "slug": "hyperbolic", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Meta-Llama-3.1-8B-Instruct", + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000001", "image": "0", "request": "0", "webSearch": "0", @@ -9940,24 +10662,64 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "logprobs", + "top_logprobs", "seed", + "logit_bias", "top_k", "min_p", "repetition_penalty" ], - "inputCost": 0.2, - "outputCost": 0.7 + "inputCost": 0.1, + "outputCost": 0.1, + "throughput": 268.153, + "latency": 981 }, { - "name": "Fireworks", - "slug": "fireworks", + "name": "Cerebras", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cerebras.ai/&size=256", + "slug": "cerebras", + "quantization": "fp16", + "context": 32000, + "maxCompletionTokens": 32000, + "providerModelId": "llama3.1-8b", + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "response_format", + "stop", + "seed", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.1, + "outputCost": 0.1, + "throughput": 3911.239, + "latency": 188 + }, + { + "name": "Friendli", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://friendli.ai/&size=256", + "slug": "friendli", "quantization": null, - "context": 1048576, - "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/llama4-maverick-instruct-basic", + "context": 131072, + "maxCompletionTokens": 8000, + "providerModelId": "meta-llama-3.1-8b-instruct", "pricing": { - "prompt": "0.00000022", - "completion": "0.00000088", + "prompt": "0.0000001", + "completion": "0.0000001", "image": "0", "request": "0", "webSearch": "0", @@ -9973,27 +10735,58 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "seed", "top_k", + "min_p", "repetition_penalty", "response_format", - "structured_outputs", - "logit_bias", - "logprobs", - "top_logprobs" + "structured_outputs" ], - "inputCost": 0.22, - "outputCost": 0.88 + "inputCost": 0.1, + "outputCost": 0.1, + "throughput": 282.859, + "latency": 304 }, { - "name": "GMICloud", - "slug": "gmiCloud", + "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", + "slug": "sambaNova", + "quantization": "bf16", + "context": 16384, + "maxCompletionTokens": 4096, + "providerModelId": "Meta-Llama-3.1-8B-Instruct", + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 0.1, + "outputCost": 0.2, + "throughput": 879.206, + "latency": 329 + }, + { + "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", + "slug": "klusterAi", "quantization": "fp8", - "context": 1048576, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct", + "context": 131000, + "maxCompletionTokens": 131000, + "providerModelId": "klusterai/Meta-Llama-3.1-8B-Instruct-Turbo", "pricing": { - "prompt": "0.00000025", - "completion": "0.0000008", + "prompt": "0.00000018", + "completion": "0.00000018", "image": "0", "request": "0", "webSearch": "0", @@ -10004,22 +10797,34 @@ export default { "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", "seed" ], - "inputCost": 0.25, - "outputCost": 0.8 + "inputCost": 0.18, + "outputCost": 0.18, + "throughput": 62.505, + "latency": 551 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "fp8", - "context": 1048576, + "context": 131072, "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", + "providerModelId": "meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo", "pricing": { - "prompt": "0.00000027", - "completion": "0.00000085", - "image": "0.00090234", + "prompt": "0.00000018", + "completion": "0.00000018", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -10040,19 +10845,22 @@ export default { "min_p", "response_format" ], - "inputCost": 0.27, - "outputCost": 0.85 + "inputCost": 0.18, + "outputCost": 0.18, + "throughput": 252.455, + "latency": 313 }, { - "name": "SambaNova", - "slug": "sambaNova", - "quantization": null, - "context": 65536, - "maxCompletionTokens": 4096, - "providerModelId": "Llama-4-Maverick-17B-128E-Instruct", + "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", + "slug": "fireworks", + "quantization": "unknown", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "accounts/fireworks/models/llama-v3p1-8b-instruct", "pricing": { - "prompt": "0.00000063", - "completion": "0.0000018", + "prompt": "0.0000002", + "completion": "0.0000002", "image": "0", "request": "0", "webSearch": "0", @@ -10063,79 +10871,128 @@ export default { "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "top_k", - "stop" + "repetition_penalty", + "response_format", + "structured_outputs", + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 0.63, - "outputCost": 1.8 + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 255.586, + "latency": 361 + }, + { + "name": "Avian.io", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://avian.io/&size=256", + "slug": "avianIo", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "Meta-Llama-3.1-8B-Instruct", + "pricing": { + "prompt": "0.0000002", + "completion": "0.0000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "response_format", + "seed" + ], + "inputCost": 0.2, + "outputCost": 0.2 } ] }, { - "slug": "deepseek/deepseek-chat", - "hfSlug": "deepseek-ai/DeepSeek-V3", + "slug": "google/gemma-3-27b-it", + "hfSlug": "", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-26T19:28:40.559917+00:00", + "createdAt": "2025-03-12T05:12:39.645813+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: DeepSeek V3", - "shortName": "DeepSeek V3", - "author": "deepseek", - "description": "DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations reveal that the model outperforms other open-source models and rivals leading closed-source models.\n\nFor model details, please visit [the DeepSeek-V3 repo](https://github.com/deepseek-ai/DeepSeek-V3) for more information, or see the [launch announcement](https://api-docs.deepseek.com/news/news1226).", - "modelVersionGroupId": null, - "contextLength": 163840, + "name": "Google: Gemma 3 27B", + "shortName": "Gemma 3 27B", + "author": "google", + "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 27B is Google's latest open source model, successor to [Gemma 2](google/gemma-2-27b-it)", + "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", + "contextLength": 131072, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "DeepSeek", - "instructType": null, + "group": "Gemini", + "instructType": "gemma", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "", + "", + "" + ], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "deepseek/deepseek-chat-v3", + "warningMessage": "", + "permaslug": "google/gemma-3-27b-it", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "5294d55f-9012-496b-8f22-8cc919432dcd", - "name": "DeepInfra | deepseek/deepseek-chat-v3", - "contextLength": 163840, + "id": "8f22002c-c045-446f-a1b9-9896133536b8", + "name": "DeepInfra | google/gemma-3-27b-it", + "contextLength": 131072, "model": { - "slug": "deepseek/deepseek-chat", - "hfSlug": "deepseek-ai/DeepSeek-V3", + "slug": "google/gemma-3-27b-it", + "hfSlug": "", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-26T19:28:40.559917+00:00", + "createdAt": "2025-03-12T05:12:39.645813+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: DeepSeek V3", - "shortName": "DeepSeek V3", - "author": "deepseek", - "description": "DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations reveal that the model outperforms other open-source models and rivals leading closed-source models.\n\nFor model details, please visit [the DeepSeek-V3 repo](https://github.com/deepseek-ai/DeepSeek-V3) for more information, or see the [launch announcement](https://api-docs.deepseek.com/news/news1226).", - "modelVersionGroupId": null, + "name": "Google: Gemma 3 27B", + "shortName": "Gemma 3 27B", + "author": "google", + "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 27B is Google's latest open source model, successor to [Gemma 2](google/gemma-2-27b-it)", + "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", "contextLength": 131072, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "DeepSeek", - "instructType": null, + "group": "Gemini", + "instructType": "gemma", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "", + "", + "" + ], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "deepseek/deepseek-chat-v3", + "warningMessage": "", + "permaslug": "google/gemma-3-27b-it", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "deepseek/deepseek-chat", - "modelVariantPermaslug": "deepseek/deepseek-chat-v3", + "modelVariantSlug": "google/gemma-3-27b-it", + "modelVariantPermaslug": "google/gemma-3-27b-it", "providerName": "DeepInfra", "providerInfo": { "name": "DeepInfra", @@ -10168,14 +11025,14 @@ export default { } }, "providerDisplayName": "DeepInfra", - "providerModelId": "deepseek-ai/DeepSeek-V3", + "providerModelId": "google/gemma-3-27b-it", "providerGroup": "DeepInfra", - "quantization": "fp8", + "quantization": "bf16", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 163840, + "maxCompletionTokens": 16384, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -10205,9 +11062,9 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000038", - "completion": "0.00000089", - "image": "0", + "prompt": "0.0000001", + "completion": "0.0000002", + "image": "0.0000256", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -10232,24 +11089,26 @@ export default { }, "sorting": { "topWeekly": 27, - "newest": 148, - "throughputHighToLow": 126, - "latencyLowToHigh": 106, - "pricingLowToHigh": 54, - "pricingHighToLow": 146 + "newest": 93, + "throughputHighToLow": 35, + "latencyLowToHigh": 143, + "pricingLowToHigh": 41, + "pricingHighToLow": 199 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", - "quantization": "fp8", - "context": 163840, - "maxCompletionTokens": 163840, - "providerModelId": "deepseek-ai/DeepSeek-V3", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": 16384, + "providerModelId": "google/gemma-3-27b-it", "pricing": { - "prompt": "0.00000038", - "completion": "0.00000089", - "image": "0", + "prompt": "0.0000001", + "completion": "0.0000002", + "image": "0.0000256", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -10268,19 +11127,22 @@ export default { "seed", "min_p" ], - "inputCost": 0.38, - "outputCost": 0.89 + "inputCost": 0.1, + "outputCost": 0.2, + "throughput": 32.525, + "latency": 868 }, { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": null, - "context": 64000, - "maxCompletionTokens": 16000, - "providerModelId": "deepseek/deepseek-v3-turbo", + "name": "InoCloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inocloud.com/&size=256", + "slug": "inoCloud", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": 16384, + "providerModelId": "google/gemma-3-27b-it", "pricing": { - "prompt": "0.0000004", - "completion": "0.0000013", + "prompt": "0.0000001", + "completion": "0.0000002", "image": "0", "request": "0", "webSearch": "0", @@ -10288,33 +11150,29 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" + "presence_penalty" ], - "inputCost": 0.4, - "outputCost": 1.3 + "inputCost": 0.1, + "outputCost": 0.2, + "throughput": 27.1265, + "latency": 1183 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", - "context": 163840, + "context": 110000, "maxCompletionTokens": null, - "providerModelId": "deepseek-ai/DeepSeek-V3", + "providerModelId": "google/gemma-3-27b-it", "pricing": { - "prompt": "0.0000005", - "completion": "0.0000015", + "prompt": "0.0000001", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -10334,19 +11192,57 @@ export default { "logprobs", "top_logprobs" ], - "inputCost": 0.5, - "outputCost": 1.5 + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 49.53, + "latency": 918 }, { - "name": "Fireworks", - "slug": "fireworks", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", "quantization": null, - "context": 131072, + "context": 32000, "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/deepseek-v3", + "providerModelId": "google/gemma-3-27b-it", "pricing": { - "prompt": "0.0000009", - "completion": "0.0000009", + "prompt": "0.000000119", + "completion": "0.0000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" + ], + "inputCost": 0.12, + "outputCost": 0.2, + "throughput": 31.0395, + "latency": 2320 + }, + { + "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", + "slug": "klusterAi", + "quantization": null, + "context": 64000, + "maxCompletionTokens": 8000, + "providerModelId": "google/gemma-3-27b-it", + "pricing": { + "prompt": "0.0000002", + "completion": "0.00000035", "image": "0", "request": "0", "webSearch": "0", @@ -10354,8 +11250,6 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -10364,29 +11258,131 @@ export default { "presence_penalty", "top_k", "repetition_penalty", - "response_format", - "structured_outputs", "logit_bias", "logprobs", + "top_logprobs", + "min_p", + "seed" + ], + "inputCost": 0.2, + "outputCost": 0.35, + "throughput": 54.094, + "latency": 1346 + }, + { + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "parasail-gemma3-27b-it", + "pricing": { + "prompt": "0.00000025", + "completion": "0.0000004", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "presence_penalty", + "frequency_penalty", + "repetition_penalty", + "top_k" + ], + "inputCost": 0.25, + "outputCost": 0.4, + "throughput": 50.5485, + "latency": 1250.5 + }, + { + "name": "nCompass", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://console.ncompass.tech/&size=256", + "slug": "nCompass", + "quantization": "bf16", + "context": 32768, + "maxCompletionTokens": 32768, + "providerModelId": "google/gemma-3-27b-it", + "pricing": { + "prompt": "0.0000003", + "completion": "0.0000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty" + ], + "inputCost": 0.3, + "outputCost": 0.3, + "throughput": 43.6385, + "latency": 998 + }, + { + "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", + "slug": "inferenceNet", + "quantization": "bf16", + "context": 128000, + "maxCompletionTokens": 8192, + "providerModelId": "google/gemma-3-27b-instruct/bf-16", + "pricing": { + "prompt": "0.0000003", + "completion": "0.0000004", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "logit_bias", "top_logprobs" ], - "inputCost": 0.9, - "outputCost": 0.9 + "inputCost": 0.3, + "outputCost": 0.4, + "throughput": 15.504, + "latency": 2100 } ] }, { - "slug": "meta-llama/llama-3.1-8b-instruct", - "hfSlug": "meta-llama/Meta-Llama-3.1-8B-Instruct", + "slug": "deepseek/deepseek-chat", + "hfSlug": "deepseek-ai/DeepSeek-V3", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-07-23T00:00:00+00:00", + "createdAt": "2024-12-26T19:28:40.559917+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.1 8B Instruct", - "shortName": "Llama 3.1 8B Instruct", - "author": "meta-llama", - "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 8B instruct-tuned version is fast and efficient.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "modelVersionGroupId": "803c32ed-9861-4abf-b5da-7d9c9e6dcf04", - "contextLength": 16384, + "name": "DeepSeek: DeepSeek V3", + "shortName": "DeepSeek V3", + "author": "deepseek", + "description": "DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations reveal that the model outperforms other open-source models and rivals leading closed-source models.\n\nFor model details, please visit [the DeepSeek-V3 repo](https://github.com/deepseek-ai/DeepSeek-V3) for more information, or see the [launch announcement](https://api-docs.deepseek.com/news/news1226).", + "modelVersionGroupId": null, + "contextLength": 163840, "inputModalities": [ "text" ], @@ -10394,34 +11390,31 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "DeepSeek", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.1-8b-instruct", + "permaslug": "deepseek/deepseek-chat-v3", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "612663e2-9ede-4c5d-b993-c55cd12e3055", - "name": "InferenceNet | meta-llama/llama-3.1-8b-instruct", - "contextLength": 16384, + "id": "5294d55f-9012-496b-8f22-8cc919432dcd", + "name": "DeepInfra | deepseek/deepseek-chat-v3", + "contextLength": 163840, "model": { - "slug": "meta-llama/llama-3.1-8b-instruct", - "hfSlug": "meta-llama/Meta-Llama-3.1-8B-Instruct", + "slug": "deepseek/deepseek-chat", + "hfSlug": "deepseek-ai/DeepSeek-V3", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-07-23T00:00:00+00:00", + "createdAt": "2024-12-26T19:28:40.559917+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.1 8B Instruct", - "shortName": "Llama 3.1 8B Instruct", - "author": "meta-llama", - "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 8B instruct-tuned version is fast and efficient.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "modelVersionGroupId": "803c32ed-9861-4abf-b5da-7d9c9e6dcf04", + "name": "DeepSeek: DeepSeek V3", + "shortName": "DeepSeek V3", + "author": "deepseek", + "description": "DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations reveal that the model outperforms other open-source models and rivals leading closed-source models.\n\nFor model details, please visit [the DeepSeek-V3 repo](https://github.com/deepseek-ai/DeepSeek-V3) for more information, or see the [launch announcement](https://api-docs.deepseek.com/news/news1226).", + "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ "text" @@ -10430,41 +11423,40 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "DeepSeek", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.1-8b-instruct", + "permaslug": "deepseek/deepseek-chat-v3", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "meta-llama/llama-3.1-8b-instruct", - "modelVariantPermaslug": "meta-llama/llama-3.1-8b-instruct", - "providerName": "InferenceNet", + "modelVariantSlug": "deepseek/deepseek-chat", + "modelVariantPermaslug": "deepseek/deepseek-chat-v3", + "providerName": "DeepInfra", "providerInfo": { - "name": "InferenceNet", - "displayName": "inference.net", - "slug": "inference-net", + "name": "DeepInfra", + "displayName": "DeepInfra", + "slug": "deepinfra", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://inference.net/privacy-policy", - "termsOfServiceUrl": "https://inference.net/terms-of-service", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false } }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, + "isAbortable": true, "moderationRequired": false, - "group": "InferenceNet", + "group": "DeepInfra", "editors": [], "owners": [], "isMultipartSupported": true, @@ -10472,19 +11464,18 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", - "invertRequired": true + "url": "/images/icons/DeepInfra.webp" } }, - "providerDisplayName": "inference.net", - "providerModelId": "meta-llama/llama-3.1-8b-instruct/fp-16", - "providerGroup": "InferenceNet", - "quantization": "fp16", + "providerDisplayName": "DeepInfra", + "providerModelId": "deepseek-ai/DeepSeek-V3", + "providerGroup": "DeepInfra", + "quantization": "fp8", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 16384, + "maxCompletionTokens": 163840, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -10494,25 +11485,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", + "repetition_penalty", + "response_format", "top_k", - "min_p", - "logit_bias", - "top_logprobs" + "seed", + "min_p" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://inference.net/privacy-policy", - "termsOfServiceUrl": "https://inference.net/terms-of-service", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false }, - "training": false + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0.00000002", - "completion": "0.00000003", + "prompt": "0.00000038", + "completion": "0.00000089", "image": "0", "request": "0", "webSearch": "0", @@ -10531,61 +11525,32 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 28, - "newest": 229, - "throughputHighToLow": 180, - "latencyLowToHigh": 132, - "pricingLowToHigh": 67, - "pricingHighToLow": 242 + "newest": 148, + "throughputHighToLow": 126, + "latencyLowToHigh": 119, + "pricingLowToHigh": 54, + "pricingHighToLow": 146 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [ { - "name": "inference.net", - "slug": "inferenceNet", - "quantization": "fp16", - "context": 16384, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/llama-3.1-8b-instruct/fp-16", - "pricing": { - "prompt": "0.00000002", - "completion": "0.00000003", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "logit_bias", - "top_logprobs" - ], - "inputCost": 0.02, - "outputCost": 0.03 - }, - { - "name": "DeepInfra (Turbo)", - "slug": "deepInfra (turbo)", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo", + "context": 163840, + "maxCompletionTokens": 163840, + "providerModelId": "deepseek-ai/DeepSeek-V3", "pricing": { - "prompt": "0.00000002", - "completion": "0.00000003", + "prompt": "0.00000038", + "completion": "0.00000089", "image": "0", "request": "0", "webSearch": "0", @@ -10593,8 +11558,6 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -10607,19 +11570,22 @@ export default { "seed", "min_p" ], - "inputCost": 0.02, - "outputCost": 0.03 + "inputCost": 0.38, + "outputCost": 0.89, + "throughput": 23.217, + "latency": 691 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", - "quantization": "fp8", - "context": 16384, - "maxCompletionTokens": 8192, - "providerModelId": "meta-llama/llama-3.1-8b-instruct", + "quantization": null, + "context": 64000, + "maxCompletionTokens": 16000, + "providerModelId": "deepseek/deepseek-v3-turbo", "pricing": { - "prompt": "0.00000002", - "completion": "0.00000005", + "prompt": "0.0000004", + "completion": "0.0000013", "image": "0", "request": "0", "webSearch": "0", @@ -10627,6 +11593,8 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", @@ -10639,19 +11607,22 @@ export default { "repetition_penalty", "logit_bias" ], - "inputCost": 0.02, - "outputCost": 0.05 + "inputCost": 0.4, + "outputCost": 1.3, + "throughput": 26.7455, + "latency": 881 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", - "context": 131072, + "context": 163840, "maxCompletionTokens": null, - "providerModelId": "meta-llama/Meta-Llama-3.1-8B-Instruct", + "providerModelId": "deepseek-ai/DeepSeek-V3", "pricing": { - "prompt": "0.00000002", - "completion": "0.00000006", + "prompt": "0.0000005", + "completion": "0.0000015", "image": "0", "request": "0", "webSearch": "0", @@ -10671,51 +11642,22 @@ export default { "logprobs", "top_logprobs" ], - "inputCost": 0.02, - "outputCost": 0.06 - }, - { - "name": "Lambda", - "slug": "lambda", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "llama3.1-8b-instruct", - "pricing": { - "prompt": "0.000000025", - "completion": "0.00000004", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format" - ], - "inputCost": 0.02, - "outputCost": 0.04 + "inputCost": 0.5, + "outputCost": 1.5, + "throughput": 20.1365, + "latency": 420 }, { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", + "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", + "slug": "fireworks", + "quantization": null, "context": 131072, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/Meta-Llama-3.1-8B-Instruct", + "maxCompletionTokens": null, + "providerModelId": "accounts/fireworks/models/deepseek-v3", "pricing": { - "prompt": "0.00000003", - "completion": "0.00000005", + "prompt": "0.0000009", + "completion": "0.0000009", "image": "0", "request": "0", "webSearch": "0", @@ -10728,95 +11670,281 @@ export default { "max_tokens", "temperature", "top_p", - "structured_outputs", - "response_format", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", "top_k", - "seed", - "min_p" + "repetition_penalty", + "response_format", + "structured_outputs", + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 0.03, - "outputCost": 0.05 - }, - { - "name": "Cloudflare", - "slug": "cloudflare", - "quantization": "fp8", - "context": 32000, - "maxCompletionTokens": null, - "providerModelId": "@cf/meta/llama-3.1-8b-instruct-fp8", - "pricing": { - "prompt": "0.000000045", - "completion": "0.000000384", - "image": "0", + "inputCost": 0.9, + "outputCost": 0.9, + "throughput": 50.8755, + "latency": 1740.5 + } + ] + }, + { + "slug": "meta-llama/llama-4-maverick", + "hfSlug": "meta-llama/Llama-4-Maverick-17B-128E-Instruct", + "updatedAt": "2025-04-05T19:58:55.362967+00:00", + "createdAt": "2025-04-05T19:37:02.129674+00:00", + "hfUpdatedAt": null, + "name": "Meta: Llama 4 Maverick (free)", + "shortName": "Llama 4 Maverick (free)", + "author": "meta-llama", + "description": "Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward pass (400B total). It supports multilingual text and image input, and produces multilingual text and code output across 12 supported languages. Optimized for vision-language tasks, Maverick is instruction-tuned for assistant-like behavior, image reasoning, and general-purpose multimodal interaction.\n\nMaverick features early fusion for native multimodality and a 1 million token context window. It was trained on a curated mixture of public, licensed, and Meta-platform data, covering ~22 trillion tokens, with a knowledge cutoff in August 2024. Released on April 5, 2025 under the Llama 4 Community License, Maverick is suited for research and commercial applications requiring advanced multimodal understanding and high model throughput.", + "modelVersionGroupId": null, + "contextLength": 128000, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Other", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "2ff5f508-d666-4051-9625-0c795e746c32", + "name": "Chutes | meta-llama/llama-4-maverick-17b-128e-instruct:free", + "contextLength": 128000, + "model": { + "slug": "meta-llama/llama-4-maverick", + "hfSlug": "meta-llama/Llama-4-Maverick-17B-128E-Instruct", + "updatedAt": "2025-04-05T19:58:55.362967+00:00", + "createdAt": "2025-04-05T19:37:02.129674+00:00", + "hfUpdatedAt": null, + "name": "Meta: Llama 4 Maverick", + "shortName": "Llama 4 Maverick", + "author": "meta-llama", + "description": "Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward pass (400B total). It supports multilingual text and image input, and produces multilingual text and code output across 12 supported languages. Optimized for vision-language tasks, Maverick is instruction-tuned for assistant-like behavior, image reasoning, and general-purpose multimodal interaction.\n\nMaverick features early fusion for native multimodality and a 1 million token context window. It was trained on a curated mixture of public, licensed, and Meta-platform data, covering ~22 trillion tokens, with a knowledge cutoff in August 2024. Released on April 5, 2025 under the Llama 4 Community License, Maverick is suited for research and commercial applications requiring advanced multimodal understanding and high model throughput.", + "modelVersionGroupId": null, + "contextLength": 1048576, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Other", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "meta-llama/llama-4-maverick-17b-128e-instruct", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "meta-llama/llama-4-maverick:free", + "modelVariantPermaslug": "meta-llama/llama-4-maverick-17b-128e-instruct:free", + "providerName": "Chutes", + "providerInfo": { + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://chutes.ai/tos", + "paidModels": { + "training": true, + "retainsPrompts": true + } + }, + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "Chutes", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + } + }, + "providerDisplayName": "Chutes", + "providerModelId": "chutesai/Llama-4-Maverick-17B-128E-Instruct-FP8", + "providerGroup": "Chutes", + "quantization": "fp8", + "variant": "free", + "isFree": true, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": 8, + "maxTokensPerImage": 3342, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "structured_outputs", + "response_format", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://chutes.ai/tos", + "paidModels": { + "training": true, + "retainsPrompts": true + }, + "training": true, + "retainsPrompts": true + }, + "pricing": { + "prompt": "0", + "completion": "0", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportedParameters": { + "responseFormat": true, + "structuredOutputs": true + }, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 23, + "newest": 61, + "throughputHighToLow": 84, + "latencyLowToHigh": 60, + "pricingLowToHigh": 26, + "pricingHighToLow": 169 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", + "providers": [ + { + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "fp8", + "context": 1048576, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", + "pricing": { + "prompt": "0.00000016", + "completion": "0.0000006", + "image": "0.0006684", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", "top_k", "seed", - "repetition_penalty", - "frequency_penalty", - "presence_penalty" + "min_p" ], - "inputCost": 0.04, - "outputCost": 0.38 + "inputCost": 0.16, + "outputCost": 0.6, + "throughput": 74.5405, + "latency": 424 }, { - "name": "Groq", - "slug": "groq", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "llama-3.1-8b-instant", + "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", + "slug": "klusterAi", + "quantization": "fp8", + "context": 1048576, + "maxCompletionTokens": 1048576, + "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", "pricing": { - "prompt": "0.00000005", - "completion": "0.00000008", - "image": "0", + "prompt": "0.00000016", + "completion": "0.0000008", + "image": "0.0008355", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "response_format", - "top_logprobs", - "logprobs", + "top_k", + "repetition_penalty", "logit_bias", + "logprobs", + "top_logprobs", + "min_p", "seed" ], - "inputCost": 0.05, - "outputCost": 0.08 + "inputCost": 0.16, + "outputCost": 0.8, + "throughput": 127.202, + "latency": 513 }, { - "name": "NextBit", - "slug": "nextBit", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "llama3.1:8b", + "context": 1048576, + "maxCompletionTokens": 1048576, + "providerModelId": "meta-llama/llama-4-maverick-17b-128e-instruct-fp8", "pricing": { - "prompt": "0.00000006", - "completion": "0.0000001", - "image": "0", + "prompt": "0.00000017", + "completion": "0.00000085", + "image": "0.0006684", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -10829,22 +11957,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "response_format", - "structured_outputs" + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" ], - "inputCost": 0.06, - "outputCost": 0.1 + "inputCost": 0.17, + "outputCost": 0.85, + "throughput": 58.138, + "latency": 708 }, { - "name": "Hyperbolic", - "slug": "hyperbolic", + "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", + "slug": "lambda", "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Meta-Llama-3.1-8B-Instruct", + "context": 1048576, + "maxCompletionTokens": 1048576, + "providerModelId": "llama-4-maverick-17b-128e-instruct-fp8", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", + "prompt": "0.00000018", + "completion": "0.0000006", "image": "0", "request": "0", "webSearch": "0", @@ -10858,58 +11992,59 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "logprobs", - "top_logprobs", "seed", "logit_bias", - "top_k", - "min_p", - "repetition_penalty" + "logprobs", + "top_logprobs", + "response_format" ], - "inputCost": 0.1, - "outputCost": 0.1 + "inputCost": 0.18, + "outputCost": 0.6, + "throughput": 149.191, + "latency": 681 }, { - "name": "Cerebras", - "slug": "cerebras", - "quantization": "fp16", - "context": 32000, - "maxCompletionTokens": 32000, - "providerModelId": "llama3.1-8b", + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", + "quantization": "fp8", + "context": 1048576, + "maxCompletionTokens": 1048576, + "providerModelId": "parasail-llama-4-maverick-instruct-fp8", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", - "image": "0", + "prompt": "0.00000019", + "completion": "0.00000085", + "image": "0.00070182", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "response_format", - "stop", - "seed", - "logprobs", - "top_logprobs" + "presence_penalty", + "frequency_penalty", + "repetition_penalty", + "top_k" ], - "inputCost": 0.1, - "outputCost": 0.1 + "inputCost": 0.19, + "outputCost": 0.85, + "throughput": 151.3725, + "latency": 492 }, { - "name": "Friendli", - "slug": "friendli", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 8000, - "providerModelId": "meta-llama-3.1-8b-instruct", + "name": "CentML", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://centml.ai/&size=256", + "slug": "centMl", + "quantization": "fp8", + "context": 1048576, + "maxCompletionTokens": 1048576, + "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", + "prompt": "0.0000002", + "completion": "0.0000002", "image": "0", "request": "0", "webSearch": "0", @@ -10917,8 +12052,6 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -10928,23 +12061,24 @@ export default { "seed", "top_k", "min_p", - "repetition_penalty", - "response_format", - "structured_outputs" + "repetition_penalty" ], - "inputCost": 0.1, - "outputCost": 0.1 + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 79.6685, + "latency": 407 }, { - "name": "SambaNova", - "slug": "sambaNova", - "quantization": "bf16", - "context": 16384, - "maxCompletionTokens": 4096, - "providerModelId": "Meta-Llama-3.1-8B-Instruct", + "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", + "slug": "groq", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 8192, + "providerModelId": "meta-llama/llama-4-maverick-17b-128e-instruct", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000002", + "prompt": "0.0000002", + "completion": "0.0000006", "image": "0", "request": "0", "webSearch": "0", @@ -10952,25 +12086,36 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "top_k", - "stop" + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "top_logprobs", + "logprobs", + "logit_bias", + "seed" ], - "inputCost": 0.1, - "outputCost": 0.2 + "inputCost": 0.2, + "outputCost": 0.6, + "throughput": 265.193, + "latency": 474 }, { - "name": "kluster.ai", - "slug": "klusterAi", + "name": "nCompass", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://console.ncompass.tech/&size=256", + "slug": "nCompass", "quantization": "fp8", - "context": 131000, - "maxCompletionTokens": 131000, - "providerModelId": "klusterai/Meta-Llama-3.1-8B-Instruct-Turbo", + "context": 400000, + "maxCompletionTokens": 400000, + "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", "pricing": { - "prompt": "0.00000018", - "completion": "0.00000018", + "prompt": "0.0000002", + "completion": "0.0000007", "image": "0", "request": "0", "webSearch": "0", @@ -10979,21 +12124,32 @@ export default { }, "supportedParameters": [ "max_tokens", - "temperature" + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty" ], - "inputCost": 0.18, - "outputCost": 0.18 + "inputCost": 0.2, + "outputCost": 0.7, + "throughput": 133.7025, + "latency": 439 }, { - "name": "Together", - "slug": "together", - "quantization": "fp8", - "context": 131072, + "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", + "slug": "fireworks", + "quantization": null, + "context": 1048576, "maxCompletionTokens": null, - "providerModelId": "meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo", + "providerModelId": "accounts/fireworks/models/llama4-maverick-instruct-basic", "pricing": { - "prompt": "0.00000018", - "completion": "0.00000018", + "prompt": "0.00000022", + "completion": "0.00000088", "image": "0", "request": "0", "webSearch": "0", @@ -11011,23 +12167,28 @@ export default { "presence_penalty", "top_k", "repetition_penalty", + "response_format", + "structured_outputs", "logit_bias", - "min_p", - "response_format" + "logprobs", + "top_logprobs" ], - "inputCost": 0.18, - "outputCost": 0.18 + "inputCost": 0.22, + "outputCost": 0.88, + "throughput": 83.029, + "latency": 634 }, { - "name": "Fireworks", - "slug": "fireworks", - "quantization": "unknown", - "context": 131072, + "name": "GMICloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://gmicloud.ai/&size=256", + "slug": "gmiCloud", + "quantization": "fp8", + "context": 1048576, "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/llama-v3p1-8b-instruct", + "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", + "prompt": "0.00000025", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -11035,6 +12196,36 @@ export default { "discount": 0 }, "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "seed" + ], + "inputCost": 0.25, + "outputCost": 0.8, + "throughput": 93.446, + "latency": 716 + }, + { + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": "fp8", + "context": 1048576, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8", + "pricing": { + "prompt": "0.00000027", + "completion": "0.00000085", + "image": "0.00090234", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", @@ -11043,25 +12234,26 @@ export default { "presence_penalty", "top_k", "repetition_penalty", - "response_format", - "structured_outputs", "logit_bias", - "logprobs", - "top_logprobs" + "min_p", + "response_format" ], - "inputCost": 0.2, - "outputCost": 0.2 + "inputCost": 0.27, + "outputCost": 0.85, + "throughput": 98.9445, + "latency": 446 }, { - "name": "Avian.io", - "slug": "avianIo", - "quantization": "fp8", + "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", + "slug": "sambaNova", + "quantization": null, "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "Meta-Llama-3.1-8B-Instruct", + "maxCompletionTokens": 4096, + "providerModelId": "Llama-4-Maverick-17B-128E-Instruct", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", + "prompt": "0.00000063", + "completion": "0.0000018", "image": "0", "request": "0", "webSearch": "0", @@ -11069,94 +12261,337 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "response_format", - "seed" + "top_k", + "stop" ], - "inputCost": 0.2, - "outputCost": 0.2 + "inputCost": 0.63, + "outputCost": 1.8, + "throughput": 647.386, + "latency": 804 } ] }, { - "slug": "google/gemma-3-27b-it", + "slug": "x-ai/grok-3-mini-beta", "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-12T05:12:39.645813+00:00", + "updatedAt": "2025-04-10T00:40:58.337579+00:00", + "createdAt": "2025-04-09T23:09:55.305821+00:00", "hfUpdatedAt": null, - "name": "Google: Gemma 3 27B", - "shortName": "Gemma 3 27B", - "author": "google", - "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 27B is Google's latest open source model, successor to [Gemma 2](google/gemma-2-27b-it)", - "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", + "name": "xAI: Grok 3 Mini Beta", + "shortName": "Grok 3 Mini Beta", + "author": "x-ai", + "description": "Grok 3 Mini is a lightweight, smaller thinking model. Unlike traditional models that generate answers immediately, Grok 3 Mini thinks before responding. It’s ideal for reasoning-heavy tasks that don’t demand extensive domain knowledge, and shines in math-specific and quantitative use cases, such as solving challenging puzzles or math problems.\n\nTransparent \"thinking\" traces accessible. Defaults to low reasoning, can boost with setting `reasoning: { effort: \"high\" }`\n\nNote: That there are two xAI endpoints for this model. By default when using this model we will always route you to the base endpoint. If you want the fast endpoint you can add `provider: { sort: throughput}`, to sort by throughput instead. \n", + "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Gemini", - "instructType": "gemma", + "group": "Grok", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "", - "", - "" - ], + "defaultStops": [], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "google/gemma-3-27b-it", + "warningMessage": null, + "permaslug": "x-ai/grok-3-mini-beta", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "8f22002c-c045-446f-a1b9-9896133536b8", - "name": "DeepInfra | google/gemma-3-27b-it", + "id": "e1b8a32c-6b6b-456b-91c9-ede164b895ad", + "name": "xAI | x-ai/grok-3-mini-beta", "contextLength": 131072, "model": { - "slug": "google/gemma-3-27b-it", + "slug": "x-ai/grok-3-mini-beta", "hfSlug": "", + "updatedAt": "2025-04-10T00:40:58.337579+00:00", + "createdAt": "2025-04-09T23:09:55.305821+00:00", + "hfUpdatedAt": null, + "name": "xAI: Grok 3 Mini Beta", + "shortName": "Grok 3 Mini Beta", + "author": "x-ai", + "description": "Grok 3 Mini is a lightweight, smaller thinking model. Unlike traditional models that generate answers immediately, Grok 3 Mini thinks before responding. It’s ideal for reasoning-heavy tasks that don’t demand extensive domain knowledge, and shines in math-specific and quantitative use cases, such as solving challenging puzzles or math problems.\n\nTransparent \"thinking\" traces accessible. Defaults to low reasoning, can boost with setting `reasoning: { effort: \"high\" }`\n\nNote: That there are two xAI endpoints for this model. By default when using this model we will always route you to the base endpoint. If you want the fast endpoint you can add `provider: { sort: throughput}`, to sort by throughput instead. \n", + "modelVersionGroupId": null, + "contextLength": 131072, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Grok", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "x-ai/grok-3-mini-beta", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "x-ai/grok-3-mini-beta", + "modelVariantPermaslug": "x-ai/grok-3-mini-beta", + "providerName": "xAI", + "providerInfo": { + "name": "xAI", + "displayName": "xAI", + "slug": "xai", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://x.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://x.ai/legal/privacy-policy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + } + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "xAI", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": "https://status.x.ai/", + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://x.ai/&size=256" + } + }, + "providerDisplayName": "xAI", + "providerModelId": "grok-3-mini-beta", + "providerGroup": "xAI", + "quantization": null, + "variant": "standard", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "seed", + "logprobs", + "top_logprobs", + "response_format" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://x.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://x.ai/legal/privacy-policy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "pricing": { + "prompt": "0.0000003", + "completion": "0.0000005", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": true, + "supportsReasoning": true, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 30, + "newest": 56, + "throughputHighToLow": 87, + "latencyLowToHigh": 45, + "pricingLowToHigh": 167, + "pricingHighToLow": 152 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://x.ai/\\u0026size=256", + "providers": [ + { + "name": "xAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://x.ai/&size=256", + "slug": "xAi", + "quantization": null, + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "grok-3-mini-beta", + "pricing": { + "prompt": "0.0000003", + "completion": "0.0000005", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "seed", + "logprobs", + "top_logprobs", + "response_format" + ], + "inputCost": 0.3, + "outputCost": 0.5, + "throughput": 87.6125, + "latency": 398 + }, + { + "name": "xAI Fast", + "icon": "", + "slug": "xAiFast", + "quantization": null, + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "grok-3-mini-fast-beta", + "pricing": { + "prompt": "0.0000006", + "completion": "0.000004", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "seed", + "logprobs", + "top_logprobs", + "response_format" + ], + "inputCost": 0.6, + "outputCost": 4, + "throughput": 124.1375, + "latency": 572 + } + ] + }, + { + "slug": "qwen/qwen-2.5-72b-instruct", + "hfSlug": "Qwen/Qwen2.5-72B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-09-19T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Qwen2.5 72B Instruct", + "shortName": "Qwen2.5 72B Instruct", + "author": "qwen", + "description": "Qwen2.5 72B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "modelVersionGroupId": null, + "contextLength": 32768, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Qwen", + "instructType": "chatml", + "defaultSystem": null, + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "qwen/qwen-2.5-72b-instruct", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "8b6b26e9-621a-4b31-b55a-c9aaa7482ede", + "name": "DeepInfra | qwen/qwen-2.5-72b-instruct", + "contextLength": 32768, + "model": { + "slug": "qwen/qwen-2.5-72b-instruct", + "hfSlug": "Qwen/Qwen2.5-72B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-12T05:12:39.645813+00:00", + "createdAt": "2024-09-19T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Google: Gemma 3 27B", - "shortName": "Gemma 3 27B", - "author": "google", - "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 27B is Google's latest open source model, successor to [Gemma 2](google/gemma-2-27b-it)", - "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", + "name": "Qwen2.5 72B Instruct", + "shortName": "Qwen2.5 72B Instruct", + "author": "qwen", + "description": "Qwen2.5 72B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Gemini", - "instructType": "gemma", + "group": "Qwen", + "instructType": "chatml", "defaultSystem": null, "defaultStops": [ - "", - "", - "" + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" ], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "google/gemma-3-27b-it", + "warningMessage": null, + "permaslug": "qwen/qwen-2.5-72b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "google/gemma-3-27b-it", - "modelVariantPermaslug": "google/gemma-3-27b-it", + "modelVariantSlug": "qwen/qwen-2.5-72b-instruct", + "modelVariantPermaslug": "qwen/qwen-2.5-72b-instruct", "providerName": "DeepInfra", "providerInfo": { "name": "DeepInfra", @@ -11189,9 +12624,9 @@ export default { } }, "providerDisplayName": "DeepInfra", - "providerModelId": "google/gemma-3-27b-it", + "providerModelId": "Qwen/Qwen2.5-72B-Instruct", "providerGroup": "DeepInfra", - "quantization": "bf16", + "quantization": "fp8", "variant": "standard", "isFree": false, "canAbort": true, @@ -11200,6 +12635,8 @@ export default { "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", @@ -11226,9 +12663,9 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.0000001", - "completion": "0.0000002", - "image": "0.0000256", + "prompt": "0.00000012", + "completion": "0.00000039", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -11238,7 +12675,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -11246,37 +12683,40 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 29, - "newest": 93, - "throughputHighToLow": 91, - "latencyLowToHigh": 152, - "pricingLowToHigh": 41, - "pricingHighToLow": 199 + "topWeekly": 31, + "newest": 204, + "throughputHighToLow": 221, + "latencyLowToHigh": 135, + "pricingLowToHigh": 64, + "pricingHighToLow": 186 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", - "quantization": "bf16", - "context": 131072, + "quantization": "fp8", + "context": 32768, "maxCompletionTokens": 16384, - "providerModelId": "google/gemma-3-27b-it", + "providerModelId": "Qwen/Qwen2.5-72B-Instruct", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000002", - "image": "0.0000256", + "prompt": "0.00000012", + "completion": "0.00000039", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", @@ -11289,46 +12729,22 @@ export default { "seed", "min_p" ], - "inputCost": 0.1, - "outputCost": 0.2 - }, - { - "name": "InoCloud", - "slug": "inoCloud", - "quantization": "bf16", - "context": 128000, - "maxCompletionTokens": 32768, - "providerModelId": "google/gemma-3-27b-it", - "pricing": { - "prompt": "0.0000001", - "completion": "0.0000002", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty" - ], - "inputCost": 0.1, - "outputCost": 0.2 + "inputCost": 0.12, + "outputCost": 0.39, + "throughput": 33.379, + "latency": 675 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", - "context": 110000, + "context": 131072, "maxCompletionTokens": null, - "providerModelId": "google/gemma-3-27b-it", + "providerModelId": "Qwen/Qwen2.5-72B-Instruct", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", + "prompt": "0.00000013", + "completion": "0.0000004", "image": "0", "request": "0", "webSearch": "0", @@ -11348,19 +12764,22 @@ export default { "logprobs", "top_logprobs" ], - "inputCost": 0.1, - "outputCost": 0.3 + "inputCost": 0.13, + "outputCost": 0.4, + "throughput": 19.889, + "latency": 1954.5 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", - "quantization": null, + "quantization": "unknown", "context": 32000, - "maxCompletionTokens": null, - "providerModelId": "google/gemma-3-27b-it", + "maxCompletionTokens": 4096, + "providerModelId": "qwen/qwen-2.5-72b-instruct", "pricing": { - "prompt": "0.000000119", - "completion": "0.0000002", + "prompt": "0.00000038", + "completion": "0.0000004", "image": "0", "request": "0", "webSearch": "0", @@ -11368,6 +12787,8 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", @@ -11380,41 +12801,21 @@ export default { "repetition_penalty", "logit_bias" ], - "inputCost": 0.12, - "outputCost": 0.2 - }, - { - "name": "kluster.ai", - "slug": "klusterAi", - "quantization": null, - "context": 64000, - "maxCompletionTokens": 8000, - "providerModelId": "google/gemma-3-27b-it", - "pricing": { - "prompt": "0.0000002", - "completion": "0.00000035", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature" - ], - "inputCost": 0.2, - "outputCost": 0.35 + "inputCost": 0.38, + "outputCost": 0.4, + "throughput": 21.3825, + "latency": 4101 }, { - "name": "Parasail", - "slug": "parasail", + "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", + "slug": "hyperbolic", "quantization": "bf16", "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "parasail-gemma3-27b-it", + "maxCompletionTokens": null, + "providerModelId": "Qwen/Qwen2.5-72B-Instruct", "pricing": { - "prompt": "0.00000025", + "prompt": "0.0000004", "completion": "0.0000004", "image": "0", "request": "0", @@ -11426,24 +12827,33 @@ export default { "max_tokens", "temperature", "top_p", - "presence_penalty", + "stop", "frequency_penalty", - "repetition_penalty", - "top_k" + "presence_penalty", + "logprobs", + "top_logprobs", + "seed", + "logit_bias", + "top_k", + "min_p", + "repetition_penalty" ], - "inputCost": 0.25, - "outputCost": 0.4 + "inputCost": 0.4, + "outputCost": 0.4, + "throughput": 32.508, + "latency": 2157 }, { - "name": "nCompass", - "slug": "nCompass", - "quantization": "bf16", + "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", + "slug": "fireworks", + "quantization": "unknown", "context": 32768, - "maxCompletionTokens": 32768, - "providerModelId": "google/gemma-3-27b-it", + "maxCompletionTokens": null, + "providerModelId": "accounts/fireworks/models/qwen2p5-72b-instruct", "pricing": { - "prompt": "0.0000003", - "completion": "0.0000003", + "prompt": "0.0000009", + "completion": "0.0000009", "image": "0", "request": "0", "webSearch": "0", @@ -11451,30 +12861,38 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "seed", "top_k", - "min_p", - "repetition_penalty" + "repetition_penalty", + "response_format", + "structured_outputs", + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 0.3, - "outputCost": 0.3 + "inputCost": 0.9, + "outputCost": 0.9, + "throughput": 70.3035, + "latency": 354 }, { - "name": "inference.net", - "slug": "inferenceNet", - "quantization": "bf16", - "context": 128000, - "maxCompletionTokens": 8192, - "providerModelId": "google/gemma-3-27b-instruct/bf-16", + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": 2048, + "providerModelId": "Qwen/Qwen2.5-72B-Instruct-Turbo", "pricing": { - "prompt": "0.0000003", - "completion": "0.0000004", + "prompt": "0.0000012", + "completion": "0.0000012", "image": "0", "request": "0", "webSearch": "0", @@ -11488,14 +12906,16 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", "top_k", - "min_p", + "repetition_penalty", "logit_bias", - "top_logprobs" + "min_p", + "response_format" ], - "inputCost": 0.3, - "outputCost": 0.4 + "inputCost": 1.2, + "outputCost": 1.2, + "throughput": 95.0335, + "latency": 808 } ] }, @@ -11683,16 +13103,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 11, + "topWeekly": 12, "newest": 143, - "throughputHighToLow": 215, - "latencyLowToHigh": 154, + "throughputHighToLow": 171, + "latencyLowToHigh": 111, "pricingLowToHigh": 53, "pricingHighToLow": 124 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", "context": 163840, @@ -11723,10 +13145,13 @@ export default { "min_p" ], "inputCost": 0.5, - "outputCost": 2.18 + "outputCost": 2.18, + "throughput": 41.738, + "latency": 785 }, { "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", "slug": "inferenceNet", "quantization": "fp8", "context": 128000, @@ -11759,10 +13184,13 @@ export default { "structured_outputs" ], "inputCost": 0.5, - "outputCost": 3 + "outputCost": 3, + "throughput": 19.955, + "latency": 1274.5 }, { "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", "slug": "lambda", "quantization": "fp8", "context": 163840, @@ -11795,10 +13223,13 @@ export default { "response_format" ], "inputCost": 0.54, - "outputCost": 2.18 + "outputCost": 2.18, + "throughput": 36.8895, + "latency": 1027 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": null, "context": 64000, @@ -11831,10 +13262,13 @@ export default { "logit_bias" ], "inputCost": 0.7, - "outputCost": 2.5 + "outputCost": 2.5, + "throughput": 28.961, + "latency": 1270.5 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 163840, @@ -11865,10 +13299,13 @@ export default { "top_logprobs" ], "inputCost": 0.8, - "outputCost": 2.4 + "outputCost": 2.4, + "throughput": 26.319, + "latency": 797 }, { "name": "DeepInfra Turbo", + "icon": "", "slug": "deepInfraTurbo", "quantization": "fp4", "context": 32768, @@ -11899,10 +13336,13 @@ export default { "min_p" ], "inputCost": 1, - "outputCost": 3 + "outputCost": 3, + "throughput": 104.019, + "latency": 522 }, { "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", "slug": "klusterAi", "quantization": "fp8", "context": 163840, @@ -11920,14 +13360,28 @@ export default { "supportedParameters": [ "max_tokens", "temperature", + "top_p", "reasoning", - "include_reasoning" + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", + "seed" ], "inputCost": 1.75, - "outputCost": 5 + "outputCost": 5, + "throughput": 29.2855, + "latency": 962 }, { "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", "slug": "parasail", "quantization": "fp8", "context": 131072, @@ -11954,10 +13408,13 @@ export default { "top_k" ], "inputCost": 1.95, - "outputCost": 5 + "outputCost": 5, + "throughput": 52.0385, + "latency": 557 }, { "name": "Nebius Fast", + "icon": "", "slug": "nebiusFast", "quantization": "fp8", "context": 163840, @@ -11988,10 +13445,13 @@ export default { "top_logprobs" ], "inputCost": 2, - "outputCost": 6 + "outputCost": 6, + "throughput": 51.42, + "latency": 829 }, { "name": "CentML", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://centml.ai/&size=256", "slug": "centMl", "quantization": "fp8", "context": 131072, @@ -12021,10 +13481,13 @@ export default { "repetition_penalty" ], "inputCost": 2.99, - "outputCost": 2.99 + "outputCost": 2.99, + "throughput": 58.293, + "latency": 964 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "fp8", "context": 163840, @@ -12055,10 +13518,13 @@ export default { "response_format" ], "inputCost": 3, - "outputCost": 7 + "outputCost": 7, + "throughput": 49.032, + "latency": 1279 }, { "name": "Friendli", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://friendli.ai/&size=256", "slug": "friendli", "quantization": "fp8", "context": 163840, @@ -12092,10 +13558,13 @@ export default { "structured_outputs" ], "inputCost": 3, - "outputCost": 7 + "outputCost": 7, + "throughput": 59.985, + "latency": 927 }, { "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", "slug": "fireworks", "quantization": "fp8", "context": 163840, @@ -12128,10 +13597,13 @@ export default { "top_logprobs" ], "inputCost": 3, - "outputCost": 8 + "outputCost": 8, + "throughput": 44.893, + "latency": 736 }, { "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", "slug": "sambaNova", "quantization": "fp8", "context": 32768, @@ -12156,10 +13628,13 @@ export default { "stop" ], "inputCost": 5, - "outputCost": 7 + "outputCost": 7, + "throughput": 114.5975, + "latency": 2452 }, { "name": "Minimax", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://minimaxi.com/&size=256", "slug": "minimax", "quantization": null, "context": 64000, @@ -12182,10 +13657,13 @@ export default { "include_reasoning" ], "inputCost": 0.55, - "outputCost": 2.19 + "outputCost": 2.19, + "throughput": 21.5065, + "latency": 2109 }, { "name": "DeepSeek", + "icon": "https://openrouter.ai/images/icons/DeepSeek.png", "slug": "deepSeek", "quantization": "fp8", "context": 64000, @@ -12206,10 +13684,13 @@ export default { "include_reasoning" ], "inputCost": 0.55, - "outputCost": 2.19 + "outputCost": 2.19, + "throughput": 21.5405, + "latency": 4052 }, { "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", "slug": "azure", "quantization": null, "context": 163840, @@ -12232,10 +13713,13 @@ export default { "include_reasoning" ], "inputCost": 1.48, - "outputCost": 5.94 + "outputCost": 5.94, + "throughput": 78.2155, + "latency": 1780 }, { "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", "slug": "featherless", "quantization": "fp8", "context": 32768, @@ -12265,79 +13749,79 @@ export default { "seed" ], "inputCost": 6.5, - "outputCost": 8 + "outputCost": 8, + "throughput": 13.743, + "latency": 2659 } ] }, { - "slug": "openai/gpt-4o-2024-11-20", + "slug": "openai/o4-mini", "hfSlug": "", - "updatedAt": "2025-04-23T22:07:38.265826+00:00", - "createdAt": "2024-11-20T18:33:14.771895+00:00", + "updatedAt": "2025-04-17T14:00:40.898359+00:00", + "createdAt": "2025-04-16T16:29:02.980764+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4o (2024-11-20)", - "shortName": "GPT-4o (2024-11-20)", + "name": "OpenAI: o4 Mini", + "shortName": "o4 Mini", "author": "openai", - "description": "The 2024-11-20 version of GPT-4o offers a leveled-up creative writing ability with more natural, engaging, and tailored writing to improve relevance & readability. It’s also better at working with uploaded files, providing deeper insights & more thorough responses.\n\nGPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.", - "modelVersionGroupId": "76e36b33-358e-477a-be24-09f954fcea74", - "contextLength": 128000, + "description": "OpenAI o4-mini is a compact reasoning model in the o-series, optimized for fast, cost-efficient performance while retaining strong multimodal and agentic capabilities. It supports tool use and demonstrates competitive reasoning and coding performance across benchmarks like AIME (99.5% with Python) and SWE-bench, outperforming its predecessor o3-mini and even approaching o3 in some domains.\n\nDespite its smaller size, o4-mini exhibits high accuracy in STEM tasks, visual problem solving (e.g., MathVista, MMMU), and code editing. It is especially well-suited for high-throughput scenarios where latency or cost is critical. Thanks to its efficient architecture and refined reinforcement learning training, o4-mini can chain tools, generate structured outputs, and solve multi-step tasks with minimal delay—often in under a minute.", + "modelVersionGroupId": null, + "contextLength": 200000, "inputModalities": [ - "text", "image", - "file" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "GPT", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "openai/gpt-4o-2024-11-20", + "warningMessage": "", + "permaslug": "openai/o4-mini-2025-04-16", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "3e86b7c5-bffe-4b60-a3dd-b36451978775", - "name": "OpenAI | openai/gpt-4o-2024-11-20", - "contextLength": 128000, + "id": "bd121898-b27c-4e2c-bc92-278627465a54", + "name": "OpenAI | openai/o4-mini-2025-04-16", + "contextLength": 200000, "model": { - "slug": "openai/gpt-4o-2024-11-20", + "slug": "openai/o4-mini", "hfSlug": "", - "updatedAt": "2025-04-23T22:07:38.265826+00:00", - "createdAt": "2024-11-20T18:33:14.771895+00:00", + "updatedAt": "2025-04-17T14:00:40.898359+00:00", + "createdAt": "2025-04-16T16:29:02.980764+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4o (2024-11-20)", - "shortName": "GPT-4o (2024-11-20)", + "name": "OpenAI: o4 Mini", + "shortName": "o4 Mini", "author": "openai", - "description": "The 2024-11-20 version of GPT-4o offers a leveled-up creative writing ability with more natural, engaging, and tailored writing to improve relevance & readability. It’s also better at working with uploaded files, providing deeper insights & more thorough responses.\n\nGPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.", - "modelVersionGroupId": "76e36b33-358e-477a-be24-09f954fcea74", - "contextLength": 128000, + "description": "OpenAI o4-mini is a compact reasoning model in the o-series, optimized for fast, cost-efficient performance while retaining strong multimodal and agentic capabilities. It supports tool use and demonstrates competitive reasoning and coding performance across benchmarks like AIME (99.5% with Python) and SWE-bench, outperforming its predecessor o3-mini and even approaching o3 in some domains.\n\nDespite its smaller size, o4-mini exhibits high accuracy in STEM tasks, visual problem solving (e.g., MathVista, MMMU), and code editing. It is especially well-suited for high-throughput scenarios where latency or cost is critical. Thanks to its efficient architecture and refined reinforcement learning training, o4-mini can chain tools, generate structured outputs, and solve multi-step tasks with minimal delay—often in under a minute.", + "modelVersionGroupId": null, + "contextLength": 200000, "inputModalities": [ - "text", "image", - "file" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "GPT", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "openai/gpt-4o-2024-11-20", + "warningMessage": "", + "permaslug": "openai/o4-mini-2025-04-16", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/gpt-4o-2024-11-20", - "modelVariantPermaslug": "openai/gpt-4o-2024-11-20", + "modelVariantSlug": "openai/o4-mini", + "modelVariantPermaslug": "openai/o4-mini-2025-04-16", "providerName": "OpenAI", "providerInfo": { "name": "OpenAI", @@ -12373,30 +13857,21 @@ export default { } }, "providerDisplayName": "OpenAI", - "providerModelId": "gpt-4o-2024-11-20", + "providerModelId": "o4-mini-2025-04-16", "providerGroup": "OpenAI", "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 16384, + "maxCompletionTokens": 100000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "tools", "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "web_search_options", "seed", - "logit_bias", - "logprobs", - "top_logprobs", + "max_tokens", "response_format", "structured_outputs" ], @@ -12417,69 +13892,59 @@ export default { "retentionDays": 30 }, "pricing": { - "prompt": "0.0000025", - "completion": "0.00001", - "image": "0.003613", + "prompt": "0.0000011", + "completion": "0.0000044", + "image": "0.0008415", "request": "0", - "inputCacheRead": "0.00000125", + "inputCacheRead": "0.000000275", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [ - { - "type": "search-threshold", - "threshold": "high", - "request": "0.05" - }, - { - "type": "search-threshold", - "threshold": "medium", - "request": "0.035" - }, - { - "type": "search-threshold", - "threshold": "low", - "request": "0.03" - } - ], + "variablePricings": [], "isHidden": false, "isDeranked": false, "isDisabled": false, "supportsToolParameters": true, - "supportsReasoning": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": { + "responseFormat": true, + "structuredOutputs": true + }, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 31, - "newest": 166, - "throughputHighToLow": 156, - "latencyLowToHigh": 87, - "pricingLowToHigh": 260, - "pricingHighToLow": 54 + "topWeekly": 33, + "newest": 45, + "throughputHighToLow": 83, + "latencyLowToHigh": 299, + "pricingLowToHigh": 230, + "pricingHighToLow": 85 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", "slug": "openAi", "quantization": null, - "context": 128000, - "maxCompletionTokens": 16384, - "providerModelId": "gpt-4o-2024-11-20", + "context": 200000, + "maxCompletionTokens": 100000, + "providerModelId": "o4-mini-2025-04-16", "pricing": { - "prompt": "0.0000025", - "completion": "0.00001", - "image": "0.003613", + "prompt": "0.0000011", + "completion": "0.0000044", + "image": "0.0008415", "request": "0", - "inputCacheRead": "0.00000125", + "inputCacheRead": "0.000000275", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -12487,37 +13952,30 @@ export default { "supportedParameters": [ "tools", "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "web_search_options", "seed", - "logit_bias", - "logprobs", - "top_logprobs", + "max_tokens", "response_format", "structured_outputs" ], - "inputCost": 2.5, - "outputCost": 10 + "inputCost": 1.1, + "outputCost": 4.4, + "throughput": 93.6425, + "latency": 4611 } ] }, { - "slug": "qwen/qwen-2.5-72b-instruct", - "hfSlug": "Qwen/Qwen2.5-72B-Instruct", + "slug": "microsoft/wizardlm-2-8x22b", + "hfSlug": "microsoft/WizardLM-2-8x22B", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-19T00:00:00+00:00", + "createdAt": "2024-04-16T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen2.5 72B Instruct", - "shortName": "Qwen2.5 72B Instruct", - "author": "qwen", - "description": "Qwen2.5 72B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "name": "WizardLM-2 8x22B", + "shortName": "WizardLM-2 8x22B", + "author": "microsoft", + "description": "WizardLM-2 8x22B is Microsoft AI's most advanced Wizard model. It demonstrates highly competitive performance compared to leading proprietary models, and it consistently outperforms all existing state-of-the-art opensource models.\n\nIt is an instruct finetune of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b).\n\nTo read more about the model release, [click here](https://wizardlm.github.io/WizardLM2/).\n\n#moe", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 65536, "inputModalities": [ "text" ], @@ -12525,36 +13983,35 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", + "group": "Mistral", + "instructType": "vicuna", "defaultSystem": null, "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" + "USER:", + "" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-2.5-72b-instruct", + "permaslug": "microsoft/wizardlm-2-8x22b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "8b6b26e9-621a-4b31-b55a-c9aaa7482ede", - "name": "DeepInfra | qwen/qwen-2.5-72b-instruct", - "contextLength": 32768, + "id": "03ac4ad1-a230-4ce7-821c-e797305733df", + "name": "DeepInfra | microsoft/wizardlm-2-8x22b", + "contextLength": 65536, "model": { - "slug": "qwen/qwen-2.5-72b-instruct", - "hfSlug": "Qwen/Qwen2.5-72B-Instruct", + "slug": "microsoft/wizardlm-2-8x22b", + "hfSlug": "microsoft/WizardLM-2-8x22B", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-19T00:00:00+00:00", + "createdAt": "2024-04-16T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen2.5 72B Instruct", - "shortName": "Qwen2.5 72B Instruct", - "author": "qwen", - "description": "Qwen2.5 72B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "name": "WizardLM-2 8x22B", + "shortName": "WizardLM-2 8x22B", + "author": "microsoft", + "description": "WizardLM-2 8x22B is Microsoft AI's most advanced Wizard model. It demonstrates highly competitive performance compared to leading proprietary models, and it consistently outperforms all existing state-of-the-art opensource models.\n\nIt is an instruct finetune of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b).\n\nTo read more about the model release, [click here](https://wizardlm.github.io/WizardLM2/).\n\n#moe", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 65536, "inputModalities": [ "text" ], @@ -12562,23 +14019,22 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", + "group": "Mistral", + "instructType": "vicuna", "defaultSystem": null, "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" + "USER:", + "" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-2.5-72b-instruct", + "permaslug": "microsoft/wizardlm-2-8x22b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "qwen/qwen-2.5-72b-instruct", - "modelVariantPermaslug": "qwen/qwen-2.5-72b-instruct", + "modelVariantSlug": "microsoft/wizardlm-2-8x22b", + "modelVariantPermaslug": "microsoft/wizardlm-2-8x22b", "providerName": "DeepInfra", "providerInfo": { "name": "DeepInfra", @@ -12611,9 +14067,9 @@ export default { } }, "providerDisplayName": "DeepInfra", - "providerModelId": "Qwen/Qwen2.5-72B-Instruct", + "providerModelId": "microsoft/WizardLM-2-8x22B", "providerGroup": "DeepInfra", - "quantization": "fp8", + "quantization": "bf16", "variant": "standard", "isFree": false, "canAbort": true, @@ -12622,8 +14078,6 @@ export default { "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -12650,8 +14104,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000012", - "completion": "0.00000039", + "prompt": "0.0000005", + "completion": "0.0000005", "image": "0", "request": "0", "webSearch": "0", @@ -12662,7 +14116,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -12675,24 +14129,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 32, - "newest": 204, - "throughputHighToLow": 222, - "latencyLowToHigh": 78, - "pricingLowToHigh": 64, - "pricingHighToLow": 186 + "topWeekly": 34, + "newest": 269, + "throughputHighToLow": 238, + "latencyLowToHigh": 139, + "pricingLowToHigh": 178, + "pricingHighToLow": 139 }, + "authorIcon": "https://openrouter.ai/images/icons/Microsoft.svg", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", - "quantization": "fp8", - "context": 32768, + "quantization": "bf16", + "context": 65536, "maxCompletionTokens": 16384, - "providerModelId": "Qwen/Qwen2.5-72B-Instruct", + "providerModelId": "microsoft/WizardLM-2-8x22B", "pricing": { - "prompt": "0.00000012", - "completion": "0.00000039", + "prompt": "0.0000005", + "completion": "0.0000005", "image": "0", "request": "0", "webSearch": "0", @@ -12700,8 +14156,6 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -12714,19 +14168,22 @@ export default { "seed", "min_p" ], - "inputCost": 0.12, - "outputCost": 0.39 + "inputCost": 0.5, + "outputCost": 0.5, + "throughput": 33.839, + "latency": 809 }, { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen2.5-72B-Instruct", + "context": 65536, + "maxCompletionTokens": 65536, + "providerModelId": "parasail-wizardlm-2-8x22b", "pricing": { - "prompt": "0.00000013", - "completion": "0.0000004", + "prompt": "0.0000005", + "completion": "0.0000006", "image": "0", "request": "0", "webSearch": "0", @@ -12737,62 +14194,27 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" + "frequency_penalty", + "repetition_penalty", + "top_k" ], - "inputCost": 0.13, - "outputCost": 0.4 + "inputCost": 0.5, + "outputCost": 0.6, + "throughput": 46.339, + "latency": 593 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "unknown", - "context": 32000, - "maxCompletionTokens": 4096, - "providerModelId": "qwen/qwen-2.5-72b-instruct", - "pricing": { - "prompt": "0.00000038", - "completion": "0.0000004", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" - ], - "inputCost": 0.38, - "outputCost": 0.4 - }, - { - "name": "Hyperbolic", - "slug": "hyperbolic", - "quantization": "bf16", - "context": 131072, + "context": 65535, "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen2.5-72B-Instruct", + "providerModelId": "microsoft/wizardlm-2-8x22b", "pricing": { - "prompt": "0.0000004", - "completion": "0.0000004", + "prompt": "0.00000062", + "completion": "0.00000062", "image": "0", "request": "0", "webSearch": "0", @@ -12806,193 +14228,126 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "logprobs", - "top_logprobs", "seed", - "logit_bias", "top_k", "min_p", - "repetition_penalty" - ], - "inputCost": 0.4, - "outputCost": 0.4 - }, - { - "name": "Fireworks", - "slug": "fireworks", - "quantization": "unknown", - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/qwen2p5-72b-instruct", - "pricing": { - "prompt": "0.0000009", - "completion": "0.0000009", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "response_format", - "structured_outputs", - "logit_bias", - "logprobs", - "top_logprobs" - ], - "inputCost": 0.9, - "outputCost": 0.9 - }, - { - "name": "Together", - "slug": "together", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 2048, - "providerModelId": "Qwen/Qwen2.5-72B-Instruct-Turbo", - "pricing": { - "prompt": "0.0000012", - "completion": "0.0000012", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "logit_bias" ], - "inputCost": 1.2, - "outputCost": 1.2 + "inputCost": 0.62, + "outputCost": 0.62, + "throughput": 33.647, + "latency": 973 } ] }, { - "slug": "microsoft/wizardlm-2-8x22b", - "hfSlug": "microsoft/WizardLM-2-8x22B", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-04-16T00:00:00+00:00", + "slug": "openai/gpt-4o", + "hfSlug": null, + "updatedAt": "2025-04-23T22:07:26.226912+00:00", + "createdAt": "2024-05-13T00:00:00+00:00", "hfUpdatedAt": null, - "name": "WizardLM-2 8x22B", - "shortName": "WizardLM-2 8x22B", - "author": "microsoft", - "description": "WizardLM-2 8x22B is Microsoft AI's most advanced Wizard model. It demonstrates highly competitive performance compared to leading proprietary models, and it consistently outperforms all existing state-of-the-art opensource models.\n\nIt is an instruct finetune of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b).\n\nTo read more about the model release, [click here](https://wizardlm.github.io/WizardLM2/).\n\n#moe", - "modelVersionGroupId": null, - "contextLength": 65536, + "name": "OpenAI: GPT-4o", + "shortName": "GPT-4o", + "author": "openai", + "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)\n\n#multimodal", + "modelVersionGroupId": "76e36b33-358e-477a-be24-09f954fcea74", + "contextLength": 128000, "inputModalities": [ - "text" + "text", + "image", + "file" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "vicuna", + "group": "GPT", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "USER:", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "microsoft/wizardlm-2-8x22b", + "permaslug": "openai/gpt-4o", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "03ac4ad1-a230-4ce7-821c-e797305733df", - "name": "DeepInfra | microsoft/wizardlm-2-8x22b", - "contextLength": 65536, + "id": "452a72a0-2c24-4e31-98cb-d6cc1084fb99", + "name": "OpenAI | openai/gpt-4o", + "contextLength": 128000, "model": { - "slug": "microsoft/wizardlm-2-8x22b", - "hfSlug": "microsoft/WizardLM-2-8x22B", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-04-16T00:00:00+00:00", + "slug": "openai/gpt-4o", + "hfSlug": null, + "updatedAt": "2025-04-23T22:07:26.226912+00:00", + "createdAt": "2024-05-13T00:00:00+00:00", "hfUpdatedAt": null, - "name": "WizardLM-2 8x22B", - "shortName": "WizardLM-2 8x22B", - "author": "microsoft", - "description": "WizardLM-2 8x22B is Microsoft AI's most advanced Wizard model. It demonstrates highly competitive performance compared to leading proprietary models, and it consistently outperforms all existing state-of-the-art opensource models.\n\nIt is an instruct finetune of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b).\n\nTo read more about the model release, [click here](https://wizardlm.github.io/WizardLM2/).\n\n#moe", - "modelVersionGroupId": null, - "contextLength": 65536, + "name": "OpenAI: GPT-4o", + "shortName": "GPT-4o", + "author": "openai", + "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)\n\n#multimodal", + "modelVersionGroupId": "76e36b33-358e-477a-be24-09f954fcea74", + "contextLength": 128000, "inputModalities": [ - "text" + "text", + "image", + "file" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "vicuna", + "group": "GPT", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "USER:", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "microsoft/wizardlm-2-8x22b", + "permaslug": "openai/gpt-4o", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "microsoft/wizardlm-2-8x22b", - "modelVariantPermaslug": "microsoft/wizardlm-2-8x22b", - "providerName": "DeepInfra", + "modelVariantSlug": "openai/gpt-4o", + "modelVariantPermaslug": "openai/gpt-4o", + "providerName": "OpenAI", "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false - } + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": false, - "group": "DeepInfra", + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/DeepInfra.webp" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "DeepInfra", - "providerModelId": "microsoft/WizardLM-2-8x22B", - "providerGroup": "DeepInfra", - "quantization": "bf16", + "providerDisplayName": "OpenAI", + "providerModelId": "gpt-4o", + "providerGroup": "OpenAI", + "quantization": "unknown", "variant": "standard", "isFree": false, "canAbort": true, @@ -13001,45 +14356,69 @@ export default { "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", + "web_search_options", "seed", - "min_p" + "logit_bias", + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, + "requiresUserIds": true, "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.0000005", - "completion": "0.0000005", - "image": "0", + "prompt": "0.0000025", + "completion": "0.00001", + "image": "0.003613", "request": "0", + "inputCacheRead": "0.00000125", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [], + "variablePricings": [ + { + "type": "search-threshold", + "threshold": "high", + "request": "0.05" + }, + { + "type": "search-threshold", + "threshold": "medium", + "request": "0.035" + }, + { + "type": "search-threshold", + "threshold": "low", + "request": "0.03" + } + ], "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -13052,123 +14431,111 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 33, - "newest": 269, - "throughputHighToLow": 227, - "latencyLowToHigh": 121, - "pricingLowToHigh": 178, - "pricingHighToLow": 139 + "topWeekly": 35, + "newest": 258, + "throughputHighToLow": 43, + "latencyLowToHigh": 110, + "pricingLowToHigh": 265, + "pricingHighToLow": 24 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", - "context": 65536, + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": "unknown", + "context": 128000, "maxCompletionTokens": 16384, - "providerModelId": "microsoft/WizardLM-2-8x22B", + "providerModelId": "gpt-4o", "pricing": { - "prompt": "0.0000005", - "completion": "0.0000005", - "image": "0", + "prompt": "0.0000025", + "completion": "0.00001", + "image": "0.003613", "request": "0", + "inputCacheRead": "0.00000125", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", + "web_search_options", "seed", - "min_p" - ], - "inputCost": 0.5, - "outputCost": 0.5 - }, - { - "name": "Parasail", - "slug": "parasail", - "quantization": "fp8", - "context": 65536, - "maxCompletionTokens": 65536, - "providerModelId": "parasail-wizardlm-2-8x22b", - "pricing": { - "prompt": "0.0000005", - "completion": "0.0000006", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "presence_penalty", - "frequency_penalty", - "repetition_penalty", - "top_k" + "logit_bias", + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" ], - "inputCost": 0.5, - "outputCost": 0.6 + "inputCost": 2.5, + "outputCost": 10, + "throughput": 49.1885, + "latency": 672 }, { - "name": "NovitaAI", - "slug": "novitaAi", + "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", + "slug": "azure", "quantization": "unknown", - "context": 65535, - "maxCompletionTokens": null, - "providerModelId": "microsoft/wizardlm-2-8x22b", + "context": 128000, + "maxCompletionTokens": 16384, + "providerModelId": "gpt-4o", "pricing": { - "prompt": "0.00000062", - "completion": "0.00000062", - "image": "0", + "prompt": "0.0000025", + "completion": "0.00001", + "image": "0.003613", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", + "web_search_options", "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" + "logit_bias", + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" ], - "inputCost": 0.62, - "outputCost": 0.62 + "inputCost": 2.5, + "outputCost": 10, + "throughput": 111.027, + "latency": 2185 } ] }, { - "slug": "openai/gpt-4o", - "hfSlug": null, - "updatedAt": "2025-04-23T22:07:26.226912+00:00", - "createdAt": "2024-05-13T00:00:00+00:00", + "slug": "openai/gpt-4.1-nano", + "hfSlug": "", + "updatedAt": "2025-05-12T18:45:58.542641+00:00", + "createdAt": "2025-04-14T17:22:49+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4o", - "shortName": "GPT-4o", + "name": "OpenAI: GPT-4.1 Nano", + "shortName": "GPT-4.1 Nano", "author": "openai", - "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)\n\n#multimodal", - "modelVersionGroupId": "76e36b33-358e-477a-be24-09f954fcea74", - "contextLength": 128000, + "description": "For tasks that demand low latency, GPT‑4.1 nano is the fastest and cheapest model in the GPT-4.1 series. It delivers exceptional performance at a small size with its 1 million token context window, and scores 80.1% on MMLU, 50.3% on GPQA, and 9.8% on Aider polyglot coding – even higher than GPT‑4o mini. It’s ideal for tasks like classification or autocompletion.", + "modelVersionGroupId": null, + "contextLength": 1047576, "inputModalities": [ - "text", "image", + "text", "file" ], "outputModalities": [ @@ -13182,28 +14549,28 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-4o", + "permaslug": "openai/gpt-4.1-nano-2025-04-14", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "452a72a0-2c24-4e31-98cb-d6cc1084fb99", - "name": "OpenAI | openai/gpt-4o", - "contextLength": 128000, + "id": "9251cee5-5503-4be9-9439-7ae21ff062a3", + "name": "OpenAI | openai/gpt-4.1-nano-2025-04-14", + "contextLength": 1047576, "model": { - "slug": "openai/gpt-4o", - "hfSlug": null, - "updatedAt": "2025-04-23T22:07:26.226912+00:00", - "createdAt": "2024-05-13T00:00:00+00:00", + "slug": "openai/gpt-4.1-nano", + "hfSlug": "", + "updatedAt": "2025-05-12T18:45:58.542641+00:00", + "createdAt": "2025-04-14T17:22:49+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4o", - "shortName": "GPT-4o", + "name": "OpenAI: GPT-4.1 Nano", + "shortName": "GPT-4.1 Nano", "author": "openai", - "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)\n\n#multimodal", - "modelVersionGroupId": "76e36b33-358e-477a-be24-09f954fcea74", - "contextLength": 128000, + "description": "For tasks that demand low latency, GPT‑4.1 nano is the fastest and cheapest model in the GPT-4.1 series. It delivers exceptional performance at a small size with its 1 million token context window, and scores 80.1% on MMLU, 50.3% on GPQA, and 9.8% on Aider polyglot coding – even higher than GPT‑4o mini. It’s ideal for tasks like classification or autocompletion.", + "modelVersionGroupId": null, + "contextLength": 1047576, "inputModalities": [ - "text", "image", + "text", "file" ], "outputModalities": [ @@ -13217,12 +14584,12 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-4o", + "permaslug": "openai/gpt-4.1-nano-2025-04-14", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/gpt-4o", - "modelVariantPermaslug": "openai/gpt-4o", + "modelVariantSlug": "openai/gpt-4.1-nano", + "modelVariantPermaslug": "openai/gpt-4.1-nano-2025-04-14", "providerName": "OpenAI", "providerInfo": { "name": "OpenAI", @@ -13258,14 +14625,14 @@ export default { } }, "providerDisplayName": "OpenAI", - "providerModelId": "gpt-4o", + "providerModelId": "gpt-4.1-nano-2025-04-14", "providerGroup": "OpenAI", - "quantization": "unknown", + "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 16384, + "maxCompletionTokens": 32768, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -13277,7 +14644,6 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "web_search_options", "seed", "logit_bias", "logprobs", @@ -13302,32 +14668,16 @@ export default { "retentionDays": 30 }, "pricing": { - "prompt": "0.0000025", - "completion": "0.00001", - "image": "0.003613", + "prompt": "0.0000001", + "completion": "0.0000004", + "image": "0", "request": "0", - "inputCacheRead": "0.00000125", + "inputCacheRead": "0.000000025", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [ - { - "type": "search-threshold", - "threshold": "high", - "request": "0.05" - }, - { - "type": "search-threshold", - "threshold": "medium", - "request": "0.035" - }, - { - "type": "search-threshold", - "threshold": "low", - "request": "0.03" - } - ], + "variablePricings": [], "isHidden": false, "isDeranked": false, "isDisabled": false, @@ -13339,68 +14689,35 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 34, - "newest": 258, - "throughputHighToLow": 44, - "latencyLowToHigh": 107, - "pricingLowToHigh": 265, - "pricingHighToLow": 23 + "topWeekly": 36, + "newest": 50, + "throughputHighToLow": 41, + "latencyLowToHigh": 44, + "pricingLowToHigh": 127, + "pricingHighToLow": 189 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", "slug": "openAi", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 16384, - "providerModelId": "gpt-4o", - "pricing": { - "prompt": "0.0000025", - "completion": "0.00001", - "image": "0.003613", - "request": "0", - "inputCacheRead": "0.00000125", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "web_search_options", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" - ], - "inputCost": 2.5, - "outputCost": 10 - }, - { - "name": "Azure", - "slug": "azure", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 16384, - "providerModelId": "gpt-4o", + "quantization": null, + "context": 1047576, + "maxCompletionTokens": 32768, + "providerModelId": "gpt-4.1-nano-2025-04-14", "pricing": { - "prompt": "0.0000025", - "completion": "0.00001", - "image": "0.003613", + "prompt": "0.0000001", + "completion": "0.0000004", + "image": "0", "request": "0", + "inputCacheRead": "0.000000025", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -13414,7 +14731,6 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "web_search_options", "seed", "logit_bias", "logprobs", @@ -13422,8 +14738,10 @@ export default { "response_format", "structured_outputs" ], - "inputCost": 2.5, - "outputCost": 10 + "inputCost": 0.1, + "outputCost": 0.4, + "throughput": 164.3215, + "latency": 409 } ] }, @@ -13580,16 +14898,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 35, + "topWeekly": 37, "newest": 131, - "throughputHighToLow": 75, - "latencyLowToHigh": 264, + "throughputHighToLow": 79, + "latencyLowToHigh": 272, "pricingLowToHigh": 49, "pricingHighToLow": 216 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { "name": "Enfer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://enfer.ai/&size=256", "slug": "enfer", "quantization": null, "context": 28000, @@ -13617,18 +14937,21 @@ export default { "logprobs" ], "inputCost": 0.06, - "outputCost": 0.12 + "outputCost": 0.12, + "throughput": 30.0135, + "latency": 6906.5 }, { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "fp8", + "name": "NextBit", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nextbit256.com/&size=256", + "slug": "nextBit", + "quantization": "bf16", "context": 32768, - "maxCompletionTokens": 16384, - "providerModelId": "mistralai/Mistral-Small-24B-Instruct-2501", + "maxCompletionTokens": null, + "providerModelId": "mistral:small3", "pricing": { "prompt": "0.00000007", - "completion": "0.00000014", + "completion": "0.00000013", "image": "0", "request": "0", "webSearch": "0", @@ -13642,25 +14965,25 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", "response_format", - "top_k", - "seed", - "min_p" + "structured_outputs" ], "inputCost": 0.07, - "outputCost": 0.14 + "outputCost": 0.13, + "throughput": 38.119, + "latency": 1667 }, { - "name": "NextBit", - "slug": "nextBit", - "quantization": "bf16", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "fp8", "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "mistral:small3", + "maxCompletionTokens": 16384, + "providerModelId": "mistralai/Mistral-Small-24B-Instruct-2501", "pricing": { - "prompt": "0.00000008", - "completion": "0.00000025", + "prompt": "0.00000007", + "completion": "0.00000014", "image": "0", "request": "0", "webSearch": "0", @@ -13674,14 +14997,20 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "repetition_penalty", "response_format", - "structured_outputs" + "top_k", + "seed", + "min_p" ], - "inputCost": 0.08, - "outputCost": 0.25 + "inputCost": 0.07, + "outputCost": 0.14, + "throughput": 77.2655, + "latency": 506 }, { "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", "slug": "mistral", "quantization": null, "context": 32768, @@ -13710,10 +15039,13 @@ export default { "seed" ], "inputCost": 0.1, - "outputCost": 0.3 + "outputCost": 0.3, + "throughput": 145.2055, + "latency": 260 }, { "name": "Ubicloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.ubicloud.com/&size=256", "slug": "ubicloud", "quantization": "bf16", "context": 32768, @@ -13741,10 +15073,13 @@ export default { "top_k" ], "inputCost": 0.3, - "outputCost": 0.3 + "outputCost": 0.3, + "throughput": 32.404, + "latency": 762 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": null, "context": 32768, @@ -13773,115 +15108,124 @@ export default { "response_format" ], "inputCost": 0.8, - "outputCost": 0.8 + "outputCost": 0.8, + "throughput": 80.305, + "latency": 654.5 } ] }, { - "slug": "x-ai/grok-3-mini-beta", + "slug": "openai/gpt-4o-2024-11-20", "hfSlug": "", - "updatedAt": "2025-04-10T00:40:58.337579+00:00", - "createdAt": "2025-04-09T23:09:55.305821+00:00", + "updatedAt": "2025-04-23T22:07:38.265826+00:00", + "createdAt": "2024-11-20T18:33:14.771895+00:00", "hfUpdatedAt": null, - "name": "xAI: Grok 3 Mini Beta", - "shortName": "Grok 3 Mini Beta", - "author": "x-ai", - "description": "Grok 3 Mini is a lightweight, smaller thinking model. Unlike traditional models that generate answers immediately, Grok 3 Mini thinks before responding. It’s ideal for reasoning-heavy tasks that don’t demand extensive domain knowledge, and shines in math-specific and quantitative use cases, such as solving challenging puzzles or math problems.\n\nTransparent \"thinking\" traces accessible. Defaults to low reasoning, can boost with setting `reasoning: { effort: \"high\" }`\n\nNote: That there are two xAI endpoints for this model. By default when using this model we will always route you to the base endpoint. If you want the fast endpoint you can add `provider: { sort: throughput}`, to sort by throughput instead. \n", - "modelVersionGroupId": null, - "contextLength": 131072, + "name": "OpenAI: GPT-4o (2024-11-20)", + "shortName": "GPT-4o (2024-11-20)", + "author": "openai", + "description": "The 2024-11-20 version of GPT-4o offers a leveled-up creative writing ability with more natural, engaging, and tailored writing to improve relevance & readability. It’s also better at working with uploaded files, providing deeper insights & more thorough responses.\n\nGPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.", + "modelVersionGroupId": "76e36b33-358e-477a-be24-09f954fcea74", + "contextLength": 128000, "inputModalities": [ - "text" + "text", + "image", + "file" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Grok", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "x-ai/grok-3-mini-beta", + "permaslug": "openai/gpt-4o-2024-11-20", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "e1b8a32c-6b6b-456b-91c9-ede164b895ad", - "name": "xAI | x-ai/grok-3-mini-beta", - "contextLength": 131072, + "id": "3e86b7c5-bffe-4b60-a3dd-b36451978775", + "name": "OpenAI | openai/gpt-4o-2024-11-20", + "contextLength": 128000, "model": { - "slug": "x-ai/grok-3-mini-beta", + "slug": "openai/gpt-4o-2024-11-20", "hfSlug": "", - "updatedAt": "2025-04-10T00:40:58.337579+00:00", - "createdAt": "2025-04-09T23:09:55.305821+00:00", + "updatedAt": "2025-04-23T22:07:38.265826+00:00", + "createdAt": "2024-11-20T18:33:14.771895+00:00", "hfUpdatedAt": null, - "name": "xAI: Grok 3 Mini Beta", - "shortName": "Grok 3 Mini Beta", - "author": "x-ai", - "description": "Grok 3 Mini is a lightweight, smaller thinking model. Unlike traditional models that generate answers immediately, Grok 3 Mini thinks before responding. It’s ideal for reasoning-heavy tasks that don’t demand extensive domain knowledge, and shines in math-specific and quantitative use cases, such as solving challenging puzzles or math problems.\n\nTransparent \"thinking\" traces accessible. Defaults to low reasoning, can boost with setting `reasoning: { effort: \"high\" }`\n\nNote: That there are two xAI endpoints for this model. By default when using this model we will always route you to the base endpoint. If you want the fast endpoint you can add `provider: { sort: throughput}`, to sort by throughput instead. \n", - "modelVersionGroupId": null, - "contextLength": 131072, + "name": "OpenAI: GPT-4o (2024-11-20)", + "shortName": "GPT-4o (2024-11-20)", + "author": "openai", + "description": "The 2024-11-20 version of GPT-4o offers a leveled-up creative writing ability with more natural, engaging, and tailored writing to improve relevance & readability. It’s also better at working with uploaded files, providing deeper insights & more thorough responses.\n\nGPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.", + "modelVersionGroupId": "76e36b33-358e-477a-be24-09f954fcea74", + "contextLength": 128000, "inputModalities": [ - "text" + "text", + "image", + "file" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Grok", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "x-ai/grok-3-mini-beta", + "permaslug": "openai/gpt-4o-2024-11-20", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "x-ai/grok-3-mini-beta", - "modelVariantPermaslug": "x-ai/grok-3-mini-beta", - "providerName": "xAI", + "modelVariantSlug": "openai/gpt-4o-2024-11-20", + "modelVariantPermaslug": "openai/gpt-4o-2024-11-20", + "providerName": "OpenAI", "providerInfo": { - "name": "xAI", - "displayName": "xAI", - "slug": "xai", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://x.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://x.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 - } + }, + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": false, - "group": "xAI", + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.x.ai/", + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://x.ai/&size=256" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "xAI", - "providerModelId": "grok-3-mini-beta", - "providerGroup": "xAI", + "providerDisplayName": "OpenAI", + "providerModelId": "gpt-4o-2024-11-20", + "providerGroup": "OpenAI", "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 16384, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -13890,108 +15234,99 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", + "frequency_penalty", + "presence_penalty", + "web_search_options", "seed", + "logit_bias", "logprobs", "top_logprobs", - "response_format" + "response_format", + "structured_outputs" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://x.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://x.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 }, + "requiresUserIds": true, "training": false, "retainsPrompts": true, "retentionDays": 30 }, "pricing": { - "prompt": "0.0000003", - "completion": "0.0000005", - "image": "0", + "prompt": "0.0000025", + "completion": "0.00001", + "image": "0.003613", "request": "0", + "inputCacheRead": "0.00000125", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [], + "variablePricings": [ + { + "type": "search-threshold", + "threshold": "high", + "request": "0.05" + }, + { + "type": "search-threshold", + "threshold": "medium", + "request": "0.035" + }, + { + "type": "search-threshold", + "threshold": "low", + "request": "0.03" + } + ], "isHidden": false, "isDeranked": false, "isDisabled": false, "supportsToolParameters": true, - "supportsReasoning": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 36, - "newest": 56, - "throughputHighToLow": 46, - "latencyLowToHigh": 44, - "pricingLowToHigh": 167, - "pricingHighToLow": 152 + "topWeekly": 38, + "newest": 166, + "throughputHighToLow": 169, + "latencyLowToHigh": 90, + "pricingLowToHigh": 260, + "pricingHighToLow": 54 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "xAI", - "slug": "xAi", - "quantization": null, - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "grok-3-mini-beta", - "pricing": { - "prompt": "0.0000003", - "completion": "0.0000005", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "seed", - "logprobs", - "top_logprobs", - "response_format" - ], - "inputCost": 0.3, - "outputCost": 0.5 - }, - { - "name": "xAI Fast", - "slug": "xAiFast", + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", "quantization": null, - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "grok-3-mini-fast-beta", + "context": 128000, + "maxCompletionTokens": 16384, + "providerModelId": "gpt-4o-2024-11-20", "pricing": { - "prompt": "0.0000006", - "completion": "0.000004", - "image": "0", + "prompt": "0.0000025", + "completion": "0.00001", + "image": "0.003613", "request": "0", + "inputCacheRead": "0.00000125", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -14002,161 +15337,167 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", + "frequency_penalty", + "presence_penalty", + "web_search_options", "seed", + "logit_bias", "logprobs", "top_logprobs", - "response_format" + "response_format", + "structured_outputs" ], - "inputCost": 0.6, - "outputCost": 4 + "inputCost": 2.5, + "outputCost": 10, + "throughput": 51.497, + "latency": 600 } ] }, { - "slug": "openai/o4-mini", - "hfSlug": "", - "updatedAt": "2025-04-17T14:00:40.898359+00:00", - "createdAt": "2025-04-16T16:29:02.980764+00:00", + "slug": "meta-llama/llama-3.2-1b-instruct", + "hfSlug": "meta-llama/Llama-3.2-1B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-09-25T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenAI: o4 Mini", - "shortName": "o4 Mini", - "author": "openai", - "description": "OpenAI o4-mini is a compact reasoning model in the o-series, optimized for fast, cost-efficient performance while retaining strong multimodal and agentic capabilities. It supports tool use and demonstrates competitive reasoning and coding performance across benchmarks like AIME (99.5% with Python) and SWE-bench, outperforming its predecessor o3-mini and even approaching o3 in some domains.\n\nDespite its smaller size, o4-mini exhibits high accuracy in STEM tasks, visual problem solving (e.g., MathVista, MMMU), and code editing. It is especially well-suited for high-throughput scenarios where latency or cost is critical. Thanks to its efficient architecture and refined reinforcement learning training, o4-mini can chain tools, generate structured outputs, and solve multi-step tasks with minimal delay—often in under a minute.", + "name": "Meta: Llama 3.2 1B Instruct", + "shortName": "Llama 3.2 1B Instruct", + "author": "meta-llama", + "description": "Llama 3.2 1B is a 1-billion-parameter language model focused on efficiently performing natural language tasks, such as summarization, dialogue, and multilingual text analysis. Its smaller size allows it to operate efficiently in low-resource environments while maintaining strong task performance.\n\nSupporting eight core languages and fine-tunable for more, Llama 1.3B is ideal for businesses or developers seeking lightweight yet powerful AI solutions that can operate in diverse multilingual settings without the high computational demand of larger models.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", "modelVersionGroupId": null, - "contextLength": 200000, + "contextLength": 131072, "inputModalities": [ - "image", "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "openai/o4-mini-2025-04-16", + "warningMessage": null, + "permaslug": "meta-llama/llama-3.2-1b-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "bd121898-b27c-4e2c-bc92-278627465a54", - "name": "OpenAI | openai/o4-mini-2025-04-16", - "contextLength": 200000, + "id": "02c4e6a2-dd12-4efc-81ec-12c8bb5c035e", + "name": "Nebius | meta-llama/llama-3.2-1b-instruct", + "contextLength": 131072, "model": { - "slug": "openai/o4-mini", - "hfSlug": "", - "updatedAt": "2025-04-17T14:00:40.898359+00:00", - "createdAt": "2025-04-16T16:29:02.980764+00:00", + "slug": "meta-llama/llama-3.2-1b-instruct", + "hfSlug": "meta-llama/Llama-3.2-1B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-09-25T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenAI: o4 Mini", - "shortName": "o4 Mini", - "author": "openai", - "description": "OpenAI o4-mini is a compact reasoning model in the o-series, optimized for fast, cost-efficient performance while retaining strong multimodal and agentic capabilities. It supports tool use and demonstrates competitive reasoning and coding performance across benchmarks like AIME (99.5% with Python) and SWE-bench, outperforming its predecessor o3-mini and even approaching o3 in some domains.\n\nDespite its smaller size, o4-mini exhibits high accuracy in STEM tasks, visual problem solving (e.g., MathVista, MMMU), and code editing. It is especially well-suited for high-throughput scenarios where latency or cost is critical. Thanks to its efficient architecture and refined reinforcement learning training, o4-mini can chain tools, generate structured outputs, and solve multi-step tasks with minimal delay—often in under a minute.", + "name": "Meta: Llama 3.2 1B Instruct", + "shortName": "Llama 3.2 1B Instruct", + "author": "meta-llama", + "description": "Llama 3.2 1B is a 1-billion-parameter language model focused on efficiently performing natural language tasks, such as summarization, dialogue, and multilingual text analysis. Its smaller size allows it to operate efficiently in low-resource environments while maintaining strong task performance.\n\nSupporting eight core languages and fine-tunable for more, Llama 1.3B is ideal for businesses or developers seeking lightweight yet powerful AI solutions that can operate in diverse multilingual settings without the high computational demand of larger models.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", "modelVersionGroupId": null, - "contextLength": 200000, + "contextLength": 131072, "inputModalities": [ - "image", "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "openai/o4-mini-2025-04-16", + "warningMessage": null, + "permaslug": "meta-llama/llama-3.2-1b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/o4-mini", - "modelVariantPermaslug": "openai/o4-mini-2025-04-16", - "providerName": "OpenAI", + "modelVariantSlug": "meta-llama/llama-3.2-1b-instruct", + "modelVariantPermaslug": "meta-llama/llama-3.2-1b-instruct", + "providerName": "Nebius", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Nebius", + "displayName": "Nebius AI Studio", + "slug": "nebius", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "retainsPrompts": false + } }, - "headquarters": "US", + "headquarters": "DE", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", + "isAbortable": false, + "moderationRequired": false, + "group": "Nebius", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "invertRequired": true } }, - "providerDisplayName": "OpenAI", - "providerModelId": "o4-mini-2025-04-16", - "providerGroup": "OpenAI", - "quantization": null, + "providerDisplayName": "Nebius AI Studio", + "providerModelId": "meta-llama/Llama-3.2-1B-Instruct", + "providerGroup": "Nebius", + "quantization": "fp8", "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 100000, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", - "seed", "max_tokens", - "response_format", - "structured_outputs" + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, - "requiresUserIds": true, "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, "pricing": { - "prompt": "0.0000011", - "completion": "0.0000044", - "image": "0.0008415", + "prompt": "0.000000005", + "completion": "0.00000001", + "image": "0", "request": "0", - "inputCacheRead": "0.000000275", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -14165,263 +15506,117 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": true, + "supportsToolParameters": false, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": { - "responseFormat": true, - "structuredOutputs": true - }, + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 37, - "newest": 45, - "throughputHighToLow": 82, - "latencyLowToHigh": 293, - "pricingLowToHigh": 230, - "pricingHighToLow": 85 + "topWeekly": 39, + "newest": 199, + "throughputHighToLow": 4, + "latencyLowToHigh": 91, + "pricingLowToHigh": 63, + "pricingHighToLow": 247 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { - "name": "OpenAI", - "slug": "openAi", - "quantization": null, - "context": 200000, - "maxCompletionTokens": 100000, - "providerModelId": "o4-mini-2025-04-16", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Llama-3.2-1B-Instruct", "pricing": { - "prompt": "0.0000011", - "completion": "0.0000044", - "image": "0.0008415", + "prompt": "0.000000005", + "completion": "0.00000001", + "image": "0", "request": "0", - "inputCacheRead": "0.000000275", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.01, + "outputCost": 0.01, + "throughput": 31.5455, + "latency": 584.5 + }, + { + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/Llama-3.2-1B-Instruct", + "pricing": { + "prompt": "0.00000001", + "completion": "0.00000001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", "response_format", - "structured_outputs" - ], - "inputCost": 1.1, - "outputCost": 4.4 - } - ] - }, - { - "slug": "openai/gpt-4.1-nano", - "hfSlug": "", - "updatedAt": "2025-05-12T18:45:58.542641+00:00", - "createdAt": "2025-04-14T17:22:49+00:00", - "hfUpdatedAt": null, - "name": "OpenAI: GPT-4.1 Nano", - "shortName": "GPT-4.1 Nano", - "author": "openai", - "description": "For tasks that demand low latency, GPT‑4.1 nano is the fastest and cheapest model in the GPT-4.1 series. It delivers exceptional performance at a small size with its 1 million token context window, and scores 80.1% on MMLU, 50.3% on GPQA, and 9.8% on Aider polyglot coding – even higher than GPT‑4o mini. It’s ideal for tasks like classification or autocompletion.", - "modelVersionGroupId": null, - "contextLength": 1047576, - "inputModalities": [ - "image", - "text", - "file" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "GPT", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "openai/gpt-4.1-nano-2025-04-14", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "9251cee5-5503-4be9-9439-7ae21ff062a3", - "name": "OpenAI | openai/gpt-4.1-nano-2025-04-14", - "contextLength": 1047576, - "model": { - "slug": "openai/gpt-4.1-nano", - "hfSlug": "", - "updatedAt": "2025-05-12T18:45:58.542641+00:00", - "createdAt": "2025-04-14T17:22:49+00:00", - "hfUpdatedAt": null, - "name": "OpenAI: GPT-4.1 Nano", - "shortName": "GPT-4.1 Nano", - "author": "openai", - "description": "For tasks that demand low latency, GPT‑4.1 nano is the fastest and cheapest model in the GPT-4.1 series. It delivers exceptional performance at a small size with its 1 million token context window, and scores 80.1% on MMLU, 50.3% on GPQA, and 9.8% on Aider polyglot coding – even higher than GPT‑4o mini. It’s ideal for tasks like classification or autocompletion.", - "modelVersionGroupId": null, - "contextLength": 1047576, - "inputModalities": [ - "image", - "text", - "file" - ], - "outputModalities": [ - "text" + "top_k", + "seed", + "min_p" ], - "hasTextOutput": true, - "group": "GPT", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "openai/gpt-4.1-nano-2025-04-14", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "openai/gpt-4.1-nano", - "modelVariantPermaslug": "openai/gpt-4.1-nano-2025-04-14", - "providerName": "OpenAI", - "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true - } - }, - "providerDisplayName": "OpenAI", - "providerModelId": "gpt-4.1-nano-2025-04-14", - "providerGroup": "OpenAI", - "quantization": null, - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 32768, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" - ], - "isByok": false, - "moderationRequired": true, - "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "pricing": { - "prompt": "0.0000001", - "completion": "0.0000004", - "image": "0", - "request": "0", - "inputCacheRead": "0.000000025", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null + "inputCost": 0.01, + "outputCost": 0.01, + "throughput": 144.546, + "latency": 649 }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 38, - "newest": 50, - "throughputHighToLow": 22, - "latencyLowToHigh": 52, - "pricingLowToHigh": 127, - "pricingHighToLow": 189 - }, - "providers": [ { - "name": "OpenAI", - "slug": "openAi", - "quantization": null, - "context": 1047576, - "maxCompletionTokens": 32768, - "providerModelId": "gpt-4.1-nano-2025-04-14", + "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", + "slug": "inferenceNet", + "quantization": "fp16", + "context": 16384, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/llama-3.2-1b-instruct/fp-16", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000004", + "prompt": "0.00000001", + "completion": "0.00000001", "image": "0", "request": "0", - "inputCacheRead": "0.000000025", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -14429,14 +15624,76 @@ export default { "frequency_penalty", "presence_penalty", "seed", + "top_k", + "min_p", "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" + "top_logprobs" ], - "inputCost": 0.1, - "outputCost": 0.4 + "inputCost": 0.01, + "outputCost": 0.01, + "throughput": 143.1995, + "latency": 1023 + }, + { + "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", + "slug": "cloudflare", + "quantization": "unknown", + "context": 60000, + "maxCompletionTokens": null, + "providerModelId": "@cf/meta/llama-3.2-1b-instruct", + "pricing": { + "prompt": "0.000000027", + "completion": "0.0000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "top_k", + "seed", + "repetition_penalty", + "frequency_penalty", + "presence_penalty" + ], + "inputCost": 0.03, + "outputCost": 0.2, + "throughput": 277.904, + "latency": 789 + }, + { + "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", + "slug": "sambaNova", + "quantization": "bf16", + "context": 16384, + "maxCompletionTokens": 4096, + "providerModelId": "Meta-Llama-3.2-1B-Instruct", + "pricing": { + "prompt": "0.00000004", + "completion": "0.00000008", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 0.04, + "outputCost": 0.08, + "throughput": 4054.054, + "latency": 826 } ] }, @@ -14602,16 +15859,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 39, + "topWeekly": 40, "newest": 218, - "throughputHighToLow": 143, - "latencyLowToHigh": 208, + "throughputHighToLow": 173, + "latencyLowToHigh": 76, "pricingLowToHigh": 131, "pricingHighToLow": 188 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nousresearch.com/\\u0026size=256", "providers": [ { "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", "slug": "lambda", "quantization": "fp8", "context": 131072, @@ -14640,10 +15899,13 @@ export default { "response_format" ], "inputCost": 0.12, - "outputCost": 0.3 + "outputCost": 0.3, + "throughput": 52.034, + "latency": 1478.5 }, { "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", "slug": "hyperbolic", "quantization": "fp8", "context": 12288, @@ -14676,22 +15938,24 @@ export default { "repetition_penalty" ], "inputCost": 0.4, - "outputCost": 0.4 + "outputCost": 0.4, + "throughput": 34.639, + "latency": 869 } ] }, { - "slug": "qwen/qwen3-32b", - "hfSlug": "Qwen/Qwen3-32B", - "updatedAt": "2025-05-11T22:45:44.273213+00:00", - "createdAt": "2025-04-28T21:32:25.189881+00:00", + "slug": "deepseek/deepseek-chat", + "hfSlug": "deepseek-ai/DeepSeek-V3", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-12-26T19:28:40.559917+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 32B", - "shortName": "Qwen3 32B", - "author": "qwen", - "description": "Qwen3-32B is a dense 32.8B parameter causal language model from the Qwen3 series, optimized for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, coding, and logical inference, and a \"non-thinking\" mode for faster, general-purpose conversation. The model demonstrates strong performance in instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling. ", + "name": "DeepSeek: DeepSeek V3 (free)", + "shortName": "DeepSeek V3 (free)", + "author": "deepseek", + "description": "DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations reveal that the model outperforms other open-source models and rivals leading closed-source models.\n\nFor model details, please visit [the DeepSeek-V3 repo](https://github.com/deepseek-ai/DeepSeek-V3) for more information, or see the [launch announcement](https://api-docs.deepseek.com/news/news1226).", "modelVersionGroupId": null, - "contextLength": 40960, + "contextLength": 163840, "inputModalities": [ "text" ], @@ -14699,41 +15963,30 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "DeepSeek", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen3-32b-04-28", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "permaslug": "deepseek/deepseek-chat-v3", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "6b8c829d-3094-45e7-8139-0a67e09060c3", - "name": "DeepInfra | qwen/qwen3-32b-04-28", - "contextLength": 40960, + "id": "cac452eb-4e38-4952-ad98-f59f68204bda", + "name": "Chutes | deepseek/deepseek-chat-v3:free", + "contextLength": 163840, "model": { - "slug": "qwen/qwen3-32b", - "hfSlug": "Qwen/Qwen3-32B", - "updatedAt": "2025-05-11T22:45:44.273213+00:00", - "createdAt": "2025-04-28T21:32:25.189881+00:00", + "slug": "deepseek/deepseek-chat", + "hfSlug": "deepseek-ai/DeepSeek-V3", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-12-26T19:28:40.559917+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 32B", - "shortName": "Qwen3 32B", - "author": "qwen", - "description": "Qwen3-32B is a dense 32.8B parameter causal language model from the Qwen3 series, optimized for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, coding, and logical inference, and a \"non-thinking\" mode for faster, general-purpose conversation. The model demonstrates strong performance in instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling. ", + "name": "DeepSeek: DeepSeek V3", + "shortName": "DeepSeek V3", + "author": "deepseek", + "description": "DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations reveal that the model outperforms other open-source models and rivals leading closed-source models.\n\nFor model details, please visit [the DeepSeek-V3 repo](https://github.com/deepseek-ai/DeepSeek-V3) for more information, or see the [launch announcement](https://api-docs.deepseek.com/news/news1226).", "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ @@ -14743,51 +15996,37 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "DeepSeek", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen3-32b-04-28", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "permaslug": "deepseek/deepseek-chat-v3", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "qwen/qwen3-32b", - "modelVariantPermaslug": "qwen/qwen3-32b-04-28", - "providerName": "DeepInfra", + "modelVariantSlug": "deepseek/deepseek-chat:free", + "modelVariantPermaslug": "deepseek/deepseek-chat-v3:free", + "providerName": "Chutes", "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "DeepInfra", + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, @@ -14795,15 +16034,15 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/DeepInfra.webp" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "DeepInfra", - "providerModelId": "Qwen/Qwen3-32B", - "providerGroup": "DeepInfra", - "quantization": "fp8", - "variant": "standard", - "isFree": false, + "providerDisplayName": "Chutes", + "providerModelId": "deepseek-ai/DeepSeek-V3", + "providerGroup": "Chutes", + "quantization": null, + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, @@ -14813,33 +16052,31 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", "seed", - "min_p" + "top_k", + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true }, - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -14851,37 +16088,38 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 40, - "newest": 28, - "throughputHighToLow": 141, - "latencyLowToHigh": 173, - "pricingLowToHigh": 12, - "pricingHighToLow": 194 + "topWeekly": 28, + "newest": 148, + "throughputHighToLow": 126, + "latencyLowToHigh": 119, + "pricingLowToHigh": 54, + "pricingHighToLow": 146 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen3-32B", + "context": 163840, + "maxCompletionTokens": 163840, + "providerModelId": "deepseek-ai/DeepSeek-V3", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", + "prompt": "0.00000038", + "completion": "0.00000089", "image": "0", "request": "0", "webSearch": "0", @@ -14892,8 +16130,6 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -14903,19 +16139,22 @@ export default { "seed", "min_p" ], - "inputCost": 0.1, - "outputCost": 0.3 + "inputCost": 0.38, + "outputCost": 0.89, + "throughput": 23.217, + "latency": 691 }, { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen3-32B", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": null, + "context": 64000, + "maxCompletionTokens": 16000, + "providerModelId": "deepseek/deepseek-v3-turbo", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", + "prompt": "0.0000004", + "completion": "0.0000013", "image": "0", "request": "0", "webSearch": "0", @@ -14923,33 +16162,36 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", "seed", "top_k", - "logit_bias", - "logprobs", - "top_logprobs" + "min_p", + "repetition_penalty", + "logit_bias" ], - "inputCost": 0.1, - "outputCost": 0.3 + "inputCost": 0.4, + "outputCost": 1.3, + "throughput": 26.7455, + "latency": 881 }, { - "name": "NovitaAI", - "slug": "novitaAi", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", "quantization": "fp8", - "context": 128000, + "context": 163840, "maxCompletionTokens": null, - "providerModelId": "qwen/qwen3-32b-fp8", + "providerModelId": "deepseek-ai/DeepSeek-V3", "pricing": { - "prompt": "0.0000001", - "completion": "0.00000045", + "prompt": "0.0000005", + "completion": "0.0000015", "image": "0", "request": "0", "webSearch": "0", @@ -14960,30 +16202,31 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", "seed", "top_k", - "min_p", - "repetition_penalty", - "logit_bias" + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 0.1, - "outputCost": 0.45 + "inputCost": 0.5, + "outputCost": 1.5, + "throughput": 20.1365, + "latency": 420 }, { - "name": "Parasail", - "slug": "parasail", - "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": 40960, - "providerModelId": "parasail-qwen3-32b", + "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", + "slug": "fireworks", + "quantization": null, + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "accounts/fireworks/models/deepseek-v3", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000005", + "prompt": "0.0000009", + "completion": "0.0000009", "image": "0", "request": "0", "webSearch": "0", @@ -14991,159 +16234,115 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", - "presence_penalty", + "stop", "frequency_penalty", + "presence_penalty", + "top_k", "repetition_penalty", - "top_k" - ], - "inputCost": 0.1, - "outputCost": 0.5 - }, - { - "name": "GMICloud", - "slug": "gmiCloud", - "quantization": "fp8", - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen3-32B-FP8", - "pricing": { - "prompt": "0.0000003", - "completion": "0.0000006", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "seed" + "response_format", + "structured_outputs", + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 0.3, - "outputCost": 0.6 - }, - { - "name": "SambaNova", - "slug": "sambaNova", - "quantization": null, - "context": 8192, - "maxCompletionTokens": 4096, - "providerModelId": "Qwen3-32B", - "pricing": { - "prompt": "0.0000004", - "completion": "0.0000008", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "top_k", - "stop" - ], - "inputCost": 0.4, - "outputCost": 0.8 + "inputCost": 0.9, + "outputCost": 0.9, + "throughput": 50.8755, + "latency": 1740.5 } ] }, { - "slug": "deepseek/deepseek-chat", - "hfSlug": "deepseek-ai/DeepSeek-V3", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-26T19:28:40.559917+00:00", + "slug": "meta-llama/llama-4-scout", + "hfSlug": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "updatedAt": "2025-04-05T19:34:05.019062+00:00", + "createdAt": "2025-04-05T19:31:59.735804+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: DeepSeek V3 (free)", - "shortName": "DeepSeek V3 (free)", - "author": "deepseek", - "description": "DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations reveal that the model outperforms other open-source models and rivals leading closed-source models.\n\nFor model details, please visit [the DeepSeek-V3 repo](https://github.com/deepseek-ai/DeepSeek-V3) for more information, or see the [launch announcement](https://api-docs.deepseek.com/news/news1226).", + "name": "Meta: Llama 4 Scout", + "shortName": "Llama 4 Scout", + "author": "meta-llama", + "description": "Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimodal input (text and image) and multilingual output (text and code) across 12 supported languages. Designed for assistant-style interaction and visual reasoning, Scout uses 16 experts per forward pass and features a context length of 10 million tokens, with a training corpus of ~40 trillion tokens.\n\nBuilt for high efficiency and local or commercial deployment, Llama 4 Scout incorporates early fusion for seamless modality integration. It is instruction-tuned for use in multilingual chat, captioning, and image understanding tasks. Released under the Llama 4 Community License, it was last trained on data up to August 2024 and launched publicly on April 5, 2025.", "modelVersionGroupId": null, - "contextLength": 163840, + "contextLength": 1048576, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "DeepSeek", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "deepseek/deepseek-chat-v3", + "permaslug": "meta-llama/llama-4-scout-17b-16e-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "cac452eb-4e38-4952-ad98-f59f68204bda", - "name": "Chutes | deepseek/deepseek-chat-v3:free", - "contextLength": 163840, + "id": "cf59e549-6141-4393-b073-3cf13b31c187", + "name": "Lambda | meta-llama/llama-4-scout-17b-16e-instruct", + "contextLength": 1048576, "model": { - "slug": "deepseek/deepseek-chat", - "hfSlug": "deepseek-ai/DeepSeek-V3", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-26T19:28:40.559917+00:00", + "slug": "meta-llama/llama-4-scout", + "hfSlug": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "updatedAt": "2025-04-05T19:34:05.019062+00:00", + "createdAt": "2025-04-05T19:31:59.735804+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: DeepSeek V3", - "shortName": "DeepSeek V3", - "author": "deepseek", - "description": "DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations reveal that the model outperforms other open-source models and rivals leading closed-source models.\n\nFor model details, please visit [the DeepSeek-V3 repo](https://github.com/deepseek-ai/DeepSeek-V3) for more information, or see the [launch announcement](https://api-docs.deepseek.com/news/news1226).", + "name": "Meta: Llama 4 Scout", + "shortName": "Llama 4 Scout", + "author": "meta-llama", + "description": "Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimodal input (text and image) and multilingual output (text and code) across 12 supported languages. Designed for assistant-style interaction and visual reasoning, Scout uses 16 experts per forward pass and features a context length of 10 million tokens, with a training corpus of ~40 trillion tokens.\n\nBuilt for high efficiency and local or commercial deployment, Llama 4 Scout incorporates early fusion for seamless modality integration. It is instruction-tuned for use in multilingual chat, captioning, and image understanding tasks. Released under the Llama 4 Community License, it was last trained on data up to August 2024 and launched publicly on April 5, 2025.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 10000000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "DeepSeek", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "deepseek/deepseek-chat-v3", + "permaslug": "meta-llama/llama-4-scout-17b-16e-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "deepseek/deepseek-chat:free", - "modelVariantPermaslug": "deepseek/deepseek-chat-v3:free", - "providerName": "Chutes", + "modelVariantSlug": "meta-llama/llama-4-scout", + "modelVariantPermaslug": "meta-llama/llama-4-scout-17b-16e-instruct", + "providerName": "Lambda", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "Lambda", + "displayName": "Lambda", + "slug": "lambda", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "privacyPolicyUrl": "https://lambda.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://lambda.ai/legal/terms-of-service", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false } }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": true, + "isAbortable": false, "moderationRequired": false, - "group": "Chutes", + "group": "Lambda", "editors": [], "owners": [], "isMultipartSupported": true, @@ -15151,18 +16350,18 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256" } }, - "providerDisplayName": "Chutes", - "providerModelId": "deepseek-ai/DeepSeek-V3", - "providerGroup": "Chutes", - "quantization": null, - "variant": "free", - "isFree": true, - "canAbort": true, + "providerDisplayName": "Lambda", + "providerModelId": "llama-4-scout-17b-16e-instruct", + "providerGroup": "Lambda", + "quantization": "fp8", + "variant": "standard", + "isFree": false, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 1048576, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -15173,27 +16372,24 @@ export default { "frequency_penalty", "presence_penalty", "seed", - "top_k", - "min_p", - "repetition_penalty", - "logprobs", "logit_bias", - "top_logprobs" + "logprobs", + "top_logprobs", + "response_format" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "privacyPolicyUrl": "https://lambda.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://lambda.ai/legal/terms-of-service", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false }, - "training": true, - "retainsPrompts": true + "training": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.00000008", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -15209,38 +16405,76 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 27, - "newest": 148, - "throughputHighToLow": 126, - "latencyLowToHigh": 106, - "pricingLowToHigh": 54, - "pricingHighToLow": 146 + "topWeekly": 42, + "newest": 63, + "throughputHighToLow": 76, + "latencyLowToHigh": 43, + "pricingLowToHigh": 27, + "pricingHighToLow": 201 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", + "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", + "slug": "lambda", "quantization": "fp8", - "context": 163840, - "maxCompletionTokens": 163840, - "providerModelId": "deepseek-ai/DeepSeek-V3", + "context": 1048576, + "maxCompletionTokens": 1048576, + "providerModelId": "llama-4-scout-17b-16e-instruct", "pricing": { - "prompt": "0.00000038", - "completion": "0.00000089", + "prompt": "0.00000008", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" + ], + "inputCost": 0.08, + "outputCost": 0.3, + "throughput": 99.061, + "latency": 792 + }, + { + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 327680, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "pricing": { + "prompt": "0.00000008", + "completion": "0.0000003", + "image": "0.0003342", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, "supportedParameters": [ "max_tokens", "temperature", @@ -15254,28 +16488,131 @@ export default { "seed", "min_p" ], - "inputCost": 0.38, - "outputCost": 0.89 + "inputCost": 0.08, + "outputCost": 0.3, + "throughput": 44.971, + "latency": 839 + }, + { + "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", + "slug": "klusterAi", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "pricing": { + "prompt": "0.00000008", + "completion": "0.00000045", + "image": "0.0005013", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", + "seed" + ], + "inputCost": 0.08, + "outputCost": 0.45, + "throughput": 59.7695, + "latency": 586 + }, + { + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", + "quantization": "fp8", + "context": 158000, + "maxCompletionTokens": 158000, + "providerModelId": "parasail-llama-4-scout-instruct", + "pricing": { + "prompt": "0.00000009", + "completion": "0.00000048", + "image": "0.00046788", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "presence_penalty", + "frequency_penalty", + "repetition_penalty", + "top_k" + ], + "inputCost": 0.09, + "outputCost": 0.48, + "throughput": 95.9025, + "latency": 595 + }, + { + "name": "CentML", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://centml.ai/&size=256", + "slug": "centMl", + "quantization": "bf16", + "context": 1048576, + "maxCompletionTokens": 1048576, + "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty" + ], + "inputCost": 0.1, + "outputCost": 0.1, + "throughput": 81.311, + "latency": 403 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": null, - "context": 64000, - "maxCompletionTokens": 16000, - "providerModelId": "deepseek/deepseek-v3-turbo", + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "meta-llama/llama-4-scout-17b-16e-instruct", "pricing": { - "prompt": "0.0000004", - "completion": "0.0000013", - "image": "0", + "prompt": "0.0000001", + "completion": "0.0000005", + "image": "0.0003342", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -15288,51 +16625,59 @@ export default { "repetition_penalty", "logit_bias" ], - "inputCost": 0.4, - "outputCost": 1.3 + "inputCost": 0.1, + "outputCost": 0.5, + "throughput": 33.7425, + "latency": 1281 }, { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 163840, - "maxCompletionTokens": null, - "providerModelId": "deepseek-ai/DeepSeek-V3", + "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", + "slug": "groq", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 8192, + "providerModelId": "meta-llama/llama-4-scout-17b-16e-instruct", "pricing": { - "prompt": "0.0000005", - "completion": "0.0000015", - "image": "0", + "prompt": "0.00000011", + "completion": "0.00000034", + "image": "0.00036762", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "logit_bias", + "response_format", + "top_logprobs", "logprobs", - "top_logprobs" + "logit_bias", + "seed" ], - "inputCost": 0.5, - "outputCost": 1.5 + "inputCost": 0.11, + "outputCost": 0.34, + "throughput": 707.6425, + "latency": 628 }, { "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", "slug": "fireworks", "quantization": null, - "context": 131072, + "context": 1048576, "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/deepseek-v3", + "providerModelId": "accounts/fireworks/models/llama4-scout-instruct-basic", "pricing": { - "prompt": "0.0000009", - "completion": "0.0000009", + "prompt": "0.00000015", + "completion": "0.0000006", "image": "0", "request": "0", "webSearch": "0", @@ -15356,23 +16701,125 @@ export default { "logprobs", "top_logprobs" ], - "inputCost": 0.9, - "outputCost": 0.9 + "inputCost": 0.15, + "outputCost": 0.6, + "throughput": 97.1715, + "latency": 651 + }, + { + "name": "GMICloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://gmicloud.ai/&size=256", + "slug": "gmiCloud", + "quantization": "bf16", + "context": 1048576, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "pricing": { + "prompt": "0.00000015", + "completion": "0.0000006", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "seed" + ], + "inputCost": 0.15, + "outputCost": 0.6, + "throughput": 100.051, + "latency": 901 + }, + { + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": null, + "context": 1048576, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "pricing": { + "prompt": "0.00000018", + "completion": "0.00000059", + "image": "0.00090234", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" + ], + "inputCost": 0.18, + "outputCost": 0.59, + "throughput": 98.8715, + "latency": 492 + }, + { + "name": "Cerebras", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cerebras.ai/&size=256", + "slug": "cerebras", + "quantization": null, + "context": 32000, + "maxCompletionTokens": 32000, + "providerModelId": "llama-4-scout-17b-16e-instruct", + "pricing": { + "prompt": "0.00000065", + "completion": "0.00000085", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "structured_outputs", + "response_format", + "stop", + "seed", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.65, + "outputCost": 0.85, + "throughput": 7090.909, + "latency": 325 } ] }, { - "slug": "meta-llama/llama-3.2-1b-instruct", - "hfSlug": "meta-llama/Llama-3.2-1B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-25T00:00:00+00:00", + "slug": "qwen/qwen3-32b", + "hfSlug": "Qwen/Qwen3-32B", + "updatedAt": "2025-05-11T22:45:44.273213+00:00", + "createdAt": "2025-04-28T21:32:25.189881+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.2 1B Instruct", - "shortName": "Llama 3.2 1B Instruct", - "author": "meta-llama", - "description": "Llama 3.2 1B is a 1-billion-parameter language model focused on efficiently performing natural language tasks, such as summarization, dialogue, and multilingual text analysis. Its smaller size allows it to operate efficiently in low-resource environments while maintaining strong task performance.\n\nSupporting eight core languages and fine-tunable for more, Llama 1.3B is ideal for businesses or developers seeking lightweight yet powerful AI solutions that can operate in diverse multilingual settings without the high computational demand of larger models.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "name": "Qwen: Qwen3 32B", + "shortName": "Qwen3 32B", + "author": "qwen", + "description": "Qwen3-32B is a dense 32.8B parameter causal language model from the Qwen3 series, optimized for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, coding, and logical inference, and a \"non-thinking\" mode for faster, general-purpose conversation. The model demonstrates strong performance in instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling. ", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 40960, "inputModalities": [ "text" ], @@ -15380,33 +16827,41 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" + "<|im_start|>", + "<|im_end|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.2-1b-instruct", - "reasoningConfig": null, - "features": {}, + "permaslug": "qwen/qwen3-32b-04-28", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "02c4e6a2-dd12-4efc-81ec-12c8bb5c035e", - "name": "Nebius | meta-llama/llama-3.2-1b-instruct", - "contextLength": 131072, + "id": "6b8c829d-3094-45e7-8139-0a67e09060c3", + "name": "DeepInfra | qwen/qwen3-32b-04-28", + "contextLength": 40960, "model": { - "slug": "meta-llama/llama-3.2-1b-instruct", - "hfSlug": "meta-llama/Llama-3.2-1B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-25T00:00:00+00:00", + "slug": "qwen/qwen3-32b", + "hfSlug": "Qwen/Qwen3-32B", + "updatedAt": "2025-05-11T22:45:44.273213+00:00", + "createdAt": "2025-04-28T21:32:25.189881+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.2 1B Instruct", - "shortName": "Llama 3.2 1B Instruct", - "author": "meta-llama", - "description": "Llama 3.2 1B is a 1-billion-parameter language model focused on efficiently performing natural language tasks, such as summarization, dialogue, and multilingual text analysis. Its smaller size allows it to operate efficiently in low-resource environments while maintaining strong task performance.\n\nSupporting eight core languages and fine-tunable for more, Llama 1.3B is ideal for businesses or developers seeking lightweight yet powerful AI solutions that can operate in diverse multilingual settings without the high computational demand of larger models.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "name": "Qwen: Qwen3 32B", + "shortName": "Qwen3 32B", + "author": "qwen", + "description": "Qwen3-32B is a dense 32.8B parameter causal language model from the Qwen3 series, optimized for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, coding, and logical inference, and a \"non-thinking\" mode for faster, general-purpose conversation. The model demonstrates strong performance in instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling. ", "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ @@ -15416,42 +16871,51 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" + "<|im_start|>", + "<|im_end|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.2-1b-instruct", - "reasoningConfig": null, - "features": {} + "permaslug": "qwen/qwen3-32b-04-28", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "meta-llama/llama-3.2-1b-instruct", - "modelVariantPermaslug": "meta-llama/llama-3.2-1b-instruct", - "providerName": "Nebius", + "modelVariantSlug": "qwen/qwen3-32b", + "modelVariantPermaslug": "qwen/qwen3-32b-04-28", + "providerName": "DeepInfra", "providerInfo": { - "name": "Nebius", - "displayName": "Nebius AI Studio", - "slug": "nebius", + "name": "DeepInfra", + "displayName": "DeepInfra", + "slug": "deepinfra", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { "training": false, "retainsPrompts": false } }, - "headquarters": "DE", + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, + "isAbortable": true, "moderationRequired": false, - "group": "Nebius", + "group": "DeepInfra", "editors": [], "owners": [], "isMultipartSupported": true, @@ -15459,17 +16923,16 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", - "invertRequired": true + "url": "/images/icons/DeepInfra.webp" } }, - "providerDisplayName": "Nebius AI Studio", - "providerModelId": "meta-llama/Llama-3.2-1B-Instruct", - "providerGroup": "Nebius", + "providerDisplayName": "DeepInfra", + "providerModelId": "Qwen/Qwen3-32B", + "providerGroup": "DeepInfra", "quantization": "fp8", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, "maxPromptImages": null, @@ -15478,20 +16941,23 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "seed", + "repetition_penalty", + "response_format", "top_k", - "logit_bias", - "logprobs", - "top_logprobs" + "seed", + "min_p" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { "training": false, "retainsPrompts": false @@ -15500,8 +16966,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.000000005", - "completion": "0.00000001", + "prompt": "0.0000001", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -15513,7 +16979,7 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, @@ -15526,24 +16992,63 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 42, - "newest": 202, - "throughputHighToLow": 53, - "latencyLowToHigh": 50, - "pricingLowToHigh": 63, - "pricingHighToLow": 247 + "topWeekly": 43, + "newest": 28, + "throughputHighToLow": 133, + "latencyLowToHigh": 142, + "pricingLowToHigh": 12, + "pricingHighToLow": 194 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ + { + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "fp8", + "context": 40960, + "maxCompletionTokens": null, + "providerModelId": "Qwen/Qwen3-32B", + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" + ], + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 45.105, + "latency": 873 + }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", - "context": 131072, + "context": 40960, "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-3.2-1B-Instruct", + "providerModelId": "Qwen/Qwen3-32B", "pricing": { - "prompt": "0.000000005", - "completion": "0.00000001", + "prompt": "0.0000001", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -15554,6 +17059,8 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -15563,19 +17070,22 @@ export default { "logprobs", "top_logprobs" ], - "inputCost": 0.01, - "outputCost": 0.01 + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 31.8215, + "latency": 802 }, { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/Llama-3.2-1B-Instruct", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": "fp8", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "qwen/qwen3-32b-fp8", "pricing": { - "prompt": "0.00000001", - "completion": "0.00000001", + "prompt": "0.0000001", + "completion": "0.00000045", "image": "0", "request": "0", "webSearch": "0", @@ -15586,28 +17096,33 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", "seed", - "min_p" + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" ], - "inputCost": 0.01, - "outputCost": 0.01 + "inputCost": 0.1, + "outputCost": 0.45, + "throughput": 30.374, + "latency": 1003.5 }, { - "name": "inference.net", - "slug": "inferenceNet", - "quantization": "fp16", - "context": 16384, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/llama-3.2-1b-instruct/fp-16", + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", + "quantization": "fp8", + "context": 40960, + "maxCompletionTokens": 40960, + "providerModelId": "parasail-qwen3-32b", "pricing": { - "prompt": "0.00000001", - "completion": "0.00000001", + "prompt": "0.0000001", + "completion": "0.0000005", "image": "0", "request": "0", "webSearch": "0", @@ -15618,28 +17133,29 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", + "reasoning", + "include_reasoning", "presence_penalty", - "seed", - "top_k", - "min_p", - "logit_bias", - "top_logprobs" + "frequency_penalty", + "repetition_penalty", + "top_k" ], - "inputCost": 0.01, - "outputCost": 0.01 + "inputCost": 0.1, + "outputCost": 0.5, + "throughput": 49.6365, + "latency": 813 }, { - "name": "Cloudflare", - "slug": "cloudflare", - "quantization": "unknown", - "context": 60000, + "name": "GMICloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://gmicloud.ai/&size=256", + "slug": "gmiCloud", + "quantization": "fp8", + "context": 32768, "maxCompletionTokens": null, - "providerModelId": "@cf/meta/llama-3.2-1b-instruct", + "providerModelId": "Qwen/Qwen3-32B-FP8", "pricing": { - "prompt": "0.000000027", - "completion": "0.0000002", + "prompt": "0.0000003", + "completion": "0.0000006", "image": "0", "request": "0", "webSearch": "0", @@ -15650,25 +17166,59 @@ export default { "max_tokens", "temperature", "top_p", - "top_k", + "reasoning", + "include_reasoning", + "seed" + ], + "inputCost": 0.3, + "outputCost": 0.6, + "throughput": 57.572, + "latency": 734 + }, + { + "name": "Cerebras", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cerebras.ai/&size=256", + "slug": "cerebras", + "quantization": null, + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "qwen-3-32b", + "pricing": { + "prompt": "0.0000004", + "completion": "0.0000008", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", "seed", - "repetition_penalty", - "frequency_penalty", - "presence_penalty" + "logprobs", + "top_logprobs" ], - "inputCost": 0.03, - "outputCost": 0.2 + "inputCost": 0.4, + "outputCost": 0.8, + "throughput": 2145.3435, + "latency": 420 }, { "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", "slug": "sambaNova", - "quantization": "bf16", - "context": 16384, + "quantization": null, + "context": 8192, "maxCompletionTokens": 4096, - "providerModelId": "Meta-Llama-3.2-1B-Instruct", + "providerModelId": "Qwen3-32B", "pricing": { - "prompt": "0.00000004", - "completion": "0.00000008", + "prompt": "0.0000004", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -15679,11 +17229,15 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "top_k", "stop" ], - "inputCost": 0.04, - "outputCost": 0.08 + "inputCost": 0.4, + "outputCost": 0.8, + "throughput": 320.4805, + "latency": 1143 } ] }, @@ -15849,16 +17403,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 43, + "topWeekly": 44, "newest": 57, - "throughputHighToLow": 175, - "latencyLowToHigh": 82, + "throughputHighToLow": 217, + "latencyLowToHigh": 151, "pricingLowToHigh": 267, "pricingHighToLow": 39 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://x.ai/\\u0026size=256", "providers": [ { "name": "xAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://x.ai/&size=256", "slug": "xAi", "quantization": null, "context": 131072, @@ -15888,10 +17444,13 @@ export default { "response_format" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 39.5675, + "latency": 860 }, { "name": "xAI Fast", + "icon": "", "slug": "xAiFast", "quantization": null, "context": 131072, @@ -15921,96 +17480,104 @@ export default { "response_format" ], "inputCost": 5, - "outputCost": 25 + "outputCost": 25, + "throughput": 49.868, + "latency": 898 } ] }, { - "slug": "meta-llama/llama-4-scout", - "hfSlug": "meta-llama/Llama-4-Scout-17B-16E-Instruct", - "updatedAt": "2025-04-05T19:34:05.019062+00:00", - "createdAt": "2025-04-05T19:31:59.735804+00:00", + "slug": "meta-llama/llama-3.2-3b-instruct", + "hfSlug": "meta-llama/Llama-3.2-3B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-09-25T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 4 Scout", - "shortName": "Llama 4 Scout", + "name": "Meta: Llama 3.2 3B Instruct", + "shortName": "Llama 3.2 3B Instruct", "author": "meta-llama", - "description": "Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimodal input (text and image) and multilingual output (text and code) across 12 supported languages. Designed for assistant-style interaction and visual reasoning, Scout uses 16 experts per forward pass and features a context length of 10 million tokens, with a training corpus of ~40 trillion tokens.\n\nBuilt for high efficiency and local or commercial deployment, Llama 4 Scout incorporates early fusion for seamless modality integration. It is instruction-tuned for use in multilingual chat, captioning, and image understanding tasks. Released under the Llama 4 Community License, it was last trained on data up to August 2024 and launched publicly on April 5, 2025.", + "description": "Llama 3.2 3B is a 3-billion-parameter multilingual large language model, optimized for advanced natural language processing tasks like dialogue generation, reasoning, and summarization. Designed with the latest transformer architecture, it supports eight languages, including English, Spanish, and Hindi, and is adaptable for additional languages.\n\nTrained on 9 trillion tokens, the Llama 3.2 3B model excels in instruction-following, complex reasoning, and tool use. Its balanced performance makes it ideal for applications needing accuracy and efficiency in text generation across multilingual settings.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", "modelVersionGroupId": null, - "contextLength": 1048576, + "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-4-scout-17b-16e-instruct", + "permaslug": "meta-llama/llama-3.2-3b-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "cf59e549-6141-4393-b073-3cf13b31c187", - "name": "Lambda | meta-llama/llama-4-scout-17b-16e-instruct", - "contextLength": 1048576, + "id": "e462c1ad-93b6-4047-b27d-239e3ba51989", + "name": "DeepInfra | meta-llama/llama-3.2-3b-instruct", + "contextLength": 131072, "model": { - "slug": "meta-llama/llama-4-scout", - "hfSlug": "meta-llama/Llama-4-Scout-17B-16E-Instruct", - "updatedAt": "2025-04-05T19:34:05.019062+00:00", - "createdAt": "2025-04-05T19:31:59.735804+00:00", + "slug": "meta-llama/llama-3.2-3b-instruct", + "hfSlug": "meta-llama/Llama-3.2-3B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-09-25T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 4 Scout", - "shortName": "Llama 4 Scout", + "name": "Meta: Llama 3.2 3B Instruct", + "shortName": "Llama 3.2 3B Instruct", "author": "meta-llama", - "description": "Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimodal input (text and image) and multilingual output (text and code) across 12 supported languages. Designed for assistant-style interaction and visual reasoning, Scout uses 16 experts per forward pass and features a context length of 10 million tokens, with a training corpus of ~40 trillion tokens.\n\nBuilt for high efficiency and local or commercial deployment, Llama 4 Scout incorporates early fusion for seamless modality integration. It is instruction-tuned for use in multilingual chat, captioning, and image understanding tasks. Released under the Llama 4 Community License, it was last trained on data up to August 2024 and launched publicly on April 5, 2025.", + "description": "Llama 3.2 3B is a 3-billion-parameter multilingual large language model, optimized for advanced natural language processing tasks like dialogue generation, reasoning, and summarization. Designed with the latest transformer architecture, it supports eight languages, including English, Spanish, and Hindi, and is adaptable for additional languages.\n\nTrained on 9 trillion tokens, the Llama 3.2 3B model excels in instruction-following, complex reasoning, and tool use. Its balanced performance makes it ideal for applications needing accuracy and efficiency in text generation across multilingual settings.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", "modelVersionGroupId": null, - "contextLength": 10000000, + "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-4-scout-17b-16e-instruct", + "permaslug": "meta-llama/llama-3.2-3b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "meta-llama/llama-4-scout", - "modelVariantPermaslug": "meta-llama/llama-4-scout-17b-16e-instruct", - "providerName": "Lambda", + "modelVariantSlug": "meta-llama/llama-3.2-3b-instruct", + "modelVariantPermaslug": "meta-llama/llama-3.2-3b-instruct", + "providerName": "DeepInfra", "providerInfo": { - "name": "Lambda", - "displayName": "Lambda", - "slug": "lambda", + "name": "DeepInfra", + "displayName": "DeepInfra", + "slug": "deepinfra", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://lambda.ai/legal/privacy-policy", - "termsOfServiceUrl": "https://lambda.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false } }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, + "isAbortable": true, "moderationRequired": false, - "group": "Lambda", + "group": "DeepInfra", "editors": [], "owners": [], "isMultipartSupported": true, @@ -16018,18 +17585,18 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256" + "url": "/images/icons/DeepInfra.webp" } }, - "providerDisplayName": "Lambda", - "providerModelId": "llama-4-scout-17b-16e-instruct", - "providerGroup": "Lambda", - "quantization": "fp8", + "providerDisplayName": "DeepInfra", + "providerModelId": "meta-llama/Llama-3.2-3B-Instruct", + "providerGroup": "DeepInfra", + "quantization": "bf16", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 1048576, + "maxCompletionTokens": 16384, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -16039,25 +17606,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format" + "min_p" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://lambda.ai/legal/privacy-policy", - "termsOfServiceUrl": "https://lambda.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false }, - "training": false + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0.00000008", - "completion": "0.0000003", + "prompt": "0.00000001", + "completion": "0.00000002", "image": "0", "request": "0", "webSearch": "0", @@ -16073,33 +17643,34 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 44, - "newest": 63, - "throughputHighToLow": 72, - "latencyLowToHigh": 172, - "pricingLowToHigh": 27, - "pricingHighToLow": 201 + "topWeekly": 45, + "newest": 197, + "throughputHighToLow": 1, + "latencyLowToHigh": 11, + "pricingLowToHigh": 61, + "pricingHighToLow": 245 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { - "name": "Lambda", - "slug": "lambda", - "quantization": "fp8", - "context": 1048576, - "maxCompletionTokens": 1048576, - "providerModelId": "llama-4-scout-17b-16e-instruct", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/Llama-3.2-3B-Instruct", "pricing": { - "prompt": "0.00000008", - "completion": "0.0000003", + "prompt": "0.00000001", + "completion": "0.00000002", "image": "0", "request": "0", "webSearch": "0", @@ -16113,26 +17684,29 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format" + "min_p" ], - "inputCost": 0.08, - "outputCost": 0.3 + "inputCost": 0.01, + "outputCost": 0.02, + "throughput": 111.9425, + "latency": 225 }, { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", - "context": 327680, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Llama-3.2-3B-Instruct", "pricing": { - "prompt": "0.00000008", - "completion": "0.0000003", - "image": "0.0003342", + "prompt": "0.00000001", + "completion": "0.00000002", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -16145,26 +17719,29 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", "seed", - "min_p" + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 0.08, - "outputCost": 0.3 + "inputCost": 0.01, + "outputCost": 0.02, + "throughput": 111.2985, + "latency": 245 }, { - "name": "kluster.ai", - "slug": "klusterAi", - "quantization": null, + "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", + "slug": "lambda", + "quantization": "bf16", "context": 131072, "maxCompletionTokens": 131072, - "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "providerModelId": "llama3.2-3b-instruct", "pricing": { - "prompt": "0.00000008", - "completion": "0.00000045", - "image": "0.0005013", + "prompt": "0.000000015", + "completion": "0.000000025", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -16172,22 +17749,34 @@ export default { }, "supportedParameters": [ "max_tokens", - "temperature" + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" ], - "inputCost": 0.08, - "outputCost": 0.45 + "inputCost": 0.01, + "outputCost": 0.02, + "throughput": 304.932, + "latency": 690 }, { - "name": "Parasail", - "slug": "parasail", - "quantization": "fp8", - "context": 158000, - "maxCompletionTokens": 158000, - "providerModelId": "parasail-llama-4-scout-instruct", + "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", + "slug": "inferenceNet", + "quantization": "fp16", + "context": 16384, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/llama-3.2-3b-instruct/fp-16", "pricing": { - "prompt": "0.00000009", - "completion": "0.00000048", - "image": "0.00046788", + "prompt": "0.00000002", + "completion": "0.00000002", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -16197,24 +17786,31 @@ export default { "max_tokens", "temperature", "top_p", - "presence_penalty", + "stop", "frequency_penalty", - "repetition_penalty", - "top_k" + "presence_penalty", + "seed", + "top_k", + "min_p", + "logit_bias", + "top_logprobs" ], - "inputCost": 0.09, - "outputCost": 0.48 + "inputCost": 0.02, + "outputCost": 0.02, + "throughput": 85.349, + "latency": 928 }, { - "name": "CentML", - "slug": "centMl", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", "quantization": "bf16", - "context": 1048576, - "maxCompletionTokens": 1048576, - "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/llama-3.2-3b-instruct", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", + "prompt": "0.00000003", + "completion": "0.00000005", "image": "0", "request": "0", "webSearch": "0", @@ -16222,37 +17818,8 @@ export default { "discount": 0 }, "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty" - ], - "inputCost": 0.1, - "outputCost": 0.1 - }, - { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "meta-llama/llama-4-scout-17b-16e-instruct", - "pricing": { - "prompt": "0.0000001", - "completion": "0.0000005", - "image": "0.0003342", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", @@ -16265,53 +17832,54 @@ export default { "repetition_penalty", "logit_bias" ], - "inputCost": 0.1, - "outputCost": 0.5 + "inputCost": 0.03, + "outputCost": 0.05, + "throughput": 108.3015, + "latency": 561 }, { - "name": "Groq", - "slug": "groq", + "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", + "slug": "cloudflare", "quantization": null, - "context": 131072, - "maxCompletionTokens": 8192, - "providerModelId": "meta-llama/llama-4-scout-17b-16e-instruct", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "@cf/meta/llama-3.2-3b-instruct", "pricing": { - "prompt": "0.00000011", + "prompt": "0.000000051", "completion": "0.00000034", - "image": "0.00036762", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "stop", + "top_k", + "seed", + "repetition_penalty", "frequency_penalty", - "presence_penalty", - "response_format", - "top_logprobs", - "logprobs", - "logit_bias", - "seed" + "presence_penalty" ], - "inputCost": 0.11, - "outputCost": 0.34 + "inputCost": 0.05, + "outputCost": 0.34, + "throughput": 157.464, + "latency": 575 }, { - "name": "Fireworks", - "slug": "fireworks", - "quantization": null, - "context": 1048576, - "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/llama4-scout-instruct-basic", + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/Llama-3.2-3B-Instruct-Turbo", "pricing": { - "prompt": "0.00000015", - "completion": "0.0000006", + "prompt": "0.00000006", + "completion": "0.00000006", "image": "0", "request": "0", "webSearch": "0", @@ -16319,8 +17887,6 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -16329,25 +17895,26 @@ export default { "presence_penalty", "top_k", "repetition_penalty", - "response_format", - "structured_outputs", "logit_bias", - "logprobs", - "top_logprobs" + "min_p", + "response_format" ], - "inputCost": 0.15, - "outputCost": 0.6 + "inputCost": 0.06, + "outputCost": 0.06, + "throughput": 162.995, + "latency": 551 }, { - "name": "GMICloud", - "slug": "gmiCloud", + "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", + "slug": "sambaNova", "quantization": "bf16", - "context": 1048576, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "context": 4096, + "maxCompletionTokens": 4096, + "providerModelId": "Meta-Llama-3.2-3B-Instruct", "pricing": { - "prompt": "0.00000015", - "completion": "0.0000006", + "prompt": "0.00000008", + "completion": "0.00000016", "image": "0", "request": "0", "webSearch": "0", @@ -16358,223 +17925,194 @@ export default { "max_tokens", "temperature", "top_p", - "seed" + "top_k", + "stop" ], - "inputCost": 0.15, - "outputCost": 0.6 + "inputCost": 0.08, + "outputCost": 0.16, + "throughput": 3051.282, + "latency": 653.5 }, { - "name": "Together", - "slug": "together", - "quantization": null, - "context": 1048576, + "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", + "slug": "hyperbolic", + "quantization": "fp8", + "context": 131072, "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "providerModelId": "meta-llama/Llama-3.2-3B-Instruct", "pricing": { - "prompt": "0.00000018", - "completion": "0.00000059", - "image": "0.00090234", + "prompt": "0.0000001", + "completion": "0.0000001", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "top_k", - "repetition_penalty", + "logprobs", + "top_logprobs", + "seed", "logit_bias", + "top_k", "min_p", - "response_format" - ], - "inputCost": 0.18, - "outputCost": 0.59 - }, - { - "name": "Cerebras", - "slug": "cerebras", - "quantization": null, - "context": 32000, - "maxCompletionTokens": 32000, - "providerModelId": "llama-4-scout-17b-16e-instruct", - "pricing": { - "prompt": "0.00000065", - "completion": "0.00000085", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "structured_outputs", - "response_format", - "stop", - "seed", - "logprobs", - "top_logprobs" + "repetition_penalty" ], - "inputCost": 0.65, - "outputCost": 0.85 + "inputCost": 0.1, + "outputCost": 0.1, + "throughput": 113.333, + "latency": 1108 } ] }, { - "slug": "meta-llama/llama-3.2-3b-instruct", - "hfSlug": "meta-llama/Llama-3.2-3B-Instruct", + "slug": "anthropic/claude-3.5-sonnet", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-25T00:00:00+00:00", + "createdAt": "2024-10-22T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.2 3B Instruct", - "shortName": "Llama 3.2 3B Instruct", - "author": "meta-llama", - "description": "Llama 3.2 3B is a 3-billion-parameter multilingual large language model, optimized for advanced natural language processing tasks like dialogue generation, reasoning, and summarization. Designed with the latest transformer architecture, it supports eight languages, including English, Spanish, and Hindi, and is adaptable for additional languages.\n\nTrained on 9 trillion tokens, the Llama 3.2 3B model excels in instruction-following, complex reasoning, and tool use. Its balanced performance makes it ideal for applications needing accuracy and efficiency in text generation across multilingual settings.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", - "modelVersionGroupId": null, - "contextLength": 131072, + "name": "Anthropic: Claude 3.5 Sonnet (self-moderated)", + "shortName": "Claude 3.5 Sonnet (self-moderated)", + "author": "anthropic", + "description": "New Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Scores ~49% on SWE-Bench Verified, higher than the last best score, and without any fancy prompt scaffolding\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\n#multimodal", + "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", + "contextLength": 200000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Claude", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.2-3b-instruct", + "permaslug": "anthropic/claude-3.5-sonnet", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "e462c1ad-93b6-4047-b27d-239e3ba51989", - "name": "DeepInfra | meta-llama/llama-3.2-3b-instruct", - "contextLength": 131072, + "id": "07be01e7-3f6a-4b0f-854b-103b7b0a7ad5", + "name": "Anthropic | anthropic/claude-3.5-sonnet:beta", + "contextLength": 200000, "model": { - "slug": "meta-llama/llama-3.2-3b-instruct", - "hfSlug": "meta-llama/Llama-3.2-3B-Instruct", + "slug": "anthropic/claude-3.5-sonnet", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-25T00:00:00+00:00", + "createdAt": "2024-10-22T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.2 3B Instruct", - "shortName": "Llama 3.2 3B Instruct", - "author": "meta-llama", - "description": "Llama 3.2 3B is a 3-billion-parameter multilingual large language model, optimized for advanced natural language processing tasks like dialogue generation, reasoning, and summarization. Designed with the latest transformer architecture, it supports eight languages, including English, Spanish, and Hindi, and is adaptable for additional languages.\n\nTrained on 9 trillion tokens, the Llama 3.2 3B model excels in instruction-following, complex reasoning, and tool use. Its balanced performance makes it ideal for applications needing accuracy and efficiency in text generation across multilingual settings.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", - "modelVersionGroupId": null, - "contextLength": 131072, + "name": "Anthropic: Claude 3.5 Sonnet", + "shortName": "Claude 3.5 Sonnet", + "author": "anthropic", + "description": "New Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Scores ~49% on SWE-Bench Verified, higher than the last best score, and without any fancy prompt scaffolding\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\n#multimodal", + "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", + "contextLength": 200000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Claude", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.2-3b-instruct", + "permaslug": "anthropic/claude-3.5-sonnet", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "meta-llama/llama-3.2-3b-instruct", - "modelVariantPermaslug": "meta-llama/llama-3.2-3b-instruct", - "providerName": "DeepInfra", + "modelVariantSlug": "anthropic/claude-3.5-sonnet:beta", + "modelVariantPermaslug": "anthropic/claude-3.5-sonnet:beta", + "providerName": "Anthropic", "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", + "name": "Anthropic", + "displayName": "Anthropic", + "slug": "anthropic", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", "paidModels": { "training": false, - "retainsPrompts": false - } + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": false, - "group": "DeepInfra", + "moderationRequired": true, + "group": "Anthropic", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.anthropic.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/DeepInfra.webp" + "url": "/images/icons/Anthropic.svg" } }, - "providerDisplayName": "DeepInfra", - "providerModelId": "meta-llama/Llama-3.2-3B-Instruct", - "providerGroup": "DeepInfra", - "quantization": "bf16", - "variant": "standard", + "providerDisplayName": "Anthropic", + "providerModelId": "claude-3-5-sonnet-20241022", + "providerGroup": "Anthropic", + "quantization": "unknown", + "variant": "beta", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 16384, + "maxCompletionTokens": 8192, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", "top_k", - "seed", - "min_p" + "stop" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, + "requiresUserIds": true, "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.00000001", - "completion": "0.00000002", - "image": "0", + "prompt": "0.000003", + "completion": "0.000015", + "image": "0.0048", "request": "0", + "inputCacheRead": "0.0000003", + "inputCacheWrite": "0.00000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -16583,7 +18121,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -16596,154 +18134,127 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 45, - "newest": 197, - "throughputHighToLow": 3, - "latencyLowToHigh": 17, - "pricingLowToHigh": 61, - "pricingHighToLow": 245 + "topWeekly": 19, + "newest": 183, + "throughputHighToLow": 137, + "latencyLowToHigh": 210, + "pricingLowToHigh": 272, + "pricingHighToLow": 44 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/Llama-3.2-3B-Instruct", + "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", + "slug": "anthropic", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 8192, + "providerModelId": "claude-3-5-sonnet-20241022", "pricing": { - "prompt": "0.00000001", - "completion": "0.00000002", - "image": "0", + "prompt": "0.000003", + "completion": "0.000015", + "image": "0.0048", "request": "0", + "inputCacheRead": "0.0000003", + "inputCacheWrite": "0.00000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", "top_k", - "seed", - "min_p" + "stop" ], - "inputCost": 0.01, - "outputCost": 0.02 + "inputCost": 3, + "outputCost": 15, + "throughput": 57.5795, + "latency": 1371 }, { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-3.2-3B-Instruct", + "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", + "slug": "vertex", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 8192, + "providerModelId": "claude-3-5-sonnet-v2@20241022", "pricing": { - "prompt": "0.00000001", - "completion": "0.00000002", - "image": "0", + "prompt": "0.000003", + "completion": "0.000015", + "image": "0.0048", "request": "0", + "inputCacheRead": "0.0000003", + "inputCacheWrite": "0.00000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", "top_k", - "logit_bias", - "logprobs", - "top_logprobs" - ], - "inputCost": 0.01, - "outputCost": 0.02 - }, - { - "name": "Lambda", - "slug": "lambda", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "llama3.2-3b-instruct", - "pricing": { - "prompt": "0.000000015", - "completion": "0.000000025", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format" + "stop" ], - "inputCost": 0.01, - "outputCost": 0.02 + "inputCost": 3, + "outputCost": 15, + "throughput": 61.168, + "latency": 1377 }, { - "name": "inference.net", - "slug": "inferenceNet", - "quantization": "fp16", - "context": 16384, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/llama-3.2-3b-instruct/fp-16", + "name": "Amazon Bedrock", + "icon": "https://openrouter.ai/images/icons/Bedrock.svg", + "slug": "amazonBedrock", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 8192, + "providerModelId": "us.anthropic.claude-3-5-sonnet-20241022-v2:0", "pricing": { - "prompt": "0.00000002", - "completion": "0.00000002", - "image": "0", + "prompt": "0.000003", + "completion": "0.000015", + "image": "0.0048", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", "top_k", - "min_p", - "logit_bias", - "top_logprobs" + "stop" ], - "inputCost": 0.02, - "outputCost": 0.02 + "inputCost": 3, + "outputCost": 15, + "throughput": 35.7255, + "latency": 1781.5 }, { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/llama-3.2-3b-instruct", + "name": "Amazon Bedrock", + "icon": "https://openrouter.ai/images/icons/Bedrock.svg", + "slug": "amazonBedrock", + "quantization": null, + "context": 200000, + "maxCompletionTokens": 8192, + "providerModelId": "us.anthropic.claude-3-5-sonnet-20241022-v2:0", "pricing": { - "prompt": "0.00000003", - "completion": "0.00000005", - "image": "0", + "prompt": "0.000003", + "completion": "0.000015", + "image": "0.0048", "request": "0", + "inputCacheRead": "0.0000003", + "inputCacheWrite": "0.00000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -16754,152 +18265,60 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", "top_k", - "min_p", - "repetition_penalty", - "logit_bias" + "stop" ], - "inputCost": 0.03, - "outputCost": 0.05 + "inputCost": 3, + "outputCost": 15, + "throughput": 37.866, + "latency": 1632 }, { - "name": "Cloudflare", - "slug": "cloudflare", - "quantization": null, - "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "@cf/meta/llama-3.2-3b-instruct", + "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", + "slug": "vertex", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 8192, + "providerModelId": "claude-3-5-sonnet-v2@20241022", "pricing": { - "prompt": "0.000000051", - "completion": "0.00000034", - "image": "0", + "prompt": "0.000003", + "completion": "0.000015", + "image": "0.0048", "request": "0", + "inputCacheRead": "0.0000003", + "inputCacheWrite": "0.00000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "top_k", - "seed", - "repetition_penalty", - "frequency_penalty", - "presence_penalty" + "stop" ], - "inputCost": 0.05, - "outputCost": 0.34 - }, - { - "name": "Together", - "slug": "together", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/Llama-3.2-3B-Instruct-Turbo", - "pricing": { - "prompt": "0.00000006", - "completion": "0.00000006", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" - ], - "inputCost": 0.06, - "outputCost": 0.06 - }, - { - "name": "SambaNova", - "slug": "sambaNova", - "quantization": "bf16", - "context": 4096, - "maxCompletionTokens": 4096, - "providerModelId": "Meta-Llama-3.2-3B-Instruct", - "pricing": { - "prompt": "0.00000008", - "completion": "0.00000016", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "top_k", - "stop" - ], - "inputCost": 0.08, - "outputCost": 0.16 - }, - { - "name": "Hyperbolic", - "slug": "hyperbolic", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-3.2-3B-Instruct", - "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "logprobs", - "top_logprobs", - "seed", - "logit_bias", - "top_k", - "min_p", - "repetition_penalty" - ], - "inputCost": 0.1, - "outputCost": 0.1 + "inputCost": 3, + "outputCost": 15, + "throughput": 59.922, + "latency": 1022 } ] }, { - "slug": "anthropic/claude-3.5-sonnet", + "slug": "anthropic/claude-3.5-haiku", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-10-22T00:00:00+00:00", + "createdAt": "2024-11-04T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3.5 Sonnet (self-moderated)", - "shortName": "Claude 3.5 Sonnet (self-moderated)", + "name": "Anthropic: Claude 3.5 Haiku", + "shortName": "Claude 3.5 Haiku", "author": "anthropic", - "description": "New Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Scores ~49% on SWE-Bench Verified, higher than the last best score, and without any fancy prompt scaffolding\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\n#multimodal", - "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", + "description": "Claude 3.5 Haiku features offers enhanced capabilities in speed, coding accuracy, and tool use. Engineered to excel in real-time applications, it delivers quick response times that are essential for dynamic tasks such as chat interactions and immediate coding suggestions.\n\nThis makes it highly suitable for environments that demand both speed and precision, such as software development, customer service bots, and data management systems.\n\nThis model is currently pointing to [Claude 3.5 Haiku (2024-10-22)](/anthropic/claude-3-5-haiku-20241022).", + "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", "contextLength": 200000, "inputModalities": [ "text", @@ -16916,24 +18335,24 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3.5-sonnet", + "permaslug": "anthropic/claude-3-5-haiku", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "07be01e7-3f6a-4b0f-854b-103b7b0a7ad5", - "name": "Anthropic | anthropic/claude-3.5-sonnet:beta", + "id": "ced809d4-9715-4634-8b4d-ea0e09d6bb85", + "name": "Anthropic | anthropic/claude-3-5-haiku", "contextLength": 200000, "model": { - "slug": "anthropic/claude-3.5-sonnet", + "slug": "anthropic/claude-3.5-haiku", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-10-22T00:00:00+00:00", + "createdAt": "2024-11-04T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3.5 Sonnet", - "shortName": "Claude 3.5 Sonnet", + "name": "Anthropic: Claude 3.5 Haiku", + "shortName": "Claude 3.5 Haiku", "author": "anthropic", - "description": "New Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Scores ~49% on SWE-Bench Verified, higher than the last best score, and without any fancy prompt scaffolding\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\n#multimodal", - "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", + "description": "Claude 3.5 Haiku features offers enhanced capabilities in speed, coding accuracy, and tool use. Engineered to excel in real-time applications, it delivers quick response times that are essential for dynamic tasks such as chat interactions and immediate coding suggestions.\n\nThis makes it highly suitable for environments that demand both speed and precision, such as software development, customer service bots, and data management systems.\n\nThis model is currently pointing to [Claude 3.5 Haiku (2024-10-22)](/anthropic/claude-3-5-haiku-20241022).", + "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", "contextLength": 200000, "inputModalities": [ "text", @@ -16950,12 +18369,12 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3.5-sonnet", + "permaslug": "anthropic/claude-3-5-haiku", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "anthropic/claude-3.5-sonnet:beta", - "modelVariantPermaslug": "anthropic/claude-3.5-sonnet:beta", + "modelVariantSlug": "anthropic/claude-3.5-haiku", + "modelVariantPermaslug": "anthropic/claude-3-5-haiku", "providerName": "Anthropic", "providerInfo": { "name": "Anthropic", @@ -16989,10 +18408,10 @@ export default { } }, "providerDisplayName": "Anthropic", - "providerModelId": "claude-3-5-sonnet-20241022", + "providerModelId": "claude-3-5-haiku-20241022", "providerGroup": "Anthropic", "quantization": "unknown", - "variant": "beta", + "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, @@ -17009,7 +18428,7 @@ export default { "stop" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", @@ -17024,12 +18443,12 @@ export default { "retentionDays": 30 }, "pricing": { - "prompt": "0.000003", - "completion": "0.000015", - "image": "0.0048", + "prompt": "0.0000008", + "completion": "0.000004", + "image": "0", "request": "0", - "inputCacheRead": "0.0000003", - "inputCacheWrite": "0.00000375", + "inputCacheRead": "0.00000008", + "inputCacheWrite": "0.000001", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -17051,28 +18470,30 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 18, - "newest": 183, + "topWeekly": 47, + "newest": 177, "throughputHighToLow": 153, - "latencyLowToHigh": 250, - "pricingLowToHigh": 272, - "pricingHighToLow": 44 + "latencyLowToHigh": 187, + "pricingLowToHigh": 221, + "pricingHighToLow": 93 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", "slug": "anthropic", "quantization": "unknown", "context": 200000, "maxCompletionTokens": 8192, - "providerModelId": "claude-3-5-sonnet-20241022", + "providerModelId": "claude-3-5-haiku-20241022", "pricing": { - "prompt": "0.000003", - "completion": "0.000015", - "image": "0.0048", + "prompt": "0.0000008", + "completion": "0.000004", + "image": "0", "request": "0", - "inputCacheRead": "0.0000003", - "inputCacheWrite": "0.00000375", + "inputCacheRead": "0.00000008", + "inputCacheWrite": "0.000001", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -17086,23 +18507,26 @@ export default { "top_k", "stop" ], - "inputCost": 3, - "outputCost": 15 + "inputCost": 0.8, + "outputCost": 4, + "throughput": 57.0545, + "latency": 1171.5 }, { "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", "slug": "vertex", "quantization": "unknown", "context": 200000, "maxCompletionTokens": 8192, - "providerModelId": "claude-3-5-sonnet-v2@20241022", + "providerModelId": "claude-3-5-haiku@20241022", "pricing": { - "prompt": "0.000003", - "completion": "0.000015", - "image": "0.0048", + "prompt": "0.0000008", + "completion": "0.000004", + "image": "0", "request": "0", - "inputCacheRead": "0.0000003", - "inputCacheWrite": "0.00000375", + "inputCacheRead": "0.00000008", + "inputCacheWrite": "0.000001", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -17116,20 +18540,23 @@ export default { "top_k", "stop" ], - "inputCost": 3, - "outputCost": 15 + "inputCost": 0.8, + "outputCost": 4, + "throughput": 61.668, + "latency": 2440.5 }, { "name": "Amazon Bedrock", + "icon": "https://openrouter.ai/images/icons/Bedrock.svg", "slug": "amazonBedrock", "quantization": "unknown", "context": 200000, "maxCompletionTokens": 8192, - "providerModelId": "us.anthropic.claude-3-5-sonnet-20241022-v2:0", + "providerModelId": "us.anthropic.claude-3-5-haiku-20241022-v1:0", "pricing": { - "prompt": "0.000003", - "completion": "0.000015", - "image": "0.0048", + "prompt": "0.0000008", + "completion": "0.000004", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -17144,53 +18571,26 @@ export default { "top_k", "stop" ], - "inputCost": 3, - "outputCost": 15 + "inputCost": 0.8, + "outputCost": 4, + "throughput": 55.818, + "latency": 1382 }, { - "name": "Amazon Bedrock", - "slug": "amazonBedrock", + "name": "Amazon Bedrock (US-WEST)", + "icon": "", + "slug": "amazonBedrock (usWest)", "quantization": null, "context": 200000, "maxCompletionTokens": 8192, - "providerModelId": "us.anthropic.claude-3-5-sonnet-20241022-v2:0", - "pricing": { - "prompt": "0.000003", - "completion": "0.000015", - "image": "0.0048", - "request": "0", - "inputCacheRead": "0.0000003", - "inputCacheWrite": "0.00000375", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "top_k", - "stop" - ], - "inputCost": 3, - "outputCost": 15 - }, - { - "name": "Google Vertex", - "slug": "vertex", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 8192, - "providerModelId": "claude-3-5-sonnet-v2@20241022", + "providerModelId": "us.anthropic.claude-3-5-haiku-20241022-v1:0", "pricing": { - "prompt": "0.000003", - "completion": "0.000015", - "image": "0.0048", + "prompt": "0.0000008", + "completion": "0.000004", + "image": "0", "request": "0", - "inputCacheRead": "0.0000003", - "inputCacheWrite": "0.00000375", + "inputCacheRead": "0.00000008", + "inputCacheWrite": "0.000001", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -17204,8 +18604,10 @@ export default { "top_k", "stop" ], - "inputCost": 3, - "outputCost": 15 + "inputCost": 0.8, + "outputCost": 4, + "throughput": 55.934, + "latency": 1298.5 } ] }, @@ -17378,16 +18780,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 47, + "topWeekly": 48, "newest": 43, - "throughputHighToLow": 84, - "latencyLowToHigh": 303, + "throughputHighToLow": 91, + "latencyLowToHigh": 295, "pricingLowToHigh": 229, "pricingHighToLow": 84 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", "slug": "openAi", "quantization": null, "context": 200000, @@ -17412,151 +18816,156 @@ export default { "structured_outputs" ], "inputCost": 1.1, - "outputCost": 4.4 + "outputCost": 4.4, + "throughput": 89.1445, + "latency": 5160.5 } ] }, { - "slug": "anthropic/claude-3.5-haiku", - "hfSlug": null, + "slug": "gryphe/mythomax-l2-13b", + "hfSlug": "Gryphe/MythoMax-L2-13b", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-04T00:00:00+00:00", + "createdAt": "2023-07-02T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3.5 Haiku", - "shortName": "Claude 3.5 Haiku", - "author": "anthropic", - "description": "Claude 3.5 Haiku features offers enhanced capabilities in speed, coding accuracy, and tool use. Engineered to excel in real-time applications, it delivers quick response times that are essential for dynamic tasks such as chat interactions and immediate coding suggestions.\n\nThis makes it highly suitable for environments that demand both speed and precision, such as software development, customer service bots, and data management systems.\n\nThis model is currently pointing to [Claude 3.5 Haiku (2024-10-22)](/anthropic/claude-3-5-haiku-20241022).", - "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", - "contextLength": 200000, + "name": "MythoMax 13B", + "shortName": "MythoMax 13B", + "author": "gryphe", + "description": "One of the highest performing and most popular fine-tunes of Llama 2 13B, with rich descriptions and roleplay. #merge", + "modelVersionGroupId": null, + "contextLength": 4096, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", - "instructType": null, + "group": "Llama2", + "instructType": "alpaca", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "###", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-5-haiku", + "permaslug": "gryphe/mythomax-l2-13b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "ced809d4-9715-4634-8b4d-ea0e09d6bb85", - "name": "Anthropic | anthropic/claude-3-5-haiku", - "contextLength": 200000, + "id": "ffd94635-42cb-47e4-988a-b905c2e7fa57", + "name": "DeepInfra | gryphe/mythomax-l2-13b", + "contextLength": 4096, "model": { - "slug": "anthropic/claude-3.5-haiku", - "hfSlug": null, + "slug": "gryphe/mythomax-l2-13b", + "hfSlug": "Gryphe/MythoMax-L2-13b", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-04T00:00:00+00:00", + "createdAt": "2023-07-02T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3.5 Haiku", - "shortName": "Claude 3.5 Haiku", - "author": "anthropic", - "description": "Claude 3.5 Haiku features offers enhanced capabilities in speed, coding accuracy, and tool use. Engineered to excel in real-time applications, it delivers quick response times that are essential for dynamic tasks such as chat interactions and immediate coding suggestions.\n\nThis makes it highly suitable for environments that demand both speed and precision, such as software development, customer service bots, and data management systems.\n\nThis model is currently pointing to [Claude 3.5 Haiku (2024-10-22)](/anthropic/claude-3-5-haiku-20241022).", - "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", - "contextLength": 200000, + "name": "MythoMax 13B", + "shortName": "MythoMax 13B", + "author": "gryphe", + "description": "One of the highest performing and most popular fine-tunes of Llama 2 13B, with rich descriptions and roleplay. #merge", + "modelVersionGroupId": null, + "contextLength": 4096, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", - "instructType": null, + "group": "Llama2", + "instructType": "alpaca", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "###", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-5-haiku", + "permaslug": "gryphe/mythomax-l2-13b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "anthropic/claude-3.5-haiku", - "modelVariantPermaslug": "anthropic/claude-3-5-haiku", - "providerName": "Anthropic", + "modelVariantSlug": "gryphe/mythomax-l2-13b", + "modelVariantPermaslug": "gryphe/mythomax-l2-13b", + "providerName": "DeepInfra", "providerInfo": { - "name": "Anthropic", - "displayName": "Anthropic", - "slug": "anthropic", + "name": "DeepInfra", + "displayName": "DeepInfra", + "slug": "deepinfra", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "retainsPrompts": false + } }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "Anthropic", + "moderationRequired": false, + "group": "DeepInfra", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.anthropic.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Anthropic.svg" + "url": "/images/icons/DeepInfra.webp" } }, - "providerDisplayName": "Anthropic", - "providerModelId": "claude-3-5-haiku-20241022", - "providerGroup": "Anthropic", - "quantization": "unknown", + "providerDisplayName": "DeepInfra", + "providerModelId": "Gryphe/MythoMax-L2-13b", + "providerGroup": "DeepInfra", + "quantization": "fp16", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 8192, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", "top_k", - "stop" + "seed", + "min_p" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, - "requiresUserIds": true, "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, "pricing": { - "prompt": "0.0000008", - "completion": "0.000004", + "prompt": "0.000000065", + "completion": "0.000000065", "image": "0", "request": "0", - "inputCacheRead": "0.00000008", - "inputCacheWrite": "0.000001", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -17565,7 +18974,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -17578,84 +18987,96 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 48, - "newest": 179, - "throughputHighToLow": 129, - "latencyLowToHigh": 164, - "pricingLowToHigh": 221, - "pricingHighToLow": 93 + "topWeekly": 49, + "newest": 313, + "throughputHighToLow": 283, + "latencyLowToHigh": 58, + "pricingLowToHigh": 101, + "pricingHighToLow": 217 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { - "name": "Anthropic", - "slug": "anthropic", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 8192, - "providerModelId": "claude-3-5-haiku-20241022", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "fp16", + "context": 4096, + "maxCompletionTokens": 4096, + "providerModelId": "Gryphe/MythoMax-L2-13b", "pricing": { - "prompt": "0.0000008", - "completion": "0.000004", + "prompt": "0.000000065", + "completion": "0.000000065", "image": "0", "request": "0", - "inputCacheRead": "0.00000008", - "inputCacheWrite": "0.000001", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", "top_k", - "stop" + "seed", + "min_p" ], - "inputCost": 0.8, - "outputCost": 4 + "inputCost": 0.07, + "outputCost": 0.07, + "throughput": 17.7385, + "latency": 477 }, { - "name": "Google Vertex", - "slug": "vertex", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 8192, - "providerModelId": "claude-3-5-haiku@20241022", + "context": 4096, + "maxCompletionTokens": null, + "providerModelId": "gryphe/mythomax-l2-13b", "pricing": { - "prompt": "0.0000008", - "completion": "0.000004", + "prompt": "0.00000009", + "completion": "0.00000009", "image": "0", "request": "0", - "inputCacheRead": "0.00000008", - "inputCacheWrite": "0.000001", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", "top_k", - "stop" + "min_p", + "repetition_penalty", + "logit_bias" ], - "inputCost": 0.8, - "outputCost": 4 + "inputCost": 0.09, + "outputCost": 0.09, + "throughput": 82.272, + "latency": 1074 }, { - "name": "Amazon Bedrock", - "slug": "amazonBedrock", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 8192, - "providerModelId": "us.anthropic.claude-3-5-haiku-20241022-v1:0", + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", + "quantization": "fp16", + "context": 4096, + "maxCompletionTokens": 4096, + "providerModelId": "parasail-mythomax-13b", "pricing": { - "prompt": "0.0000008", - "completion": "0.000004", + "prompt": "0.00000011", + "completion": "0.00000011", "image": "0", "request": "0", "webSearch": "0", @@ -17663,348 +19084,57 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "top_k", - "stop" + "presence_penalty", + "frequency_penalty", + "repetition_penalty", + "top_k" ], - "inputCost": 0.8, - "outputCost": 4 + "inputCost": 0.11, + "outputCost": 0.11, + "throughput": 47.97, + "latency": 546 }, { - "name": "Amazon Bedrock (US-WEST)", - "slug": "amazonBedrock (usWest)", - "quantization": null, - "context": 200000, - "maxCompletionTokens": 8192, - "providerModelId": "us.anthropic.claude-3-5-haiku-20241022-v1:0", + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": "int4", + "context": 4096, + "maxCompletionTokens": null, + "providerModelId": "Gryphe/MythoMax-L2-13b-Lite", "pricing": { - "prompt": "0.0000008", - "completion": "0.000004", + "prompt": "0.0000002", + "completion": "0.0000004", "image": "0", "request": "0", - "inputCacheRead": "0.00000008", - "inputCacheWrite": "0.000001", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "top_k", - "stop" - ], - "inputCost": 0.8, - "outputCost": 4 - } - ] - }, - { - "slug": "gryphe/mythomax-l2-13b", - "hfSlug": "Gryphe/MythoMax-L2-13b", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-07-02T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "MythoMax 13B", - "shortName": "MythoMax 13B", - "author": "gryphe", - "description": "One of the highest performing and most popular fine-tunes of Llama 2 13B, with rich descriptions and roleplay. #merge", - "modelVersionGroupId": null, - "contextLength": 4096, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Llama2", - "instructType": "alpaca", - "defaultSystem": null, - "defaultStops": [ - "###", - "" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "gryphe/mythomax-l2-13b", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "ffd94635-42cb-47e4-988a-b905c2e7fa57", - "name": "DeepInfra | gryphe/mythomax-l2-13b", - "contextLength": 4096, - "model": { - "slug": "gryphe/mythomax-l2-13b", - "hfSlug": "Gryphe/MythoMax-L2-13b", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-07-02T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "MythoMax 13B", - "shortName": "MythoMax 13B", - "author": "gryphe", - "description": "One of the highest performing and most popular fine-tunes of Llama 2 13B, with rich descriptions and roleplay. #merge", - "modelVersionGroupId": null, - "contextLength": 4096, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Llama2", - "instructType": "alpaca", - "defaultSystem": null, - "defaultStops": [ - "###", - "" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "gryphe/mythomax-l2-13b", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "gryphe/mythomax-l2-13b", - "modelVariantPermaslug": "gryphe/mythomax-l2-13b", - "providerName": "DeepInfra", - "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", - "baseUrl": "url", - "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", - "paidModels": { - "training": false, - "retainsPrompts": false - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "DeepInfra", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/DeepInfra.webp" - } - }, - "providerDisplayName": "DeepInfra", - "providerModelId": "Gryphe/MythoMax-L2-13b", - "providerGroup": "DeepInfra", - "quantization": "fp16", - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 4096, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", - "paidModels": { - "training": false, - "retainsPrompts": false - }, - "training": false, - "retainsPrompts": false - }, - "pricing": { - "prompt": "0.000000065", - "completion": "0.000000065", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 49, - "newest": 313, - "throughputHighToLow": 284, - "latencyLowToHigh": 71, - "pricingLowToHigh": 101, - "pricingHighToLow": 217 - }, - "providers": [ - { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "fp16", - "context": 4096, - "maxCompletionTokens": 4096, - "providerModelId": "Gryphe/MythoMax-L2-13b", - "pricing": { - "prompt": "0.000000065", - "completion": "0.000000065", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "inputCost": 0.07, - "outputCost": 0.07 - }, - { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": "unknown", - "context": 4096, - "maxCompletionTokens": null, - "providerModelId": "gryphe/mythomax-l2-13b", - "pricing": { - "prompt": "0.00000009", - "completion": "0.00000009", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" - ], - "inputCost": 0.09, - "outputCost": 0.09 - }, - { - "name": "Parasail", - "slug": "parasail", - "quantization": "fp16", - "context": 4096, - "maxCompletionTokens": 4096, - "providerModelId": "parasail-mythomax-13b", - "pricing": { - "prompt": "0.00000011", - "completion": "0.00000011", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "presence_penalty", - "frequency_penalty", - "repetition_penalty", - "top_k" - ], - "inputCost": 0.11, - "outputCost": 0.11 - }, - { - "name": "Together", - "slug": "together", - "quantization": "int4", - "context": 4096, - "maxCompletionTokens": null, - "providerModelId": "Gryphe/MythoMax-L2-13b-Lite", - "pricing": { - "prompt": "0.0000002", - "completion": "0.0000004", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" ], "inputCost": 0.2, - "outputCost": 0.4 + "outputCost": 0.4, + "throughput": 110.497, + "latency": 523.5 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "unknown", "context": 4096, @@ -18033,10 +19163,13 @@ export default { "response_format" ], "inputCost": 0.3, - "outputCost": 0.3 + "outputCost": 0.3, + "throughput": 135.3875, + "latency": 418 }, { "name": "Mancer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", "slug": "mancer", "quantization": "unknown", "context": 8192, @@ -18044,7 +19177,7 @@ export default { "providerModelId": "mythomax", "pricing": { "prompt": "0.0000005625", - "completion": "0.000001125", + "completion": "0.00000084375", "image": "0", "request": "0", "webSearch": "0", @@ -18066,17 +19199,20 @@ export default { "top_a" ], "inputCost": 0.56, - "outputCost": 1.13 + "outputCost": 0.84, + "throughput": 42.159, + "latency": 933 }, { "name": "Mancer (private)", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", "slug": "mancer (private)", "quantization": "unknown", "context": 8192, "maxCompletionTokens": 1024, "providerModelId": "mythomax", "pricing": { - "prompt": "0.00000075", + "prompt": "0.000001", "completion": "0.0000015", "image": "0", "request": "0", @@ -18098,11 +19234,204 @@ export default { "seed", "top_a" ], - "inputCost": 0.75, - "outputCost": 1.5 + "inputCost": 1, + "outputCost": 1.5, + "throughput": 43.6625, + "latency": 807 } ] }, + { + "slug": "microsoft/mai-ds-r1", + "hfSlug": "microsoft/MAI-DS-R1", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-04-21T00:08:20.257185+00:00", + "hfUpdatedAt": null, + "name": "Microsoft: MAI DS R1 (free)", + "shortName": "MAI DS R1 (free)", + "author": "microsoft", + "description": "MAI-DS-R1 is a post-trained variant of DeepSeek-R1 developed by the Microsoft AI team to improve the model’s responsiveness on previously blocked topics while enhancing its safety profile. Built on top of DeepSeek-R1’s reasoning foundation, it integrates 110k examples from the Tulu-3 SFT dataset and 350k internally curated multilingual safety-alignment samples. The model retains strong reasoning, coding, and problem-solving capabilities, while unblocking a wide range of prompts previously restricted in R1.\n\nMAI-DS-R1 demonstrates improved performance on harm mitigation benchmarks and maintains competitive results across general reasoning tasks. It surpasses R1-1776 in satisfaction metrics for blocked queries and reduces leakage in harmful content categories. The model is based on a transformer MoE architecture and is suitable for general-purpose use cases, excluding high-stakes domains such as legal, medical, or autonomous systems.", + "modelVersionGroupId": null, + "contextLength": 163840, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "DeepSeek", + "instructType": "deepseek-r1", + "defaultSystem": null, + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "microsoft/mai-ds-r1", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, + "endpoint": { + "id": "0cab7863-8f43-47de-9cc6-c3bbf3354c9a", + "name": "Chutes | microsoft/mai-ds-r1:free", + "contextLength": 163840, + "model": { + "slug": "microsoft/mai-ds-r1", + "hfSlug": "microsoft/MAI-DS-R1", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-04-21T00:08:20.257185+00:00", + "hfUpdatedAt": null, + "name": "Microsoft: MAI DS R1", + "shortName": "MAI DS R1", + "author": "microsoft", + "description": "MAI-DS-R1 is a post-trained variant of DeepSeek-R1 developed by the Microsoft AI team to improve the model’s responsiveness on previously blocked topics while enhancing its safety profile. Built on top of DeepSeek-R1’s reasoning foundation, it integrates 110k examples from the Tulu-3 SFT dataset and 350k internally curated multilingual safety-alignment samples. The model retains strong reasoning, coding, and problem-solving capabilities, while unblocking a wide range of prompts previously restricted in R1.\n\nMAI-DS-R1 demonstrates improved performance on harm mitigation benchmarks and maintains competitive results across general reasoning tasks. It surpasses R1-1776 in satisfaction metrics for blocked queries and reduces leakage in harmful content categories. The model is based on a transformer MoE architecture and is suitable for general-purpose use cases, excluding high-stakes domains such as legal, medical, or autonomous systems.", + "modelVersionGroupId": null, + "contextLength": 163840, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "DeepSeek", + "instructType": "deepseek-r1", + "defaultSystem": null, + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "microsoft/mai-ds-r1", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } + }, + "modelVariantSlug": "microsoft/mai-ds-r1:free", + "modelVariantPermaslug": "microsoft/mai-ds-r1:free", + "providerName": "Chutes", + "providerInfo": { + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://chutes.ai/tos", + "paidModels": { + "training": true, + "retainsPrompts": true + } + }, + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "Chutes", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + } + }, + "providerDisplayName": "Chutes", + "providerModelId": "microsoft/MAI-DS-R1-FP8", + "providerGroup": "Chutes", + "quantization": "fp8", + "variant": "free", + "isFree": true, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://chutes.ai/tos", + "paidModels": { + "training": true, + "retainsPrompts": true + }, + "training": true, + "retainsPrompts": true + }, + "pricing": { + "prompt": "0", + "completion": "0", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": true, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 50, + "newest": 36, + "throughputHighToLow": 196, + "latencyLowToHigh": 223, + "pricingLowToHigh": 17, + "pricingHighToLow": 265 + }, + "authorIcon": "https://openrouter.ai/images/icons/Microsoft.svg", + "providers": [] + }, { "slug": "nousresearch/hermes-3-llama-3.1-405b", "hfSlug": "NousResearch/Hermes-3-Llama-3.1-405B", @@ -18265,16 +19594,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 50, + "topWeekly": 51, "newest": 219, - "throughputHighToLow": 232, - "latencyLowToHigh": 283, + "throughputHighToLow": 239, + "latencyLowToHigh": 163, "pricingLowToHigh": 200, "pricingHighToLow": 116 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nousresearch.com/\\u0026size=256", "providers": [ { "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", "slug": "lambda", "quantization": "fp8", "context": 131072, @@ -18303,10 +19634,13 @@ export default { "response_format" ], "inputCost": 0.8, - "outputCost": 0.8 + "outputCost": 0.8, + "throughput": 32.7495, + "latency": 1412 }, { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", "context": 131072, @@ -18335,10 +19669,13 @@ export default { "min_p" ], "inputCost": 0.8, - "outputCost": 0.8 + "outputCost": 0.8, + "throughput": 11.712, + "latency": 1277 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 131072, @@ -18367,458 +19704,33 @@ export default { "top_logprobs" ], "inputCost": 1, - "outputCost": 3 + "outputCost": 3, + "throughput": 32.051, + "latency": 682 } ] }, { - "slug": "openai/gpt-4o-mini-2024-07-18", - "hfSlug": null, - "updatedAt": "2025-04-23T22:07:14.802248+00:00", - "createdAt": "2024-07-18T00:00:00+00:00", + "slug": "qwen/qwen3-235b-a22b", + "hfSlug": "Qwen/Qwen3-235B-A22B", + "updatedAt": "2025-05-11T22:44:42.939095+00:00", + "createdAt": "2025-04-28T21:29:17.25671+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4o-mini (2024-07-18)", - "shortName": "GPT-4o-mini (2024-07-18)", - "author": "openai", - "description": "GPT-4o mini is OpenAI's newest model after [GPT-4 Omni](/models/openai/gpt-4o), supporting both text and image inputs with text outputs.\n\nAs their most advanced small model, it is many multiples more affordable than other recent frontier models, and more than 60% cheaper than [GPT-3.5 Turbo](/models/openai/gpt-3.5-turbo). It maintains SOTA intelligence, while being significantly more cost-effective.\n\nGPT-4o mini achieves an 82% score on MMLU and presently ranks higher than GPT-4 on chat preferences [common leaderboards](https://arena.lmsys.org/).\n\nCheck out the [launch announcement](https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/) to learn more.\n\n#multimodal", + "name": "Qwen: Qwen3 235B A22B (free)", + "shortName": "Qwen3 235B A22B (free)", + "author": "qwen", + "description": "Qwen3-235B-A22B is a 235B parameter mixture-of-experts (MoE) model developed by Qwen, activating 22B parameters per forward pass. It supports seamless switching between a \"thinking\" mode for complex reasoning, math, and code tasks, and a \"non-thinking\" mode for general conversational efficiency. The model demonstrates strong reasoning ability, multilingual support (100+ languages and dialects), advanced instruction-following, and agent tool-calling capabilities. It natively handles a 32K token context window and extends up to 131K tokens using YaRN-based scaling.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 40960, "inputModalities": [ - "text", - "image", - "file" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "openai/gpt-4o-mini-2024-07-18", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "ebcc1f0a-6621-4cdc-a93f-88a6e2cc2e15", - "name": "OpenAI | openai/gpt-4o-mini-2024-07-18", - "contextLength": 128000, - "model": { - "slug": "openai/gpt-4o-mini-2024-07-18", - "hfSlug": null, - "updatedAt": "2025-04-23T22:07:14.802248+00:00", - "createdAt": "2024-07-18T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "OpenAI: GPT-4o-mini (2024-07-18)", - "shortName": "GPT-4o-mini (2024-07-18)", - "author": "openai", - "description": "GPT-4o mini is OpenAI's newest model after [GPT-4 Omni](/models/openai/gpt-4o), supporting both text and image inputs with text outputs.\n\nAs their most advanced small model, it is many multiples more affordable than other recent frontier models, and more than 60% cheaper than [GPT-3.5 Turbo](/models/openai/gpt-3.5-turbo). It maintains SOTA intelligence, while being significantly more cost-effective.\n\nGPT-4o mini achieves an 82% score on MMLU and presently ranks higher than GPT-4 on chat preferences [common leaderboards](https://arena.lmsys.org/).\n\nCheck out the [launch announcement](https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/) to learn more.\n\n#multimodal", - "modelVersionGroupId": null, - "contextLength": 128000, - "inputModalities": [ - "text", - "image", - "file" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "GPT", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "openai/gpt-4o-mini-2024-07-18", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "openai/gpt-4o-mini-2024-07-18", - "modelVariantPermaslug": "openai/gpt-4o-mini-2024-07-18", - "providerName": "OpenAI", - "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true - } - }, - "providerDisplayName": "OpenAI", - "providerModelId": "gpt-4o-mini-2024-07-18", - "providerGroup": "OpenAI", - "quantization": "unknown", - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 16384, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "web_search_options", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" - ], - "isByok": false, - "moderationRequired": true, - "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "pricing": { - "prompt": "0.00000015", - "completion": "0.0000006", - "image": "0.007225", - "request": "0", - "inputCacheRead": "0.000000075", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [ - { - "type": "search-threshold", - "threshold": "high", - "request": "0.03" - }, - { - "type": "search-threshold", - "threshold": "medium", - "request": "0.0275" - }, - { - "type": "search-threshold", - "threshold": "low", - "request": "0.025" - } - ], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 51, - "newest": 236, - "throughputHighToLow": 137, - "latencyLowToHigh": 58, - "pricingLowToHigh": 146, - "pricingHighToLow": 174 - }, - "providers": [ - { - "name": "OpenAI", - "slug": "openAi", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 16384, - "providerModelId": "gpt-4o-mini-2024-07-18", - "pricing": { - "prompt": "0.00000015", - "completion": "0.0000006", - "image": "0.007225", - "request": "0", - "inputCacheRead": "0.000000075", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "web_search_options", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" - ], - "inputCost": 0.15, - "outputCost": 0.6 - } - ] - }, - { - "slug": "microsoft/mai-ds-r1", - "hfSlug": "microsoft/MAI-DS-R1", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-04-21T00:08:20.257185+00:00", - "hfUpdatedAt": null, - "name": "Microsoft: MAI DS R1 (free)", - "shortName": "MAI DS R1 (free)", - "author": "microsoft", - "description": "MAI-DS-R1 is a post-trained variant of DeepSeek-R1 developed by the Microsoft AI team to improve the model’s responsiveness on previously blocked topics while enhancing its safety profile. Built on top of DeepSeek-R1’s reasoning foundation, it integrates 110k examples from the Tulu-3 SFT dataset and 350k internally curated multilingual safety-alignment samples. The model retains strong reasoning, coding, and problem-solving capabilities, while unblocking a wide range of prompts previously restricted in R1.\n\nMAI-DS-R1 demonstrates improved performance on harm mitigation benchmarks and maintains competitive results across general reasoning tasks. It surpasses R1-1776 in satisfaction metrics for blocked queries and reduces leakage in harmful content categories. The model is based on a transformer MoE architecture and is suitable for general-purpose use cases, excluding high-stakes domains such as legal, medical, or autonomous systems.", - "modelVersionGroupId": null, - "contextLength": 163840, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "DeepSeek", - "instructType": "deepseek-r1", - "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "microsoft/mai-ds-r1", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, - "endpoint": { - "id": "0cab7863-8f43-47de-9cc6-c3bbf3354c9a", - "name": "Chutes | microsoft/mai-ds-r1:free", - "contextLength": 163840, - "model": { - "slug": "microsoft/mai-ds-r1", - "hfSlug": "microsoft/MAI-DS-R1", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-04-21T00:08:20.257185+00:00", - "hfUpdatedAt": null, - "name": "Microsoft: MAI DS R1", - "shortName": "MAI DS R1", - "author": "microsoft", - "description": "MAI-DS-R1 is a post-trained variant of DeepSeek-R1 developed by the Microsoft AI team to improve the model’s responsiveness on previously blocked topics while enhancing its safety profile. Built on top of DeepSeek-R1’s reasoning foundation, it integrates 110k examples from the Tulu-3 SFT dataset and 350k internally curated multilingual safety-alignment samples. The model retains strong reasoning, coding, and problem-solving capabilities, while unblocking a wide range of prompts previously restricted in R1.\n\nMAI-DS-R1 demonstrates improved performance on harm mitigation benchmarks and maintains competitive results across general reasoning tasks. It surpasses R1-1776 in satisfaction metrics for blocked queries and reduces leakage in harmful content categories. The model is based on a transformer MoE architecture and is suitable for general-purpose use cases, excluding high-stakes domains such as legal, medical, or autonomous systems.", - "modelVersionGroupId": null, - "contextLength": 163840, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "DeepSeek", - "instructType": "deepseek-r1", - "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "microsoft/mai-ds-r1", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } - }, - "modelVariantSlug": "microsoft/mai-ds-r1:free", - "modelVariantPermaslug": "microsoft/mai-ds-r1:free", - "providerName": "Chutes", - "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", - "paidModels": { - "training": true, - "retainsPrompts": true - } - }, - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "Chutes", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" - } - }, - "providerDisplayName": "Chutes", - "providerModelId": "microsoft/MAI-DS-R1-FP8", - "providerGroup": "Chutes", - "quantization": "fp8", - "variant": "free", - "isFree": true, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", - "paidModels": { - "training": true, - "retainsPrompts": true - }, - "training": true, - "retainsPrompts": true - }, - "pricing": { - "prompt": "0", - "completion": "0", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": true, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 52, - "newest": 36, - "throughputHighToLow": 183, - "latencyLowToHigh": 201, - "pricingLowToHigh": 17, - "pricingHighToLow": 265 - }, - "providers": [] - }, - { - "slug": "qwen/qwen3-235b-a22b", - "hfSlug": "Qwen/Qwen3-235B-A22B", - "updatedAt": "2025-05-11T22:44:42.939095+00:00", - "createdAt": "2025-04-28T21:29:17.25671+00:00", - "hfUpdatedAt": null, - "name": "Qwen: Qwen3 235B A22B (free)", - "shortName": "Qwen3 235B A22B (free)", - "author": "qwen", - "description": "Qwen3-235B-A22B is a 235B parameter mixture-of-experts (MoE) model developed by Qwen, activating 22B parameters per forward pass. It supports seamless switching between a \"thinking\" mode for complex reasoning, math, and code tasks, and a \"non-thinking\" mode for general conversational efficiency. The model demonstrates strong reasoning ability, multilingual support (100+ languages and dialects), advanced instruction-following, and agent tool-calling capabilities. It natively handles a 32K token context window and extends up to 131K tokens using YaRN-based scaling.", - "modelVersionGroupId": null, - "contextLength": 40960, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, "defaultStops": [ "<|im_start|>", @@ -18979,16 +19891,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 22, + "topWeekly": 24, "newest": 30, - "throughputHighToLow": 238, - "latencyLowToHigh": 192, + "throughputHighToLow": 244, + "latencyLowToHigh": 180, "pricingLowToHigh": 13, - "pricingHighToLow": 175 + "pricingHighToLow": 178 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", "context": 40960, @@ -19019,10 +19933,13 @@ export default { "min_p" ], "inputCost": 0.14, - "outputCost": 0.6 + "outputCost": 0.6, + "throughput": 23.8015, + "latency": 1157 }, { "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", "slug": "klusterAi", "quantization": "fp8", "context": 40960, @@ -19038,16 +19955,32 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", + "top_p", "reasoning", - "include_reasoning" + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", + "seed" ], "inputCost": 0.14, - "outputCost": 2 + "outputCost": 2, + "throughput": 40.982, + "latency": 797.5 }, { "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", "slug": "parasail", "quantization": "fp8", "context": 40960, @@ -19074,10 +20007,13 @@ export default { "top_k" ], "inputCost": 0.18, - "outputCost": 0.85 + "outputCost": 0.85, + "throughput": 55.8185, + "latency": 968 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "fp8", "context": 40960, @@ -19108,10 +20044,13 @@ export default { "response_format" ], "inputCost": 0.2, - "outputCost": 0.6 + "outputCost": 0.6, + "throughput": 30.284, + "latency": 833 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 40960, @@ -19142,10 +20081,13 @@ export default { "top_logprobs" ], "inputCost": 0.2, - "outputCost": 0.6 + "outputCost": 0.6, + "throughput": 32.01, + "latency": 699 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "fp8", "context": 128000, @@ -19178,10 +20120,13 @@ export default { "logit_bias" ], "inputCost": 0.2, - "outputCost": 0.8 + "outputCost": 0.8, + "throughput": 55.828, + "latency": 1290 }, { "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", "slug": "fireworks", "quantization": null, "context": 128000, @@ -19216,10 +20161,13 @@ export default { "top_logprobs" ], "inputCost": 0.22, - "outputCost": 0.88 + "outputCost": 0.88, + "throughput": 53.7815, + "latency": 857.5 }, { "name": "GMICloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://gmicloud.ai/&size=256", "slug": "gmiCloud", "quantization": "fp8", "context": 32768, @@ -19243,22 +20191,24 @@ export default { "seed" ], "inputCost": 0.25, - "outputCost": 1.09 + "outputCost": 1.09, + "throughput": 48.529, + "latency": 934 } ] }, { - "slug": "cohere/command-r-08-2024", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-30T00:00:00+00:00", + "slug": "tngtech/deepseek-r1t-chimera", + "hfSlug": "tngtech/DeepSeek-R1T-Chimera", + "updatedAt": "2025-05-13T18:40:57.812622+00:00", + "createdAt": "2025-04-27T13:34:35.172638+00:00", "hfUpdatedAt": null, - "name": "Cohere: Command R (08-2024)", - "shortName": "Command R (08-2024)", - "author": "cohere", - "description": "command-r-08-2024 is an update of the [Command R](/models/cohere/command-r) with improved performance for multilingual retrieval-augmented generation (RAG) and tool use. More broadly, it is better at math, code and reasoning and is competitive with the previous version of the larger Command R+ model.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", + "name": "TNG: DeepSeek R1T Chimera (free)", + "shortName": "DeepSeek R1T Chimera (free)", + "author": "tngtech", + "description": "DeepSeek-R1T-Chimera is created by merging DeepSeek-R1 and DeepSeek-V3 (0324), combining the reasoning capabilities of R1 with the token efficiency improvements of V3. It is based on a DeepSeek-MoE Transformer architecture and is optimized for general text generation tasks.\n\nThe model merges pretrained weights from both source models to balance performance across reasoning, efficiency, and instruction-following tasks. It is released under the MIT license and intended for research and commercial use.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 163840, "inputModalities": [ "text" ], @@ -19266,32 +20216,40 @@ export default { "text" ], "hasTextOutput": true, - "group": "Cohere", + "group": "DeepSeek", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cohere/command-r-08-2024", - "reasoningConfig": null, - "features": {}, + "permaslug": "tngtech/deepseek-r1t-chimera", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "3d8c4986-1fc9-4ce1-b07d-d379bb75eb36", - "name": "Cohere | cohere/command-r-08-2024", - "contextLength": 128000, + "id": "817eb65c-3217-4a55-9d95-834243ed3e3b", + "name": "Chutes | tngtech/deepseek-r1t-chimera:free", + "contextLength": 163840, "model": { - "slug": "cohere/command-r-08-2024", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-30T00:00:00+00:00", + "slug": "tngtech/deepseek-r1t-chimera", + "hfSlug": "tngtech/DeepSeek-R1T-Chimera", + "updatedAt": "2025-05-13T18:40:57.812622+00:00", + "createdAt": "2025-04-27T13:34:35.172638+00:00", "hfUpdatedAt": null, - "name": "Cohere: Command R (08-2024)", - "shortName": "Command R (08-2024)", - "author": "cohere", - "description": "command-r-08-2024 is an update of the [Command R](/models/cohere/command-r) with improved performance for multilingual retrieval-augmented generation (RAG) and tool use. More broadly, it is better at math, code and reasoning and is competitive with the previous version of the larger Command R+ model.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", + "name": "TNG: DeepSeek R1T Chimera", + "shortName": "DeepSeek R1T Chimera", + "author": "tngtech", + "description": "DeepSeek-R1T-Chimera is created by merging DeepSeek-R1 and DeepSeek-V3 (0324), combining the reasoning capabilities of R1 with the token efficiency improvements of V3. It is based on a DeepSeek-MoE Transformer architecture and is optimized for general text generation tasks.\n\nThe model merges pretrained weights from both source models to balance performance across reasoning, efficiency, and instruction-following tasks. It is released under the MIT license and intended for research and commercial use.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 163840, "inputModalities": [ "text" ], @@ -19299,218 +20257,7 @@ export default { "text" ], "hasTextOutput": true, - "group": "Cohere", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "cohere/command-r-08-2024", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "cohere/command-r-08-2024", - "modelVariantPermaslug": "cohere/command-r-08-2024", - "providerName": "Cohere", - "providerInfo": { - "name": "Cohere", - "displayName": "Cohere", - "slug": "cohere", - "baseUrl": "url", - "dataPolicy": { - "privacyPolicyUrl": "https://cohere.com/privacy", - "termsOfServiceUrl": "https://cohere.com/terms-of-use", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": true, - "moderationRequired": false, - "group": "Cohere", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": "https://status.cohere.ai/", - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/Cohere.png" - } - }, - "providerDisplayName": "Cohere", - "providerModelId": "command-r-08-2024", - "providerGroup": "Cohere", - "quantization": "unknown", - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 4000, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "tools", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "seed", - "response_format", - "structured_outputs" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "privacyPolicyUrl": "https://cohere.com/privacy", - "termsOfServiceUrl": "https://cohere.com/terms-of-use", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "pricing": { - "prompt": "0.00000015", - "completion": "0.0000006", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": false, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 54, - "newest": 212, - "throughputHighToLow": 128, - "latencyLowToHigh": 41, - "pricingLowToHigh": 145, - "pricingHighToLow": 172 - }, - "providers": [ - { - "name": "Cohere", - "slug": "cohere", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 4000, - "providerModelId": "command-r-08-2024", - "pricing": { - "prompt": "0.00000015", - "completion": "0.0000006", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "seed", - "response_format", - "structured_outputs" - ], - "inputCost": 0.15, - "outputCost": 0.6 - } - ] - }, - { - "slug": "tngtech/deepseek-r1t-chimera", - "hfSlug": "tngtech/DeepSeek-R1T-Chimera", - "updatedAt": "2025-05-13T18:40:57.812622+00:00", - "createdAt": "2025-04-27T13:34:35.172638+00:00", - "hfUpdatedAt": null, - "name": "TNG: DeepSeek R1T Chimera (free)", - "shortName": "DeepSeek R1T Chimera (free)", - "author": "tngtech", - "description": "DeepSeek-R1T-Chimera is created by merging DeepSeek-R1 and DeepSeek-V3 (0324), combining the reasoning capabilities of R1 with the token efficiency improvements of V3. It is based on a DeepSeek-MoE Transformer architecture and is optimized for general text generation tasks.\n\nThe model merges pretrained weights from both source models to balance performance across reasoning, efficiency, and instruction-following tasks. It is released under the MIT license and intended for research and commercial use.", - "modelVersionGroupId": null, - "contextLength": 163840, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "DeepSeek", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "tngtech/deepseek-r1t-chimera", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, - "endpoint": { - "id": "817eb65c-3217-4a55-9d95-834243ed3e3b", - "name": "Chutes | tngtech/deepseek-r1t-chimera:free", - "contextLength": 163840, - "model": { - "slug": "tngtech/deepseek-r1t-chimera", - "hfSlug": "tngtech/DeepSeek-R1T-Chimera", - "updatedAt": "2025-05-13T18:40:57.812622+00:00", - "createdAt": "2025-04-27T13:34:35.172638+00:00", - "hfUpdatedAt": null, - "name": "TNG: DeepSeek R1T Chimera", - "shortName": "DeepSeek R1T Chimera", - "author": "tngtech", - "description": "DeepSeek-R1T-Chimera is created by merging DeepSeek-R1 and DeepSeek-V3 (0324), combining the reasoning capabilities of R1 with the token efficiency improvements of V3. It is based on a DeepSeek-MoE Transformer architecture and is optimized for general text generation tasks.\n\nThe model merges pretrained weights from both source models to balance performance across reasoning, efficiency, and instruction-following tasks. It is released under the MIT license and intended for research and commercial use.", - "modelVersionGroupId": null, - "contextLength": 163840, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "DeepSeek", + "group": "DeepSeek", "instructType": null, "defaultSystem": null, "defaultStops": [], @@ -19625,13 +20372,14 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 55, + "topWeekly": 53, "newest": 32, - "throughputHighToLow": 248, - "latencyLowToHigh": 262, + "throughputHighToLow": 233, + "latencyLowToHigh": 263, "pricingLowToHigh": 14, "pricingHighToLow": 262 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -19785,16 +20533,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 56, + "topWeekly": 54, "newest": 124, - "throughputHighToLow": 71, - "latencyLowToHigh": 124, + "throughputHighToLow": 73, + "latencyLowToHigh": 102, "pricingLowToHigh": 100, "pricingHighToLow": 218 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { "name": "Alibaba", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256", "slug": "alibaba", "quantization": null, "context": 1000000, @@ -19820,150 +20570,186 @@ export default { "presence_penalty" ], "inputCost": 0.05, - "outputCost": 0.2 + "outputCost": 0.2, + "throughput": 104.6415, + "latency": 677.5 } ] }, { - "slug": "amazon/nova-pro-v1", - "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-05T22:05:03.587216+00:00", + "slug": "openai/gpt-4o-mini-2024-07-18", + "hfSlug": null, + "updatedAt": "2025-04-23T22:07:14.802248+00:00", + "createdAt": "2024-07-18T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Amazon: Nova Pro 1.0", - "shortName": "Nova Pro 1.0", - "author": "amazon", - "description": "Amazon Nova Pro 1.0 is a capable multimodal model from Amazon focused on providing a combination of accuracy, speed, and cost for a wide range of tasks. As of December 2024, it achieves state-of-the-art performance on key benchmarks including visual question answering (TextVQA) and video understanding (VATEX).\n\nAmazon Nova Pro demonstrates strong capabilities in processing both visual and textual information and at analyzing financial documents.\n\n**NOTE**: Video input is not supported at this time.", + "name": "OpenAI: GPT-4o-mini (2024-07-18)", + "shortName": "GPT-4o-mini (2024-07-18)", + "author": "openai", + "description": "GPT-4o mini is OpenAI's newest model after [GPT-4 Omni](/models/openai/gpt-4o), supporting both text and image inputs with text outputs.\n\nAs their most advanced small model, it is many multiples more affordable than other recent frontier models, and more than 60% cheaper than [GPT-3.5 Turbo](/models/openai/gpt-3.5-turbo). It maintains SOTA intelligence, while being significantly more cost-effective.\n\nGPT-4o mini achieves an 82% score on MMLU and presently ranks higher than GPT-4 on chat preferences [common leaderboards](https://arena.lmsys.org/).\n\nCheck out the [launch announcement](https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/) to learn more.\n\n#multimodal", "modelVersionGroupId": null, - "contextLength": 300000, + "contextLength": 128000, "inputModalities": [ "text", - "image" + "image", + "file" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Nova", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "amazon/nova-pro-v1", + "permaslug": "openai/gpt-4o-mini-2024-07-18", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "959381a4-8054-450f-9daf-5fcab64ba9aa", - "name": "Amazon Bedrock | amazon/nova-pro-v1", - "contextLength": 300000, + "id": "ebcc1f0a-6621-4cdc-a93f-88a6e2cc2e15", + "name": "OpenAI | openai/gpt-4o-mini-2024-07-18", + "contextLength": 128000, "model": { - "slug": "amazon/nova-pro-v1", - "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-05T22:05:03.587216+00:00", + "slug": "openai/gpt-4o-mini-2024-07-18", + "hfSlug": null, + "updatedAt": "2025-04-23T22:07:14.802248+00:00", + "createdAt": "2024-07-18T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Amazon: Nova Pro 1.0", - "shortName": "Nova Pro 1.0", - "author": "amazon", - "description": "Amazon Nova Pro 1.0 is a capable multimodal model from Amazon focused on providing a combination of accuracy, speed, and cost for a wide range of tasks. As of December 2024, it achieves state-of-the-art performance on key benchmarks including visual question answering (TextVQA) and video understanding (VATEX).\n\nAmazon Nova Pro demonstrates strong capabilities in processing both visual and textual information and at analyzing financial documents.\n\n**NOTE**: Video input is not supported at this time.", + "name": "OpenAI: GPT-4o-mini (2024-07-18)", + "shortName": "GPT-4o-mini (2024-07-18)", + "author": "openai", + "description": "GPT-4o mini is OpenAI's newest model after [GPT-4 Omni](/models/openai/gpt-4o), supporting both text and image inputs with text outputs.\n\nAs their most advanced small model, it is many multiples more affordable than other recent frontier models, and more than 60% cheaper than [GPT-3.5 Turbo](/models/openai/gpt-3.5-turbo). It maintains SOTA intelligence, while being significantly more cost-effective.\n\nGPT-4o mini achieves an 82% score on MMLU and presently ranks higher than GPT-4 on chat preferences [common leaderboards](https://arena.lmsys.org/).\n\nCheck out the [launch announcement](https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/) to learn more.\n\n#multimodal", "modelVersionGroupId": null, - "contextLength": 300000, + "contextLength": 128000, "inputModalities": [ "text", - "image" + "image", + "file" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Nova", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "amazon/nova-pro-v1", + "permaslug": "openai/gpt-4o-mini-2024-07-18", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "amazon/nova-pro-v1", - "modelVariantPermaslug": "amazon/nova-pro-v1", - "providerName": "Amazon Bedrock", + "modelVariantSlug": "openai/gpt-4o-mini-2024-07-18", + "modelVariantPermaslug": "openai/gpt-4o-mini-2024-07-18", + "providerName": "OpenAI", "providerInfo": { - "name": "Amazon Bedrock", - "displayName": "Amazon Bedrock", - "slug": "amazon-bedrock", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://aws.amazon.com/service-terms/", - "privacyPolicyUrl": "https://aws.amazon.com/privacy", - "dataPolicyUrl": "https://docs.aws.amazon.com/bedrock/latest/userguide/data-protection.html", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false - } + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, + "hasCompletions": true, + "isAbortable": true, "moderationRequired": true, - "group": "Amazon Bedrock", + "group": "OpenAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Bedrock.svg" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Amazon Bedrock", - "providerModelId": "us.amazon.nova-pro-v1:0", - "providerGroup": "Amazon Bedrock", - "quantization": null, + "providerDisplayName": "OpenAI", + "providerModelId": "gpt-4o-mini-2024-07-18", + "providerGroup": "OpenAI", + "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 5120, + "maxCompletionTokens": 16384, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "top_k", - "stop" + "stop", + "frequency_penalty", + "presence_penalty", + "web_search_options", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" ], "isByok": false, "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://aws.amazon.com/service-terms/", - "privacyPolicyUrl": "https://aws.amazon.com/privacy", - "dataPolicyUrl": "https://docs.aws.amazon.com/bedrock/latest/userguide/data-protection.html", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, + "requiresUserIds": true, "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.0000008", - "completion": "0.0000032", - "image": "0.0012", + "prompt": "0.00000015", + "completion": "0.0000006", + "image": "0.007225", "request": "0", + "inputCacheRead": "0.000000075", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [], + "variablePricings": [ + { + "type": "search-threshold", + "threshold": "high", + "request": "0.03" + }, + { + "type": "search-threshold", + "threshold": "medium", + "request": "0.0275" + }, + { + "type": "search-threshold", + "threshold": "low", + "request": "0.025" + } + ], "isHidden": false, "isDeranked": false, "isDisabled": false, @@ -19972,7 +20758,7 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -19980,184 +20766,196 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 57, - "newest": 161, - "throughputHighToLow": 51, - "latencyLowToHigh": 108, - "pricingLowToHigh": 218, - "pricingHighToLow": 101 + "topWeekly": 55, + "newest": 237, + "throughputHighToLow": 120, + "latencyLowToHigh": 53, + "pricingLowToHigh": 143, + "pricingHighToLow": 177 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "Amazon Bedrock", - "slug": "amazonBedrock", - "quantization": null, - "context": 300000, - "maxCompletionTokens": 5120, - "providerModelId": "us.amazon.nova-pro-v1:0", + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": "unknown", + "context": 128000, + "maxCompletionTokens": 16384, + "providerModelId": "gpt-4o-mini-2024-07-18", "pricing": { - "prompt": "0.0000008", - "completion": "0.0000032", - "image": "0.0012", + "prompt": "0.00000015", + "completion": "0.0000006", + "image": "0.007225", "request": "0", + "inputCacheRead": "0.000000075", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "top_k", - "stop" + "stop", + "frequency_penalty", + "presence_penalty", + "web_search_options", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" ], - "inputCost": 0.8, - "outputCost": 3.2 + "inputCost": 0.15, + "outputCost": 0.6, + "throughput": 64.408, + "latency": 435 } ] }, { - "slug": "anthropic/claude-3-haiku", + "slug": "cohere/command-r-08-2024", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-03-13T00:00:00+00:00", + "createdAt": "2024-08-30T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3 Haiku", - "shortName": "Claude 3 Haiku", - "author": "anthropic", - "description": "Claude 3 Haiku is Anthropic's fastest and most compact model for\nnear-instant responsiveness. Quick and accurate targeted performance.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-haiku)\n\n#multimodal", - "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", - "contextLength": 200000, + "name": "Cohere: Command R (08-2024)", + "shortName": "Command R (08-2024)", + "author": "cohere", + "description": "command-r-08-2024 is an update of the [Command R](/models/cohere/command-r) with improved performance for multilingual retrieval-augmented generation (RAG) and tool use. More broadly, it is better at math, code and reasoning and is competitive with the previous version of the larger Command R+ model.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", + "modelVersionGroupId": null, + "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", + "group": "Cohere", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-haiku", + "permaslug": "cohere/command-r-08-2024", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "8661a1db-b0cf-4eb2-ba04-c2a79f698682", - "name": "Anthropic | anthropic/claude-3-haiku", - "contextLength": 200000, + "id": "3d8c4986-1fc9-4ce1-b07d-d379bb75eb36", + "name": "Cohere | cohere/command-r-08-2024", + "contextLength": 128000, "model": { - "slug": "anthropic/claude-3-haiku", + "slug": "cohere/command-r-08-2024", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-03-13T00:00:00+00:00", + "createdAt": "2024-08-30T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3 Haiku", - "shortName": "Claude 3 Haiku", - "author": "anthropic", - "description": "Claude 3 Haiku is Anthropic's fastest and most compact model for\nnear-instant responsiveness. Quick and accurate targeted performance.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-haiku)\n\n#multimodal", - "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", - "contextLength": 200000, + "name": "Cohere: Command R (08-2024)", + "shortName": "Command R (08-2024)", + "author": "cohere", + "description": "command-r-08-2024 is an update of the [Command R](/models/cohere/command-r) with improved performance for multilingual retrieval-augmented generation (RAG) and tool use. More broadly, it is better at math, code and reasoning and is competitive with the previous version of the larger Command R+ model.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", + "modelVersionGroupId": null, + "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", + "group": "Cohere", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-haiku", + "permaslug": "cohere/command-r-08-2024", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "anthropic/claude-3-haiku", - "modelVariantPermaslug": "anthropic/claude-3-haiku", - "providerName": "Anthropic", + "modelVariantSlug": "cohere/command-r-08-2024", + "modelVariantPermaslug": "cohere/command-r-08-2024", + "providerName": "Cohere", "providerInfo": { - "name": "Anthropic", - "displayName": "Anthropic", - "slug": "anthropic", + "name": "Cohere", + "displayName": "Cohere", + "slug": "cohere", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "privacyPolicyUrl": "https://cohere.com/privacy", + "termsOfServiceUrl": "https://cohere.com/terms-of-use", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 - }, - "requiresUserIds": true + } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, + "hasCompletions": false, "isAbortable": true, - "moderationRequired": true, - "group": "Anthropic", + "moderationRequired": false, + "group": "Cohere", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.anthropic.com/", + "statusPageUrl": "https://status.cohere.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Anthropic.svg" + "url": "/images/icons/Cohere.png" } }, - "providerDisplayName": "Anthropic", - "providerModelId": "claude-3-haiku-20240307", - "providerGroup": "Anthropic", + "providerDisplayName": "Cohere", + "providerModelId": "command-r-08-2024", + "providerGroup": "Cohere", "quantization": "unknown", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": 4000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "top_k", - "stop" + "seed", + "response_format", + "structured_outputs" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "privacyPolicyUrl": "https://cohere.com/privacy", + "termsOfServiceUrl": "https://cohere.com/terms-of-use", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 }, - "requiresUserIds": true, "training": false, "retainsPrompts": true, "retentionDays": 30 }, "pricing": { - "prompt": "0.00000025", - "completion": "0.00000125", - "image": "0.0004", + "prompt": "0.00000015", + "completion": "0.0000006", + "image": "0", "request": "0", - "inputCacheRead": "0.00000003", - "inputCacheWrite": "0.0000003", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -20171,81 +20969,321 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 58, - "newest": 277, - "throughputHighToLow": 36, - "latencyLowToHigh": 134, - "pricingLowToHigh": 168, - "pricingHighToLow": 149 + "topWeekly": 56, + "newest": 213, + "throughputHighToLow": 127, + "latencyLowToHigh": 38, + "pricingLowToHigh": 142, + "pricingHighToLow": 175 }, + "authorIcon": "https://openrouter.ai/images/icons/Cohere.png", "providers": [ { - "name": "Anthropic", - "slug": "anthropic", + "name": "Cohere", + "icon": "https://openrouter.ai/images/icons/Cohere.png", + "slug": "cohere", "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 4096, - "providerModelId": "claude-3-haiku-20240307", + "context": 128000, + "maxCompletionTokens": 4000, + "providerModelId": "command-r-08-2024", "pricing": { - "prompt": "0.00000025", - "completion": "0.00000125", - "image": "0.0004", + "prompt": "0.00000015", + "completion": "0.0000006", + "image": "0", "request": "0", - "inputCacheRead": "0.00000003", - "inputCacheWrite": "0.0000003", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "top_k", - "stop" + "seed", + "response_format", + "structured_outputs" ], - "inputCost": 0.25, - "outputCost": 1.25 + "inputCost": 0.15, + "outputCost": 0.6, + "throughput": 61.619, + "latency": 359 + } + ] + }, + { + "slug": "google/gemini-pro-1.5", + "hfSlug": null, + "updatedAt": "2025-04-12T05:01:34.756668+00:00", + "createdAt": "2024-04-09T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Google: Gemini 1.5 Pro", + "shortName": "Gemini 1.5 Pro", + "author": "google", + "description": "Google's latest multimodal model, supports image and video[0] in text or chat prompts.\n\nOptimized for language tasks including:\n\n- Code generation\n- Text generation\n- Text editing\n- Problem solving\n- Recommendations\n- Information extraction\n- Data extraction or generation\n- AI agents\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n* [0]: Video input is not available through OpenRouter at this time.", + "modelVersionGroupId": null, + "contextLength": 2000000, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Gemini", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": "", + "permaslug": "google/gemini-pro-1.5", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "089465ba-a451-4730-a9dd-6a2ff05281f7", + "name": "Google | google/gemini-pro-1.5", + "contextLength": 2000000, + "model": { + "slug": "google/gemini-pro-1.5", + "hfSlug": null, + "updatedAt": "2025-04-12T05:01:34.756668+00:00", + "createdAt": "2024-04-09T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Google: Gemini 1.5 Pro", + "shortName": "Gemini 1.5 Pro", + "author": "google", + "description": "Google's latest multimodal model, supports image and video[0] in text or chat prompts.\n\nOptimized for language tasks including:\n\n- Code generation\n- Text generation\n- Text editing\n- Problem solving\n- Recommendations\n- Information extraction\n- Data extraction or generation\n- AI agents\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n* [0]: Video input is not available through OpenRouter at this time.", + "modelVersionGroupId": null, + "contextLength": 2000000, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Gemini", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": "", + "permaslug": "google/gemini-pro-1.5", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "google/gemini-pro-1.5", + "modelVariantPermaslug": "google/gemini-pro-1.5", + "providerName": "Google", + "providerInfo": { + "name": "Google", + "displayName": "Google Vertex", + "slug": "google-vertex", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://cloud.google.com/terms/", + "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", + "dataPolicyUrl": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/abuse-monitoring", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "freeModels": { + "training": true, + "retainsPrompts": true + } + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": false, + "isAbortable": false, + "moderationRequired": false, + "group": "Google", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": "https://status.cloud.google.com/products/sdXM79fz1FS6ekNpu37K/history", + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/GoogleVertex.svg" + } + }, + "providerDisplayName": "Google Vertex", + "providerModelId": "gemini-1.5-pro-002", + "providerGroup": "Google", + "quantization": "unknown", + "variant": "standard", + "isFree": false, + "canAbort": false, + "maxPromptTokens": null, + "maxCompletionTokens": 8192, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "tools", + "tool_choice", + "seed", + "response_format", + "structured_outputs" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://cloud.google.com/terms/", + "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", + "dataPolicyUrl": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/abuse-monitoring", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "freeModels": { + "training": true, + "retainsPrompts": true + }, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "pricing": { + "prompt": "0.00000125", + "completion": "0.000005", + "image": "0.0006575", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [ + { + "type": "prompt-threshold", + "threshold": 128000, + "prompt": "0.0000025", + "completions": "0.00001" + } + ], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": true, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": false, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 57, + "newest": 270, + "throughputHighToLow": 123, + "latencyLowToHigh": 160, + "pricingLowToHigh": 237, + "pricingHighToLow": 81 + }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", + "providers": [ { "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", "slug": "vertex", "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 4096, - "providerModelId": "claude-3-haiku@20240307", + "context": 2000000, + "maxCompletionTokens": 8192, + "providerModelId": "gemini-1.5-pro-002", "pricing": { - "prompt": "0.00000025", - "completion": "0.00000125", - "image": "0.0004", + "prompt": "0.00000125", + "completion": "0.000005", + "image": "0.0006575", "request": "0", - "inputCacheRead": "0.00000003", - "inputCacheWrite": "0.0000003", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "tools", "tool_choice", + "seed", + "response_format", + "structured_outputs" + ], + "inputCost": 1.25, + "outputCost": 5, + "throughput": 66.699, + "latency": 965.5 + }, + { + "name": "Google AI Studio", + "icon": "https://openrouter.ai/images/icons/GoogleAIStudio.svg", + "slug": "google", + "quantization": "unknown", + "context": 2000000, + "maxCompletionTokens": 8192, + "providerModelId": "gemini-1.5-pro-002", + "pricing": { + "prompt": "0.00000125", + "completion": "0.000005", + "image": "0", + "request": "0", + "inputCacheRead": "0.000000625", + "inputCacheWrite": "0.000002875", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ "max_tokens", "temperature", "top_p", - "top_k", - "stop" + "stop", + "frequency_penalty", + "presence_penalty", + "tools", + "tool_choice", + "seed", + "response_format", + "structured_outputs" ], - "inputCost": 0.25, - "outputCost": 1.25 + "inputCost": 1.25, + "outputCost": 5, + "throughput": 61.717, + "latency": 834 } ] }, @@ -20410,16 +21448,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 59, + "topWeekly": 58, "newest": 290, - "throughputHighToLow": 50, - "latencyLowToHigh": 26, + "throughputHighToLow": 46, + "latencyLowToHigh": 25, "pricingLowToHigh": 161, "pricingHighToLow": 158 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", "slug": "mistral", "quantization": "unknown", "context": 32768, @@ -20448,7 +21488,225 @@ export default { "seed" ], "inputCost": 0.25, - "outputCost": 0.25 + "outputCost": 0.25, + "throughput": 135.926, + "latency": 288 + } + ] + }, + { + "slug": "google/gemma-3-4b-it", + "hfSlug": "google/gemma-3-4b-it", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-03-13T22:38:30.653142+00:00", + "hfUpdatedAt": null, + "name": "Google: Gemma 3 4B", + "shortName": "Gemma 3 4B", + "author": "google", + "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling.", + "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", + "contextLength": 131072, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Gemini", + "instructType": "gemma", + "defaultSystem": null, + "defaultStops": [ + "", + "", + "" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "google/gemma-3-4b-it", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "d3de3bd4-81bc-48fb-924f-2a87b2a36e75", + "name": "DeepInfra | google/gemma-3-4b-it", + "contextLength": 131072, + "model": { + "slug": "google/gemma-3-4b-it", + "hfSlug": "google/gemma-3-4b-it", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-03-13T22:38:30.653142+00:00", + "hfUpdatedAt": null, + "name": "Google: Gemma 3 4B", + "shortName": "Gemma 3 4B", + "author": "google", + "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling.", + "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", + "contextLength": 131072, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Gemini", + "instructType": "gemma", + "defaultSystem": null, + "defaultStops": [ + "", + "", + "" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "google/gemma-3-4b-it", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "google/gemma-3-4b-it", + "modelVariantPermaslug": "google/gemma-3-4b-it", + "providerName": "DeepInfra", + "providerInfo": { + "name": "DeepInfra", + "displayName": "DeepInfra", + "slug": "deepinfra", + "baseUrl": "url", + "dataPolicy": { + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", + "paidModels": { + "training": false, + "retainsPrompts": false + } + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "DeepInfra", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/DeepInfra.webp" + } + }, + "providerDisplayName": "DeepInfra", + "providerModelId": "google/gemma-3-4b-it", + "providerGroup": "DeepInfra", + "quantization": "bf16", + "variant": "standard", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", + "paidModels": { + "training": false, + "retainsPrompts": false + }, + "training": false, + "retainsPrompts": false + }, + "pricing": { + "prompt": "0.00000002", + "completion": "0.00000004", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 59, + "newest": 83, + "throughputHighToLow": 70, + "latencyLowToHigh": 26, + "pricingLowToHigh": 38, + "pricingHighToLow": 241 + }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", + "providers": [ + { + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "google/gemma-3-4b-it", + "pricing": { + "prompt": "0.00000002", + "completion": "0.00000004", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" + ], + "inputCost": 0.02, + "outputCost": 0.04, + "throughput": 80.2565, + "latency": 299 } ] }, @@ -20619,14 +21877,16 @@ export default { "sorting": { "topWeekly": 60, "newest": 292, - "throughputHighToLow": 57, - "latencyLowToHigh": 9, - "pricingLowToHigh": 108, - "pricingHighToLow": 211 + "throughputHighToLow": 72, + "latencyLowToHigh": 12, + "pricingLowToHigh": 109, + "pricingHighToLow": 210 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 32768, @@ -20655,10 +21915,13 @@ export default { "top_logprobs" ], "inputCost": 0.08, - "outputCost": 0.24 + "outputCost": 0.24, + "throughput": 111.173, + "latency": 238 }, { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", "context": 32768, @@ -20689,10 +21952,13 @@ export default { "min_p" ], "inputCost": 0.24, - "outputCost": 0.24 + "outputCost": 0.24, + "throughput": 109.594, + "latency": 517 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "unknown", "context": 32768, @@ -20723,22 +21989,230 @@ export default { "response_format" ], "inputCost": 0.6, - "outputCost": 0.6 + "outputCost": 0.6, + "throughput": 62.552, + "latency": 491 } ] }, { - "slug": "google/gemini-pro-1.5", - "hfSlug": null, - "updatedAt": "2025-04-12T05:01:34.756668+00:00", - "createdAt": "2024-04-09T00:00:00+00:00", + "slug": "liquid/lfm-7b", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-01-25T12:08:03.736586+00:00", "hfUpdatedAt": null, - "name": "Google: Gemini 1.5 Pro", - "shortName": "Gemini 1.5 Pro", - "author": "google", - "description": "Google's latest multimodal model, supports image and video[0] in text or chat prompts.\n\nOptimized for language tasks including:\n\n- Code generation\n- Text generation\n- Text editing\n- Problem solving\n- Recommendations\n- Information extraction\n- Data extraction or generation\n- AI agents\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n* [0]: Video input is not available through OpenRouter at this time.", + "name": "Liquid: LFM 7B", + "shortName": "LFM 7B", + "author": "liquid", + "description": "LFM-7B, a new best-in-class language model. LFM-7B is designed for exceptional chat capabilities, including languages like Arabic and Japanese. Powered by the Liquid Foundation Model (LFM) architecture, it exhibits unique features like low memory footprint and fast inference speed. \n\nLFM-7B is the world’s best-in-class multilingual language model in English, Arabic, and Japanese.\n\nSee the [launch announcement](https://www.liquid.ai/lfm-7b) for benchmarks and more info.", "modelVersionGroupId": null, - "contextLength": 2000000, + "contextLength": 32768, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Other", + "instructType": "chatml", + "defaultSystem": null, + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "liquid/lfm-7b", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "25b2cca5-53f6-40e7-b47b-191ec968b7c2", + "name": "Liquid | liquid/lfm-7b", + "contextLength": 32768, + "model": { + "slug": "liquid/lfm-7b", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-01-25T12:08:03.736586+00:00", + "hfUpdatedAt": null, + "name": "Liquid: LFM 7B", + "shortName": "LFM 7B", + "author": "liquid", + "description": "LFM-7B, a new best-in-class language model. LFM-7B is designed for exceptional chat capabilities, including languages like Arabic and Japanese. Powered by the Liquid Foundation Model (LFM) architecture, it exhibits unique features like low memory footprint and fast inference speed. \n\nLFM-7B is the world’s best-in-class multilingual language model in English, Arabic, and Japanese.\n\nSee the [launch announcement](https://www.liquid.ai/lfm-7b) for benchmarks and more info.", + "modelVersionGroupId": null, + "contextLength": 32768, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Other", + "instructType": "chatml", + "defaultSystem": null, + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "liquid/lfm-7b", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "liquid/lfm-7b", + "modelVariantPermaslug": "liquid/lfm-7b", + "providerName": "Liquid", + "providerInfo": { + "name": "Liquid", + "displayName": "Liquid", + "slug": "liquid", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://www.liquid.ai/terms-conditions", + "privacyPolicyUrl": "https://www.liquid.ai/privacy-policy", + "paidModels": { + "training": false + } + }, + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "Liquid", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.liquid.ai/&size=256", + "invertRequired": true + } + }, + "providerDisplayName": "Liquid", + "providerModelId": "lfm-7b", + "providerGroup": "Liquid", + "quantization": null, + "variant": "standard", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://www.liquid.ai/terms-conditions", + "privacyPolicyUrl": "https://www.liquid.ai/privacy-policy", + "paidModels": { + "training": false + }, + "training": false + }, + "pricing": { + "prompt": "0.00000001", + "completion": "0.00000001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 61, + "newest": 139, + "throughputHighToLow": 140, + "latencyLowToHigh": 131, + "pricingLowToHigh": 72, + "pricingHighToLow": 246 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://www.liquid.ai/\\u0026size=256", + "providers": [ + { + "name": "Liquid", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.liquid.ai/&size=256", + "slug": "liquid", + "quantization": null, + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "lfm-7b", + "pricing": { + "prompt": "0.00000001", + "completion": "0.00000001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty" + ], + "inputCost": 0.01, + "outputCost": 0.01, + "throughput": 56.0045, + "latency": 720 + } + ] + }, + { + "slug": "anthropic/claude-3-haiku", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-03-13T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Anthropic: Claude 3 Haiku", + "shortName": "Claude 3 Haiku", + "author": "anthropic", + "description": "Claude 3 Haiku is Anthropic's fastest and most compact model for\nnear-instant responsiveness. Quick and accurate targeted performance.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-haiku)\n\n#multimodal", + "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", + "contextLength": 200000, "inputModalities": [ "text", "image" @@ -20747,32 +22221,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Gemini", + "group": "Claude", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "google/gemini-pro-1.5", + "warningMessage": null, + "permaslug": "anthropic/claude-3-haiku", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "089465ba-a451-4730-a9dd-6a2ff05281f7", - "name": "Google | google/gemini-pro-1.5", - "contextLength": 2000000, + "id": "8661a1db-b0cf-4eb2-ba04-c2a79f698682", + "name": "Anthropic | anthropic/claude-3-haiku", + "contextLength": 200000, "model": { - "slug": "google/gemini-pro-1.5", + "slug": "anthropic/claude-3-haiku", "hfSlug": null, - "updatedAt": "2025-04-12T05:01:34.756668+00:00", - "createdAt": "2024-04-09T00:00:00+00:00", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-03-13T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Google: Gemini 1.5 Pro", - "shortName": "Gemini 1.5 Pro", - "author": "google", - "description": "Google's latest multimodal model, supports image and video[0] in text or chat prompts.\n\nOptimized for language tasks including:\n\n- Code generation\n- Text generation\n- Text editing\n- Problem solving\n- Recommendations\n- Information extraction\n- Data extraction or generation\n- AI agents\n\nUsage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).\n\n* [0]: Video input is not available through OpenRouter at this time.", - "modelVersionGroupId": null, - "contextLength": 2000000, + "name": "Anthropic: Claude 3 Haiku", + "shortName": "Claude 3 Haiku", + "author": "anthropic", + "description": "Claude 3 Haiku is Anthropic's fastest and most compact model for\nnear-instant responsiveness. Quick and accurate targeted performance.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-haiku)\n\n#multimodal", + "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", + "contextLength": 200000, "inputModalities": [ "text", "image" @@ -20781,115 +22255,98 @@ export default { "text" ], "hasTextOutput": true, - "group": "Gemini", + "group": "Claude", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "google/gemini-pro-1.5", + "warningMessage": null, + "permaslug": "anthropic/claude-3-haiku", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "google/gemini-pro-1.5", - "modelVariantPermaslug": "google/gemini-pro-1.5", - "providerName": "Google", + "modelVariantSlug": "anthropic/claude-3-haiku", + "modelVariantPermaslug": "anthropic/claude-3-haiku", + "providerName": "Anthropic", "providerInfo": { - "name": "Google", - "displayName": "Google Vertex", - "slug": "google-vertex", + "name": "Anthropic", + "displayName": "Anthropic", + "slug": "anthropic", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://cloud.google.com/terms/", - "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", - "dataPolicyUrl": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/abuse-monitoring", + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 }, - "freeModels": { - "training": true, - "retainsPrompts": true - } + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, - "moderationRequired": false, - "group": "Google", + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": true, + "group": "Anthropic", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.cloud.google.com/products/sdXM79fz1FS6ekNpu37K/history", + "statusPageUrl": "https://status.anthropic.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/GoogleVertex.svg" + "url": "/images/icons/Anthropic.svg" } }, - "providerDisplayName": "Google Vertex", - "providerModelId": "gemini-1.5-pro-002", - "providerGroup": "Google", + "providerDisplayName": "Anthropic", + "providerModelId": "claude-3-haiku-20240307", + "providerGroup": "Anthropic", "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 8192, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "tools", - "tool_choice", - "seed", - "response_format", - "structured_outputs" + "top_k", + "stop" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://cloud.google.com/terms/", - "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", - "dataPolicyUrl": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/abuse-monitoring", + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 }, - "freeModels": { - "training": true, - "retainsPrompts": true - }, + "requiresUserIds": true, "training": false, "retainsPrompts": true, "retentionDays": 30 }, "pricing": { - "prompt": "0.00000125", - "completion": "0.000005", - "image": "0.0006575", + "prompt": "0.00000025", + "completion": "0.00000125", + "image": "0.0004", "request": "0", + "inputCacheRead": "0.00000003", + "inputCacheWrite": "0.0000003", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [ - { - "type": "prompt-threshold", - "threshold": 128000, - "prompt": "0.0000025", - "completions": "0.00001" - } - ], + "variablePricings": [], "isHidden": false, "isDeranked": false, "isDisabled": false, @@ -20898,103 +22355,103 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 61, - "newest": 271, - "throughputHighToLow": 117, - "latencyLowToHigh": 168, - "pricingLowToHigh": 239, - "pricingHighToLow": 79 + "topWeekly": 62, + "newest": 277, + "throughputHighToLow": 28, + "latencyLowToHigh": 133, + "pricingLowToHigh": 168, + "pricingHighToLow": 149 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { - "name": "Google Vertex", - "slug": "vertex", + "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", + "slug": "anthropic", "quantization": "unknown", - "context": 2000000, - "maxCompletionTokens": 8192, - "providerModelId": "gemini-1.5-pro-002", + "context": 200000, + "maxCompletionTokens": 4096, + "providerModelId": "claude-3-haiku-20240307", "pricing": { - "prompt": "0.00000125", - "completion": "0.000005", - "image": "0.0006575", + "prompt": "0.00000025", + "completion": "0.00000125", + "image": "0.0004", "request": "0", + "inputCacheRead": "0.00000003", + "inputCacheWrite": "0.0000003", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "tools", - "tool_choice", - "seed", - "response_format", - "structured_outputs" + "top_k", + "stop" ], - "inputCost": 1.25, - "outputCost": 5 + "inputCost": 0.25, + "outputCost": 1.25, + "throughput": 152.853, + "latency": 780.5 }, { - "name": "Google AI Studio", - "slug": "google", + "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", + "slug": "vertex", "quantization": "unknown", - "context": 2000000, - "maxCompletionTokens": 8192, - "providerModelId": "gemini-1.5-pro-002", + "context": 200000, + "maxCompletionTokens": 4096, + "providerModelId": "claude-3-haiku@20240307", "pricing": { - "prompt": "0.00000125", - "completion": "0.000005", - "image": "0", + "prompt": "0.00000025", + "completion": "0.00000125", + "image": "0.0004", "request": "0", - "inputCacheRead": "0.000000625", - "inputCacheWrite": "0.000002875", + "inputCacheRead": "0.00000003", + "inputCacheWrite": "0.0000003", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "tools", - "tool_choice", - "seed", - "response_format", - "structured_outputs" + "top_k", + "stop" ], - "inputCost": 1.25, - "outputCost": 5 + "inputCost": 0.25, + "outputCost": 1.25, + "throughput": 152.9625, + "latency": 1510 } ] }, { - "slug": "qwen/qwen2.5-vl-72b-instruct", - "hfSlug": "Qwen/Qwen2.5-VL-72B-Instruct", + "slug": "mistralai/mistral-small-3.1-24b-instruct", + "hfSlug": "mistralai/Mistral-Small-3.1-24B-Instruct-2503", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-01T11:45:11.997326+00:00", + "createdAt": "2025-03-17T19:15:37.00423+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen2.5 VL 72B Instruct", - "shortName": "Qwen2.5 VL 72B Instruct", - "author": "qwen", - "description": "Qwen2.5-VL is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.", + "name": "Mistral: Mistral Small 3.1 24B", + "shortName": "Mistral Small 3.1 24B", + "author": "mistralai", + "description": "Mistral Small 3.1 24B Instruct is an upgraded variant of Mistral Small 3 (2501), featuring 24 billion parameters with advanced multimodal capabilities. It provides state-of-the-art performance in text-based reasoning and vision tasks, including image analysis, programming, mathematical reasoning, and multilingual support across dozens of languages. Equipped with an extensive 128k token context window and optimized for efficient local inference, it supports use cases such as conversational agents, function calling, long-document comprehension, and privacy-sensitive deployments.", "modelVersionGroupId": null, - "contextLength": 32000, + "contextLength": 131072, "inputModalities": [ "text", "image" @@ -21003,32 +22460,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", + "group": "Mistral", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen2.5-vl-72b-instruct", + "permaslug": "mistralai/mistral-small-3.1-24b-instruct-2503", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "2fff2f07-5438-4dcd-854c-6a8ca11fd420", - "name": "Nebius | qwen/qwen2.5-vl-72b-instruct", - "contextLength": 32000, + "id": "b4449f43-003d-40ba-949d-5888358e25ef", + "name": "Nebius | mistralai/mistral-small-3.1-24b-instruct-2503", + "contextLength": 131072, "model": { - "slug": "qwen/qwen2.5-vl-72b-instruct", - "hfSlug": "Qwen/Qwen2.5-VL-72B-Instruct", + "slug": "mistralai/mistral-small-3.1-24b-instruct", + "hfSlug": "mistralai/Mistral-Small-3.1-24B-Instruct-2503", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-01T11:45:11.997326+00:00", + "createdAt": "2025-03-17T19:15:37.00423+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen2.5 VL 72B Instruct", - "shortName": "Qwen2.5 VL 72B Instruct", - "author": "qwen", - "description": "Qwen2.5-VL is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.", + "name": "Mistral: Mistral Small 3.1 24B", + "shortName": "Mistral Small 3.1 24B", + "author": "mistralai", + "description": "Mistral Small 3.1 24B Instruct is an upgraded variant of Mistral Small 3 (2501), featuring 24 billion parameters with advanced multimodal capabilities. It provides state-of-the-art performance in text-based reasoning and vision tasks, including image analysis, programming, mathematical reasoning, and multilingual support across dozens of languages. Equipped with an extensive 128k token context window and optimized for efficient local inference, it supports use cases such as conversational agents, function calling, long-document comprehension, and privacy-sensitive deployments.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 128000, "inputModalities": [ "text", "image" @@ -21037,19 +22494,19 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", + "group": "Mistral", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen2.5-vl-72b-instruct", + "permaslug": "mistralai/mistral-small-3.1-24b-instruct-2503", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "qwen/qwen2.5-vl-72b-instruct", - "modelVariantPermaslug": "qwen/qwen2.5-vl-72b-instruct", + "modelVariantSlug": "mistralai/mistral-small-3.1-24b-instruct", + "modelVariantPermaslug": "mistralai/mistral-small-3.1-24b-instruct-2503", "providerName": "Nebius", "providerInfo": { "name": "Nebius", @@ -21082,7 +22539,7 @@ export default { } }, "providerDisplayName": "Nebius AI Studio", - "providerModelId": "Qwen/Qwen2.5-VL-72B-Instruct", + "providerModelId": "mistralai/Mistral-Small-3.1-24B-Instruct-2503", "providerGroup": "Nebius", "quantization": "fp8", "variant": "standard", @@ -21118,8 +22575,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000025", - "completion": "0.00000075", + "prompt": "0.00000005", + "completion": "0.00000015", "image": "0", "request": "0", "webSearch": "0", @@ -21144,24 +22601,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 62, - "newest": 125, - "throughputHighToLow": 297, - "latencyLowToHigh": 233, - "pricingLowToHigh": 48, - "pricingHighToLow": 154 + "topWeekly": 63, + "newest": 79, + "throughputHighToLow": 104, + "latencyLowToHigh": 33, + "pricingLowToHigh": 35, + "pricingHighToLow": 219 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", - "context": 32000, + "context": 131072, "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen2.5-VL-72B-Instruct", + "providerModelId": "mistralai/Mistral-Small-3.1-24B-Instruct-2503", "pricing": { - "prompt": "0.00000025", - "completion": "0.00000075", + "prompt": "0.00000005", + "completion": "0.00000015", "image": "0", "request": "0", "webSearch": "0", @@ -21181,22 +22640,25 @@ export default { "logprobs", "top_logprobs" ], - "inputCost": 0.25, - "outputCost": 0.75 + "inputCost": 0.05, + "outputCost": 0.15, + "throughput": 58.5635, + "latency": 635 }, { "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", "slug": "parasail", - "quantization": "fp8", + "quantization": "bf16", "context": 128000, "maxCompletionTokens": 128000, - "providerModelId": "parasail-qwen25-vl-72b-instruct", + "providerModelId": "parasail-mistral-small-31-24b-instruct", "pricing": { - "prompt": "0.0000007", - "completion": "0.0000007", - "image": "0", - "request": "0", - "webSearch": "0", + "prompt": "0.0000001", + "completion": "0.0000003", + "image": "0.000926", + "request": "0", + "webSearch": "0", "internalReasoning": "0", "discount": 0 }, @@ -21209,51 +22671,57 @@ export default { "repetition_penalty", "top_k" ], - "inputCost": 0.7, - "outputCost": 0.7 + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 67.322, + "latency": 1379 }, { - "name": "NovitaAI", - "slug": "novitaAi", + "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", + "slug": "mistral", "quantization": null, - "context": 96000, + "context": 131072, "maxCompletionTokens": null, - "providerModelId": "qwen/qwen2.5-vl-72b-instruct", + "providerModelId": "mistral-small-2503", "pricing": { - "prompt": "0.0000008", - "completion": "0.0000008", - "image": "0", + "prompt": "0.0000001", + "completion": "0.0000003", + "image": "0.0009264", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" + "response_format", + "structured_outputs", + "seed" ], - "inputCost": 0.8, - "outputCost": 0.8 + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 140.218, + "latency": 220 }, { - "name": "Together", - "slug": "together", + "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", + "slug": "cloudflare", "quantization": null, - "context": 32768, + "context": 128000, "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen2.5-VL-72B-Instruct", + "providerModelId": "@cf/mistralai/mistral-small-3.1-24b-instruct", "pricing": { - "prompt": "0.00000195", - "completion": "0.000008", + "prompt": "0.00000035", + "completion": "0.00000056", "image": "0", "request": "0", "webSearch": "0", @@ -21264,229 +22732,16 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", "top_k", + "seed", "repetition_penalty", - "logit_bias", - "min_p", - "response_format" - ], - "inputCost": 1.95, - "outputCost": 8 - } - ] - }, - { - "slug": "google/gemma-3-4b-it", - "hfSlug": "google/gemma-3-4b-it", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-13T22:38:30.653142+00:00", - "hfUpdatedAt": null, - "name": "Google: Gemma 3 4B", - "shortName": "Gemma 3 4B", - "author": "google", - "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling.", - "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", - "contextLength": 131072, - "inputModalities": [ - "text", - "image" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Gemini", - "instructType": "gemma", - "defaultSystem": null, - "defaultStops": [ - "", - "", - "" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "google/gemma-3-4b-it", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "d3de3bd4-81bc-48fb-924f-2a87b2a36e75", - "name": "DeepInfra | google/gemma-3-4b-it", - "contextLength": 131072, - "model": { - "slug": "google/gemma-3-4b-it", - "hfSlug": "google/gemma-3-4b-it", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-13T22:38:30.653142+00:00", - "hfUpdatedAt": null, - "name": "Google: Gemma 3 4B", - "shortName": "Gemma 3 4B", - "author": "google", - "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling.", - "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", - "contextLength": 131072, - "inputModalities": [ - "text", - "image" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Gemini", - "instructType": "gemma", - "defaultSystem": null, - "defaultStops": [ - "", - "", - "" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "google/gemma-3-4b-it", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "google/gemma-3-4b-it", - "modelVariantPermaslug": "google/gemma-3-4b-it", - "providerName": "DeepInfra", - "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", - "baseUrl": "url", - "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", - "paidModels": { - "training": false, - "retainsPrompts": false - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "DeepInfra", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/DeepInfra.webp" - } - }, - "providerDisplayName": "DeepInfra", - "providerModelId": "google/gemma-3-4b-it", - "providerGroup": "DeepInfra", - "quantization": "bf16", - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", - "paidModels": { - "training": false, - "retainsPrompts": false - }, - "training": false, - "retainsPrompts": false - }, - "pricing": { - "prompt": "0.00000002", - "completion": "0.00000004", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 63, - "newest": 83, - "throughputHighToLow": 62, - "latencyLowToHigh": 34, - "pricingLowToHigh": 38, - "pricingHighToLow": 241 - }, - "providers": [ - { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "google/gemma-3-4b-it", - "pricing": { - "prompt": "0.00000002", - "completion": "0.00000004", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" + "presence_penalty" ], - "inputCost": 0.02, - "outputCost": 0.04 + "inputCost": 0.35, + "outputCost": 0.56, + "throughput": 43.0765, + "latency": 390.5 } ] }, @@ -21649,14 +22904,16 @@ export default { "sorting": { "topWeekly": 64, "newest": 176, - "throughputHighToLow": 140, + "throughputHighToLow": 113, "latencyLowToHigh": 297, "pricingLowToHigh": 173, "pricingHighToLow": 145 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "Enfer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://enfer.ai/&size=256", "slug": "enfer", "quantization": "fp8", "context": 32000, @@ -21682,10 +22939,13 @@ export default { "logprobs" ], "inputCost": 0.45, - "outputCost": 0.45 + "outputCost": 0.45, + "throughput": 66.058, + "latency": 5451.5 }, { "name": "Infermatic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://infermatic.ai/&size=256", "slug": "infermatic", "quantization": null, "context": 32000, @@ -21714,209 +22974,9 @@ export default { "seed" ], "inputCost": 0.5, - "outputCost": 0.5 - } - ] - }, - { - "slug": "liquid/lfm-7b", - "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-01-25T12:08:03.736586+00:00", - "hfUpdatedAt": null, - "name": "Liquid: LFM 7B", - "shortName": "LFM 7B", - "author": "liquid", - "description": "LFM-7B, a new best-in-class language model. LFM-7B is designed for exceptional chat capabilities, including languages like Arabic and Japanese. Powered by the Liquid Foundation Model (LFM) architecture, it exhibits unique features like low memory footprint and fast inference speed. \n\nLFM-7B is the world’s best-in-class multilingual language model in English, Arabic, and Japanese.\n\nSee the [launch announcement](https://www.liquid.ai/lfm-7b) for benchmarks and more info.", - "modelVersionGroupId": null, - "contextLength": 32768, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": "chatml", - "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "liquid/lfm-7b", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "25b2cca5-53f6-40e7-b47b-191ec968b7c2", - "name": "Liquid | liquid/lfm-7b", - "contextLength": 32768, - "model": { - "slug": "liquid/lfm-7b", - "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-01-25T12:08:03.736586+00:00", - "hfUpdatedAt": null, - "name": "Liquid: LFM 7B", - "shortName": "LFM 7B", - "author": "liquid", - "description": "LFM-7B, a new best-in-class language model. LFM-7B is designed for exceptional chat capabilities, including languages like Arabic and Japanese. Powered by the Liquid Foundation Model (LFM) architecture, it exhibits unique features like low memory footprint and fast inference speed. \n\nLFM-7B is the world’s best-in-class multilingual language model in English, Arabic, and Japanese.\n\nSee the [launch announcement](https://www.liquid.ai/lfm-7b) for benchmarks and more info.", - "modelVersionGroupId": null, - "contextLength": 32768, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": "chatml", - "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "liquid/lfm-7b", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "liquid/lfm-7b", - "modelVariantPermaslug": "liquid/lfm-7b", - "providerName": "Liquid", - "providerInfo": { - "name": "Liquid", - "displayName": "Liquid", - "slug": "liquid", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://www.liquid.ai/terms-conditions", - "privacyPolicyUrl": "https://www.liquid.ai/privacy-policy", - "paidModels": { - "training": false - } - }, - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "Liquid", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.liquid.ai/&size=256", - "invertRequired": true - } - }, - "providerDisplayName": "Liquid", - "providerModelId": "lfm-7b", - "providerGroup": "Liquid", - "quantization": null, - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://www.liquid.ai/terms-conditions", - "privacyPolicyUrl": "https://www.liquid.ai/privacy-policy", - "paidModels": { - "training": false - }, - "training": false - }, - "pricing": { - "prompt": "0.00000001", - "completion": "0.00000001", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 65, - "newest": 139, - "throughputHighToLow": 177, - "latencyLowToHigh": 97, - "pricingLowToHigh": 72, - "pricingHighToLow": 246 - }, - "providers": [ - { - "name": "Liquid", - "slug": "liquid", - "quantization": null, - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "lfm-7b", - "pricing": { - "prompt": "0.00000001", - "completion": "0.00000001", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty" - ], - "inputCost": 0.01, - "outputCost": 0.01 + "outputCost": 0.5, + "throughput": 95.3575, + "latency": 525 } ] }, @@ -22103,16 +23163,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 66, + "topWeekly": 65, "newest": 102, - "throughputHighToLow": 48, - "latencyLowToHigh": 98, + "throughputHighToLow": 51, + "latencyLowToHigh": 83, "pricingLowToHigh": 43, - "pricingHighToLow": 180 + "pricingHighToLow": 183 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "bf16", "context": 131072, @@ -22143,10 +23205,13 @@ export default { "min_p" ], "inputCost": 0.15, - "outputCost": 0.2 + "outputCost": 0.2, + "throughput": 39.107, + "latency": 345 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 131072, @@ -22177,10 +23242,13 @@ export default { "top_logprobs" ], "inputCost": 0.15, - "outputCost": 0.45 + "outputCost": 0.45, + "throughput": 32.285, + "latency": 405 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": null, "context": 32768, @@ -22213,10 +23281,13 @@ export default { "logit_bias" ], "inputCost": 0.18, - "outputCost": 0.2 + "outputCost": 0.2, + "throughput": 33.979, + "latency": 715 }, { "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", "slug": "inferenceNet", "quantization": "fp8", "context": 16384, @@ -22251,6 +23322,7 @@ export default { }, { "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", "slug": "groq", "quantization": null, "context": 131072, @@ -22281,10 +23353,13 @@ export default { "seed" ], "inputCost": 0.29, - "outputCost": 0.39 + "outputCost": 0.39, + "throughput": 563.8855, + "latency": 696.5 }, { "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", "slug": "hyperbolic", "quantization": "bf16", "context": 131072, @@ -22317,10 +23392,13 @@ export default { "repetition_penalty" ], "inputCost": 0.4, - "outputCost": 0.4 + "outputCost": 0.4, + "throughput": 33.561, + "latency": 1027 }, { "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", "slug": "sambaNova", "quantization": "bf16", "context": 16384, @@ -22345,10 +23423,13 @@ export default { "stop" ], "inputCost": 0.5, - "outputCost": 1 + "outputCost": 1, + "throughput": 262.86, + "latency": 681 }, { "name": "Nebius AI Studio (Fast)", + "icon": "", "slug": "nebiusAiStudio (fast)", "quantization": "fp8", "context": 131072, @@ -22379,10 +23460,13 @@ export default { "top_logprobs" ], "inputCost": 0.5, - "outputCost": 1.5 + "outputCost": 1.5, + "throughput": 75.252, + "latency": 285 }, { "name": "CentML", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://centml.ai/&size=256", "slug": "centMl", "quantization": "fp16", "context": 40960, @@ -22412,10 +23496,13 @@ export default { "repetition_penalty" ], "inputCost": 0.65, - "outputCost": 0.65 + "outputCost": 0.65, + "throughput": 58.551, + "latency": 606 }, { "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", "slug": "fireworks", "quantization": null, "context": 131072, @@ -22448,10 +23535,13 @@ export default { "top_logprobs" ], "inputCost": 0.9, - "outputCost": 0.9 + "outputCost": 0.9, + "throughput": 133.165, + "latency": 705 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": null, "context": 131072, @@ -22482,22 +23572,24 @@ export default { "response_format" ], "inputCost": 1.2, - "outputCost": 1.2 + "outputCost": 1.2, + "throughput": 88.417, + "latency": 962 } ] }, { - "slug": "anthropic/claude-3.5-haiku-20241022", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-04T00:00:00+00:00", + "slug": "mistralai/mistral-medium-3", + "hfSlug": "", + "updatedAt": "2025-05-07T14:30:43.39556+00:00", + "createdAt": "2025-05-07T14:15:41.980763+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3.5 Haiku (2024-10-22) (self-moderated)", - "shortName": "Claude 3.5 Haiku (2024-10-22) (self-moderated)", - "author": "anthropic", - "description": "Claude 3.5 Haiku features enhancements across all skill sets including coding, tool use, and reasoning. As the fastest model in the Anthropic lineup, it offers rapid response times suitable for applications that require high interactivity and low latency, such as user-facing chatbots and on-the-fly code completions. It also excels in specialized tasks like data extraction and real-time content moderation, making it a versatile tool for a broad range of industries.\n\nIt does not support image inputs.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/3-5-models-and-computer-use)", - "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", - "contextLength": 200000, + "name": "Mistral: Mistral Medium 3", + "shortName": "Mistral Medium 3", + "author": "mistralai", + "description": "Mistral Medium 3 is a high-performance enterprise-grade language model designed to deliver frontier-level capabilities at significantly reduced operational cost. It balances state-of-the-art reasoning and multimodal performance with 8× lower cost compared to traditional large models, making it suitable for scalable deployments across professional and industrial use cases.\n\nThe model excels in domains such as coding, STEM reasoning, and enterprise adaptation. It supports hybrid, on-prem, and in-VPC deployments and is optimized for integration into custom workflows. Mistral Medium 3 offers competitive accuracy relative to larger models like Claude Sonnet 3.5/3.7, Llama 4 Maverick, and Command R+, while maintaining broad compatibility across cloud environments.", + "modelVersionGroupId": null, + "contextLength": 131072, "inputModalities": [ "text", "image" @@ -22506,23 +23598,232 @@ export default { "text" ], "hasTextOutput": true, - "group": "Claude", + "group": "Mistral", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-5-haiku-20241022", + "permaslug": "mistralai/mistral-medium-3", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "81dc58ef-b297-4540-ae41-b232d2bf5c3b", - "name": "Anthropic | anthropic/claude-3-5-haiku-20241022:beta", - "contextLength": 200000, + "id": "9d5ba5bf-8465-46df-9185-1330820338f5", + "name": "Mistral | mistralai/mistral-medium-3", + "contextLength": 131072, "model": { - "slug": "anthropic/claude-3.5-haiku-20241022", - "hfSlug": null, + "slug": "mistralai/mistral-medium-3", + "hfSlug": "", + "updatedAt": "2025-05-07T14:30:43.39556+00:00", + "createdAt": "2025-05-07T14:15:41.980763+00:00", + "hfUpdatedAt": null, + "name": "Mistral: Mistral Medium 3", + "shortName": "Mistral Medium 3", + "author": "mistralai", + "description": "Mistral Medium 3 is a high-performance enterprise-grade language model designed to deliver frontier-level capabilities at significantly reduced operational cost. It balances state-of-the-art reasoning and multimodal performance with 8× lower cost compared to traditional large models, making it suitable for scalable deployments across professional and industrial use cases.\n\nThe model excels in domains such as coding, STEM reasoning, and enterprise adaptation. It supports hybrid, on-prem, and in-VPC deployments and is optimized for integration into custom workflows. Mistral Medium 3 offers competitive accuracy relative to larger models like Claude Sonnet 3.5/3.7, Llama 4 Maverick, and Command R+, while maintaining broad compatibility across cloud environments.", + "modelVersionGroupId": null, + "contextLength": 131072, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Mistral", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "mistralai/mistral-medium-3", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "mistralai/mistral-medium-3", + "modelVariantPermaslug": "mistralai/mistral-medium-3", + "providerName": "Mistral", + "providerInfo": { + "name": "Mistral", + "displayName": "Mistral", + "slug": "mistral", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + } + }, + "headquarters": "FR", + "hasChatCompletions": true, + "hasCompletions": false, + "isAbortable": false, + "moderationRequired": false, + "group": "Mistral", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/Mistral.png" + } + }, + "providerDisplayName": "Mistral", + "providerModelId": "mistral-medium-2505", + "providerGroup": "Mistral", + "quantization": null, + "variant": "standard", + "isFree": false, + "canAbort": false, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "structured_outputs", + "seed" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "pricing": { + "prompt": "0.0000004", + "completion": "0.000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": true, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": false, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 66, + "newest": 1, + "throughputHighToLow": 136, + "latencyLowToHigh": 85, + "pricingLowToHigh": 185, + "pricingHighToLow": 133 + }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", + "providers": [ + { + "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", + "slug": "mistral", + "quantization": null, + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "mistral-medium-2505", + "pricing": { + "prompt": "0.0000004", + "completion": "0.000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "structured_outputs", + "seed" + ], + "inputCost": 0.4, + "outputCost": 2, + "throughput": 45.5665, + "latency": 478 + } + ] + }, + { + "slug": "anthropic/claude-3.5-haiku-20241022", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-11-04T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Anthropic: Claude 3.5 Haiku (2024-10-22) (self-moderated)", + "shortName": "Claude 3.5 Haiku (2024-10-22) (self-moderated)", + "author": "anthropic", + "description": "Claude 3.5 Haiku features enhancements across all skill sets including coding, tool use, and reasoning. As the fastest model in the Anthropic lineup, it offers rapid response times suitable for applications that require high interactivity and low latency, such as user-facing chatbots and on-the-fly code completions. It also excels in specialized tasks like data extraction and real-time content moderation, making it a versatile tool for a broad range of industries.\n\nIt does not support image inputs.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/3-5-models-and-computer-use)", + "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", + "contextLength": 200000, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Claude", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "anthropic/claude-3-5-haiku-20241022", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "81dc58ef-b297-4540-ae41-b232d2bf5c3b", + "name": "Anthropic | anthropic/claude-3-5-haiku-20241022:beta", + "contextLength": 200000, + "model": { + "slug": "anthropic/claude-3.5-haiku-20241022", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-11-04T00:00:00+00:00", "hfUpdatedAt": null, @@ -22649,15 +23950,17 @@ export default { }, "sorting": { "topWeekly": 67, - "newest": 177, - "throughputHighToLow": 167, - "latencyLowToHigh": 206, + "newest": 179, + "throughputHighToLow": 163, + "latencyLowToHigh": 221, "pricingLowToHigh": 219, "pricingHighToLow": 95 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", "slug": "anthropic", "quantization": "unknown", "context": 200000, @@ -22684,10 +23987,13 @@ export default { "stop" ], "inputCost": 0.8, - "outputCost": 4 + "outputCost": 4, + "throughput": 52.1635, + "latency": 1537 }, { "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", "slug": "vertex", "quantization": "unknown", "context": 200000, @@ -22714,7 +24020,9 @@ export default { "stop" ], "inputCost": 0.8, - "outputCost": 4 + "outputCost": 4, + "throughput": 69.812, + "latency": 2450 } ] }, @@ -22886,9 +24194,11 @@ export default { "pricingLowToHigh": 232, "pricingHighToLow": 87 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", "slug": "openAi", "quantization": null, "context": 200000, @@ -22913,200 +24223,9 @@ export default { "structured_outputs" ], "inputCost": 1.1, - "outputCost": 4.4 - } - ] - }, - { - "slug": "amazon/nova-lite-v1", - "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-05T22:22:43.403315+00:00", - "hfUpdatedAt": null, - "name": "Amazon: Nova Lite 1.0", - "shortName": "Nova Lite 1.0", - "author": "amazon", - "description": "Amazon Nova Lite 1.0 is a very low-cost multimodal model from Amazon that focused on fast processing of image, video, and text inputs to generate text output. Amazon Nova Lite can handle real-time customer interactions, document analysis, and visual question-answering tasks with high accuracy.\n\nWith an input context of 300K tokens, it can analyze multiple images or up to 30 minutes of video in a single input.", - "modelVersionGroupId": null, - "contextLength": 300000, - "inputModalities": [ - "text", - "image" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Nova", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "amazon/nova-lite-v1", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "72eda073-d180-4482-8e4f-81051cb66f7e", - "name": "Amazon Bedrock | amazon/nova-lite-v1", - "contextLength": 300000, - "model": { - "slug": "amazon/nova-lite-v1", - "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-05T22:22:43.403315+00:00", - "hfUpdatedAt": null, - "name": "Amazon: Nova Lite 1.0", - "shortName": "Nova Lite 1.0", - "author": "amazon", - "description": "Amazon Nova Lite 1.0 is a very low-cost multimodal model from Amazon that focused on fast processing of image, video, and text inputs to generate text output. Amazon Nova Lite can handle real-time customer interactions, document analysis, and visual question-answering tasks with high accuracy.\n\nWith an input context of 300K tokens, it can analyze multiple images or up to 30 minutes of video in a single input.", - "modelVersionGroupId": null, - "contextLength": 300000, - "inputModalities": [ - "text", - "image" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Nova", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "amazon/nova-lite-v1", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "amazon/nova-lite-v1", - "modelVariantPermaslug": "amazon/nova-lite-v1", - "providerName": "Amazon Bedrock", - "providerInfo": { - "name": "Amazon Bedrock", - "displayName": "Amazon Bedrock", - "slug": "amazon-bedrock", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://aws.amazon.com/service-terms/", - "privacyPolicyUrl": "https://aws.amazon.com/privacy", - "dataPolicyUrl": "https://docs.aws.amazon.com/bedrock/latest/userguide/data-protection.html", - "paidModels": { - "training": false, - "retainsPrompts": false - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, - "moderationRequired": true, - "group": "Amazon Bedrock", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/Bedrock.svg" - } - }, - "providerDisplayName": "Amazon Bedrock", - "providerModelId": "us.amazon.nova-lite-v1:0", - "providerGroup": "Amazon Bedrock", - "quantization": null, - "variant": "standard", - "isFree": false, - "canAbort": false, - "maxPromptTokens": null, - "maxCompletionTokens": 5120, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "tools", - "max_tokens", - "temperature", - "top_p", - "top_k", - "stop" - ], - "isByok": false, - "moderationRequired": true, - "dataPolicy": { - "termsOfServiceUrl": "https://aws.amazon.com/service-terms/", - "privacyPolicyUrl": "https://aws.amazon.com/privacy", - "dataPolicyUrl": "https://docs.aws.amazon.com/bedrock/latest/userguide/data-protection.html", - "paidModels": { - "training": false, - "retainsPrompts": false - }, - "training": false, - "retainsPrompts": false - }, - "pricing": { - "prompt": "0.00000006", - "completion": "0.00000024", - "image": "0.00009", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": false, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 69, - "newest": 159, - "throughputHighToLow": 18, - "latencyLowToHigh": 49, - "pricingLowToHigh": 105, - "pricingHighToLow": 214 - }, - "providers": [ - { - "name": "Amazon Bedrock", - "slug": "amazonBedrock", - "quantization": null, - "context": 300000, - "maxCompletionTokens": 5120, - "providerModelId": "us.amazon.nova-lite-v1:0", - "pricing": { - "prompt": "0.00000006", - "completion": "0.00000024", - "image": "0.00009", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "max_tokens", - "temperature", - "top_p", - "top_k", - "stop" - ], - "inputCost": 0.06, - "outputCost": 0.24 + "outputCost": 4.4, + "throughput": 92.77, + "latency": 5858.5 } ] }, @@ -23276,16 +24395,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 29, + "topWeekly": 27, "newest": 93, - "throughputHighToLow": 91, - "latencyLowToHigh": 152, + "throughputHighToLow": 35, + "latencyLowToHigh": 143, "pricingLowToHigh": 41, "pricingHighToLow": 199 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "bf16", "context": 131072, @@ -23314,14 +24435,17 @@ export default { "min_p" ], "inputCost": 0.1, - "outputCost": 0.2 + "outputCost": 0.2, + "throughput": 32.525, + "latency": 868 }, { "name": "InoCloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inocloud.com/&size=256", "slug": "inoCloud", "quantization": "bf16", - "context": 128000, - "maxCompletionTokens": 32768, + "context": 131072, + "maxCompletionTokens": 16384, "providerModelId": "google/gemma-3-27b-it", "pricing": { "prompt": "0.0000001", @@ -23341,10 +24465,13 @@ export default { "presence_penalty" ], "inputCost": 0.1, - "outputCost": 0.2 + "outputCost": 0.2, + "throughput": 27.1265, + "latency": 1183 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 110000, @@ -23373,10 +24500,13 @@ export default { "top_logprobs" ], "inputCost": 0.1, - "outputCost": 0.3 + "outputCost": 0.3, + "throughput": 49.53, + "latency": 918 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": null, "context": 32000, @@ -23405,10 +24535,13 @@ export default { "logit_bias" ], "inputCost": 0.12, - "outputCost": 0.2 + "outputCost": 0.2, + "throughput": 31.0395, + "latency": 2320 }, { "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", "slug": "klusterAi", "quantization": null, "context": 64000, @@ -23425,13 +24558,27 @@ export default { }, "supportedParameters": [ "max_tokens", - "temperature" + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", + "seed" ], "inputCost": 0.2, - "outputCost": 0.35 + "outputCost": 0.35, + "throughput": 54.094, + "latency": 1346 }, { "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", "slug": "parasail", "quantization": "bf16", "context": 131072, @@ -23456,10 +24603,13 @@ export default { "top_k" ], "inputCost": 0.25, - "outputCost": 0.4 + "outputCost": 0.4, + "throughput": 50.5485, + "latency": 1250.5 }, { "name": "nCompass", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://console.ncompass.tech/&size=256", "slug": "nCompass", "quantization": "bf16", "context": 32768, @@ -23487,10 +24637,13 @@ export default { "repetition_penalty" ], "inputCost": 0.3, - "outputCost": 0.3 + "outputCost": 0.3, + "throughput": 43.6385, + "latency": 998 }, { "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", "slug": "inferenceNet", "quantization": "bf16", "context": 128000, @@ -23519,7 +24672,9 @@ export default { "top_logprobs" ], "inputCost": 0.3, - "outputCost": 0.4 + "outputCost": 0.4, + "throughput": 15.504, + "latency": 2100 } ] }, @@ -23685,16 +24840,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 71, - "newest": 250, + "topWeekly": 70, + "newest": 251, "throughputHighToLow": 33, - "latencyLowToHigh": 163, + "latencyLowToHigh": 28, "pricingLowToHigh": 81, "pricingHighToLow": 237 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nousresearch.com/\\u0026size=256", "providers": [ { "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", "slug": "lambda", "quantization": "bf16", "context": 131072, @@ -23723,10 +24880,13 @@ export default { "response_format" ], "inputCost": 0.02, - "outputCost": 0.04 + "outputCost": 0.04, + "throughput": 153.6935, + "latency": 310 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "unknown", "context": 8192, @@ -23755,22 +24915,24 @@ export default { "logit_bias" ], "inputCost": 0.14, - "outputCost": 0.14 + "outputCost": 0.14, + "throughput": 132.7985, + "latency": 683 } ] }, { - "slug": "mistralai/mistral-small-3.1-24b-instruct", - "hfSlug": "mistralai/Mistral-Small-3.1-24B-Instruct-2503", + "slug": "amazon/nova-lite-v1", + "hfSlug": "", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-17T19:15:37.00423+00:00", + "createdAt": "2024-12-05T22:22:43.403315+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral Small 3.1 24B", - "shortName": "Mistral Small 3.1 24B", - "author": "mistralai", - "description": "Mistral Small 3.1 24B Instruct is an upgraded variant of Mistral Small 3 (2501), featuring 24 billion parameters with advanced multimodal capabilities. It provides state-of-the-art performance in text-based reasoning and vision tasks, including image analysis, programming, mathematical reasoning, and multilingual support across dozens of languages. Equipped with an extensive 128k token context window and optimized for efficient local inference, it supports use cases such as conversational agents, function calling, long-document comprehension, and privacy-sensitive deployments.", + "name": "Amazon: Nova Lite 1.0", + "shortName": "Nova Lite 1.0", + "author": "amazon", + "description": "Amazon Nova Lite 1.0 is a very low-cost multimodal model from Amazon that focused on fast processing of image, video, and text inputs to generate text output. Amazon Nova Lite can handle real-time customer interactions, document analysis, and visual question-answering tasks with high accuracy.\n\nWith an input context of 300K tokens, it can analyze multiple images or up to 30 minutes of video in a single input.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 300000, "inputModalities": [ "text", "image" @@ -23779,32 +24941,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "Nova", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-small-3.1-24b-instruct-2503", + "permaslug": "amazon/nova-lite-v1", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "b4449f43-003d-40ba-949d-5888358e25ef", - "name": "Nebius | mistralai/mistral-small-3.1-24b-instruct-2503", - "contextLength": 131072, + "id": "72eda073-d180-4482-8e4f-81051cb66f7e", + "name": "Amazon Bedrock | amazon/nova-lite-v1", + "contextLength": 300000, "model": { - "slug": "mistralai/mistral-small-3.1-24b-instruct", - "hfSlug": "mistralai/Mistral-Small-3.1-24B-Instruct-2503", + "slug": "amazon/nova-lite-v1", + "hfSlug": "", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-17T19:15:37.00423+00:00", + "createdAt": "2024-12-05T22:22:43.403315+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral Small 3.1 24B", - "shortName": "Mistral Small 3.1 24B", - "author": "mistralai", - "description": "Mistral Small 3.1 24B Instruct is an upgraded variant of Mistral Small 3 (2501), featuring 24 billion parameters with advanced multimodal capabilities. It provides state-of-the-art performance in text-based reasoning and vision tasks, including image analysis, programming, mathematical reasoning, and multilingual support across dozens of languages. Equipped with an extensive 128k token context window and optimized for efficient local inference, it supports use cases such as conversational agents, function calling, long-document comprehension, and privacy-sensitive deployments.", + "name": "Amazon: Nova Lite 1.0", + "shortName": "Nova Lite 1.0", + "author": "amazon", + "description": "Amazon Nova Lite 1.0 is a very low-cost multimodal model from Amazon that focused on fast processing of image, video, and text inputs to generate text output. Amazon Nova Lite can handle real-time customer interactions, document analysis, and visual question-answering tasks with high accuracy.\n\nWith an input context of 300K tokens, it can analyze multiple images or up to 30 minutes of video in a single input.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 300000, "inputModalities": [ "text", "image" @@ -23813,19 +24975,216 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "Nova", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-small-3.1-24b-instruct-2503", + "permaslug": "amazon/nova-lite-v1", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "mistralai/mistral-small-3.1-24b-instruct", - "modelVariantPermaslug": "mistralai/mistral-small-3.1-24b-instruct-2503", + "modelVariantSlug": "amazon/nova-lite-v1", + "modelVariantPermaslug": "amazon/nova-lite-v1", + "providerName": "Amazon Bedrock", + "providerInfo": { + "name": "Amazon Bedrock", + "displayName": "Amazon Bedrock", + "slug": "amazon-bedrock", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://aws.amazon.com/service-terms/", + "privacyPolicyUrl": "https://aws.amazon.com/privacy", + "dataPolicyUrl": "https://docs.aws.amazon.com/bedrock/latest/userguide/data-protection.html", + "paidModels": { + "training": false, + "retainsPrompts": false + } + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": false, + "isAbortable": false, + "moderationRequired": true, + "group": "Amazon Bedrock", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/Bedrock.svg" + } + }, + "providerDisplayName": "Amazon Bedrock", + "providerModelId": "us.amazon.nova-lite-v1:0", + "providerGroup": "Amazon Bedrock", + "quantization": null, + "variant": "standard", + "isFree": false, + "canAbort": false, + "maxPromptTokens": null, + "maxCompletionTokens": 5120, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "tools", + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "isByok": false, + "moderationRequired": true, + "dataPolicy": { + "termsOfServiceUrl": "https://aws.amazon.com/service-terms/", + "privacyPolicyUrl": "https://aws.amazon.com/privacy", + "dataPolicyUrl": "https://docs.aws.amazon.com/bedrock/latest/userguide/data-protection.html", + "paidModels": { + "training": false, + "retainsPrompts": false + }, + "training": false, + "retainsPrompts": false + }, + "pricing": { + "prompt": "0.00000006", + "completion": "0.00000024", + "image": "0.00009", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": true, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": false, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 71, + "newest": 159, + "throughputHighToLow": 17, + "latencyLowToHigh": 49, + "pricingLowToHigh": 105, + "pricingHighToLow": 214 + }, + "authorIcon": "https://openrouter.ai/images/icons/Bedrock.svg", + "providers": [ + { + "name": "Amazon Bedrock", + "icon": "https://openrouter.ai/images/icons/Bedrock.svg", + "slug": "amazonBedrock", + "quantization": null, + "context": 300000, + "maxCompletionTokens": 5120, + "providerModelId": "us.amazon.nova-lite-v1:0", + "pricing": { + "prompt": "0.00000006", + "completion": "0.00000024", + "image": "0.00009", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 0.06, + "outputCost": 0.24, + "throughput": 172.302, + "latency": 412 + } + ] + }, + { + "slug": "qwen/qwen2.5-vl-72b-instruct", + "hfSlug": "Qwen/Qwen2.5-VL-72B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-01T11:45:11.997326+00:00", + "hfUpdatedAt": null, + "name": "Qwen: Qwen2.5 VL 72B Instruct", + "shortName": "Qwen2.5 VL 72B Instruct", + "author": "qwen", + "description": "Qwen2.5-VL is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.", + "modelVersionGroupId": null, + "contextLength": 32000, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Qwen", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "qwen/qwen2.5-vl-72b-instruct", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "2fff2f07-5438-4dcd-854c-6a8ca11fd420", + "name": "Nebius | qwen/qwen2.5-vl-72b-instruct", + "contextLength": 32000, + "model": { + "slug": "qwen/qwen2.5-vl-72b-instruct", + "hfSlug": "Qwen/Qwen2.5-VL-72B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-01T11:45:11.997326+00:00", + "hfUpdatedAt": null, + "name": "Qwen: Qwen2.5 VL 72B Instruct", + "shortName": "Qwen2.5 VL 72B Instruct", + "author": "qwen", + "description": "Qwen2.5-VL is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.", + "modelVersionGroupId": null, + "contextLength": 131072, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Qwen", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "qwen/qwen2.5-vl-72b-instruct", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "qwen/qwen2.5-vl-72b-instruct", + "modelVariantPermaslug": "qwen/qwen2.5-vl-72b-instruct", "providerName": "Nebius", "providerInfo": { "name": "Nebius", @@ -23858,7 +25217,7 @@ export default { } }, "providerDisplayName": "Nebius AI Studio", - "providerModelId": "mistralai/Mistral-Small-3.1-24B-Instruct-2503", + "providerModelId": "Qwen/Qwen2.5-VL-72B-Instruct", "providerGroup": "Nebius", "quantization": "fp8", "variant": "standard", @@ -23894,8 +25253,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000005", - "completion": "0.00000015", + "prompt": "0.00000025", + "completion": "0.00000075", "image": "0", "request": "0", "webSearch": "0", @@ -23921,23 +25280,25 @@ export default { }, "sorting": { "topWeekly": 72, - "newest": 79, - "throughputHighToLow": 96, - "latencyLowToHigh": 42, - "pricingLowToHigh": 35, - "pricingHighToLow": 219 + "newest": 125, + "throughputHighToLow": 249, + "latencyLowToHigh": 257, + "pricingLowToHigh": 48, + "pricingHighToLow": 154 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", - "context": 131072, + "context": 32000, "maxCompletionTokens": null, - "providerModelId": "mistralai/Mistral-Small-3.1-24B-Instruct-2503", + "providerModelId": "Qwen/Qwen2.5-VL-72B-Instruct", "pricing": { - "prompt": "0.00000005", - "completion": "0.00000015", + "prompt": "0.00000025", + "completion": "0.00000075", "image": "0", "request": "0", "webSearch": "0", @@ -23957,20 +25318,23 @@ export default { "logprobs", "top_logprobs" ], - "inputCost": 0.05, - "outputCost": 0.15 + "inputCost": 0.25, + "outputCost": 0.75, + "throughput": 13.064, + "latency": 2040.5 }, { "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", "slug": "parasail", - "quantization": "bf16", + "quantization": "fp8", "context": 128000, "maxCompletionTokens": 128000, - "providerModelId": "parasail-mistral-small-31-24b-instruct", + "providerModelId": "parasail-qwen25-vl-72b-instruct", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", - "image": "0.000926", + "prompt": "0.0000007", + "completion": "0.0000007", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -23985,51 +25349,57 @@ export default { "repetition_penalty", "top_k" ], - "inputCost": 0.1, - "outputCost": 0.3 + "inputCost": 0.7, + "outputCost": 0.7, + "throughput": 33.3855, + "latency": 1933 }, { - "name": "Mistral", - "slug": "mistral", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", "quantization": null, - "context": 131072, + "context": 96000, "maxCompletionTokens": null, - "providerModelId": "mistral-small-2503", + "providerModelId": "qwen/qwen2.5-vl-72b-instruct", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", - "image": "0.0009264", + "prompt": "0.0000008", + "completion": "0.0000008", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "response_format", - "structured_outputs", - "seed" + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" ], - "inputCost": 0.1, - "outputCost": 0.3 + "inputCost": 0.8, + "outputCost": 0.8, + "throughput": 17.542, + "latency": 4798 }, { - "name": "Cloudflare", - "slug": "cloudflare", + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", "quantization": null, - "context": 128000, + "context": 32768, "maxCompletionTokens": null, - "providerModelId": "@cf/mistralai/mistral-small-3.1-24b-instruct", + "providerModelId": "Qwen/Qwen2.5-VL-72B-Instruct", "pricing": { - "prompt": "0.00000035", - "completion": "0.00000056", + "prompt": "0.00000195", + "completion": "0.000008", "image": "0", "request": "0", "webSearch": "0", @@ -24040,29 +25410,34 @@ export default { "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "top_k", - "seed", "repetition_penalty", - "frequency_penalty", - "presence_penalty" + "logit_bias", + "min_p", + "response_format" ], - "inputCost": 0.35, - "outputCost": 0.56 + "inputCost": 1.95, + "outputCost": 8, + "throughput": 31.478, + "latency": 1337 } ] }, { - "slug": "mistralai/mistral-medium-3", + "slug": "amazon/nova-pro-v1", "hfSlug": "", - "updatedAt": "2025-05-07T14:30:43.39556+00:00", - "createdAt": "2025-05-07T14:15:41.980763+00:00", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-12-05T22:05:03.587216+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral Medium 3", - "shortName": "Mistral Medium 3", - "author": "mistralai", - "description": "Mistral Medium 3 is a high-performance enterprise-grade language model designed to deliver frontier-level capabilities at significantly reduced operational cost. It balances state-of-the-art reasoning and multimodal performance with 8× lower cost compared to traditional large models, making it suitable for scalable deployments across professional and industrial use cases.\n\nThe model excels in domains such as coding, STEM reasoning, and enterprise adaptation. It supports hybrid, on-prem, and in-VPC deployments and is optimized for integration into custom workflows. Mistral Medium 3 offers competitive accuracy relative to larger models like Claude Sonnet 3.5/3.7, Llama 4 Maverick, and Command R+, while maintaining broad compatibility across cloud environments.", + "name": "Amazon: Nova Pro 1.0", + "shortName": "Nova Pro 1.0", + "author": "amazon", + "description": "Amazon Nova Pro 1.0 is a capable multimodal model from Amazon focused on providing a combination of accuracy, speed, and cost for a wide range of tasks. As of December 2024, it achieves state-of-the-art performance on key benchmarks including visual question answering (TextVQA) and video understanding (VATEX).\n\nAmazon Nova Pro demonstrates strong capabilities in processing both visual and textual information and at analyzing financial documents.\n\n**NOTE**: Video input is not supported at this time.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 300000, "inputModalities": [ "text", "image" @@ -24071,32 +25446,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "Nova", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-medium-3", + "permaslug": "amazon/nova-pro-v1", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "9d5ba5bf-8465-46df-9185-1330820338f5", - "name": "Mistral | mistralai/mistral-medium-3", - "contextLength": 131072, + "id": "959381a4-8054-450f-9daf-5fcab64ba9aa", + "name": "Amazon Bedrock | amazon/nova-pro-v1", + "contextLength": 300000, "model": { - "slug": "mistralai/mistral-medium-3", + "slug": "amazon/nova-pro-v1", "hfSlug": "", - "updatedAt": "2025-05-07T14:30:43.39556+00:00", - "createdAt": "2025-05-07T14:15:41.980763+00:00", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-12-05T22:05:03.587216+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral Medium 3", - "shortName": "Mistral Medium 3", - "author": "mistralai", - "description": "Mistral Medium 3 is a high-performance enterprise-grade language model designed to deliver frontier-level capabilities at significantly reduced operational cost. It balances state-of-the-art reasoning and multimodal performance with 8× lower cost compared to traditional large models, making it suitable for scalable deployments across professional and industrial use cases.\n\nThe model excels in domains such as coding, STEM reasoning, and enterprise adaptation. It supports hybrid, on-prem, and in-VPC deployments and is optimized for integration into custom workflows. Mistral Medium 3 offers competitive accuracy relative to larger models like Claude Sonnet 3.5/3.7, Llama 4 Maverick, and Command R+, while maintaining broad compatibility across cloud environments.", + "name": "Amazon: Nova Pro 1.0", + "shortName": "Nova Pro 1.0", + "author": "amazon", + "description": "Amazon Nova Pro 1.0 is a capable multimodal model from Amazon focused on providing a combination of accuracy, speed, and cost for a wide range of tasks. As of December 2024, it achieves state-of-the-art performance on key benchmarks including visual question answering (TextVQA) and video understanding (VATEX).\n\nAmazon Nova Pro demonstrates strong capabilities in processing both visual and textual information and at analyzing financial documents.\n\n**NOTE**: Video input is not supported at this time.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 300000, "inputModalities": [ "text", "image" @@ -24105,40 +25480,40 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "Nova", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-medium-3", + "permaslug": "amazon/nova-pro-v1", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "mistralai/mistral-medium-3", - "modelVariantPermaslug": "mistralai/mistral-medium-3", - "providerName": "Mistral", + "modelVariantSlug": "amazon/nova-pro-v1", + "modelVariantPermaslug": "amazon/nova-pro-v1", + "providerName": "Amazon Bedrock", "providerInfo": { - "name": "Mistral", - "displayName": "Mistral", - "slug": "mistral", + "name": "Amazon Bedrock", + "displayName": "Amazon Bedrock", + "slug": "amazon-bedrock", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "termsOfServiceUrl": "https://aws.amazon.com/service-terms/", + "privacyPolicyUrl": "https://aws.amazon.com/privacy", + "dataPolicyUrl": "https://docs.aws.amazon.com/bedrock/latest/userguide/data-protection.html", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false } }, - "headquarters": "FR", + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": false, "isAbortable": false, - "moderationRequired": false, - "group": "Mistral", + "moderationRequired": true, + "group": "Amazon Bedrock", "editors": [], "owners": [], "isMultipartSupported": true, @@ -24146,51 +25521,45 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Mistral.png" + "url": "/images/icons/Bedrock.svg" } }, - "providerDisplayName": "Mistral", - "providerModelId": "mistral-medium-2505", - "providerGroup": "Mistral", + "providerDisplayName": "Amazon Bedrock", + "providerModelId": "us.amazon.nova-pro-v1:0", + "providerGroup": "Amazon Bedrock", "quantization": null, "variant": "standard", "isFree": false, "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 5120, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "structured_outputs", - "seed" + "top_k", + "stop" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "termsOfServiceUrl": "https://aws.amazon.com/service-terms/", + "privacyPolicyUrl": "https://aws.amazon.com/privacy", + "dataPolicyUrl": "https://docs.aws.amazon.com/bedrock/latest/userguide/data-protection.html", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, "pricing": { - "prompt": "0.0000004", - "completion": "0.000002", - "image": "0", + "prompt": "0.0000008", + "completion": "0.0000032", + "image": "0.0012", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -24208,31 +25577,32 @@ export default { "hasCompletions": false, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 73, - "newest": 1, - "throughputHighToLow": 228, - "latencyLowToHigh": 64, - "pricingLowToHigh": 185, - "pricingHighToLow": 133 + "newest": 161, + "throughputHighToLow": 47, + "latencyLowToHigh": 101, + "pricingLowToHigh": 218, + "pricingHighToLow": 101 }, + "authorIcon": "https://openrouter.ai/images/icons/Bedrock.svg", "providers": [ { - "name": "Mistral", - "slug": "mistral", + "name": "Amazon Bedrock", + "icon": "https://openrouter.ai/images/icons/Bedrock.svg", + "slug": "amazonBedrock", "quantization": null, - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "mistral-medium-2505", + "context": 300000, + "maxCompletionTokens": 5120, + "providerModelId": "us.amazon.nova-pro-v1:0", "pricing": { - "prompt": "0.0000004", - "completion": "0.000002", - "image": "0", + "prompt": "0.0000008", + "completion": "0.0000032", + "image": "0.0012", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -24240,19 +25610,16 @@ export default { }, "supportedParameters": [ "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "structured_outputs", - "seed" + "top_k", + "stop" ], - "inputCost": 0.4, - "outputCost": 2 + "inputCost": 0.8, + "outputCost": 3.2, + "throughput": 127.811, + "latency": 630.5 } ] }, @@ -24415,14 +25782,16 @@ export default { "sorting": { "topWeekly": 74, "newest": 19, - "throughputHighToLow": 105, - "latencyLowToHigh": 139, + "throughputHighToLow": 110, + "latencyLowToHigh": 197, "pricingLowToHigh": 8, "pricingHighToLow": 123 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [ { "name": "GMICloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://gmicloud.ai/&size=256", "slug": "gmiCloud", "quantization": "fp8", "context": 131072, @@ -24444,10 +25813,13 @@ export default { "seed" ], "inputCost": 0.5, - "outputCost": 2.18 + "outputCost": 2.18, + "throughput": 62.1385, + "latency": 1209 }, { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", "context": 163840, @@ -24476,10 +25848,13 @@ export default { "min_p" ], "inputCost": 0.7, - "outputCost": 2.18 + "outputCost": 2.18, + "throughput": 52.849, + "latency": 1226 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "fp8", "context": 160000, @@ -24508,81 +25883,79 @@ export default { "logit_bias" ], "inputCost": 0.7, - "outputCost": 2.5 + "outputCost": 2.5, + "throughput": 24.435, + "latency": 1327 } ] }, { - "slug": "mistralai/mistral-nemo", - "hfSlug": "mistralai/Mistral-Nemo-Instruct-2407", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-07-19T00:00:00+00:00", + "slug": "meta-llama/llama-4-scout", + "hfSlug": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "updatedAt": "2025-04-05T19:34:05.019062+00:00", + "createdAt": "2025-04-05T19:31:59.735804+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral Nemo (free)", - "shortName": "Mistral Nemo (free)", - "author": "mistralai", - "description": "A 12B parameter model with a 128k token context length built by Mistral in collaboration with NVIDIA.\n\nThe model is multilingual, supporting English, French, German, Spanish, Italian, Portuguese, Chinese, Japanese, Korean, Arabic, and Hindi.\n\nIt supports function calling and is released under the Apache 2.0 license.", + "name": "Meta: Llama 4 Scout (free)", + "shortName": "Llama 4 Scout (free)", + "author": "meta-llama", + "description": "Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimodal input (text and image) and multilingual output (text and code) across 12 supported languages. Designed for assistant-style interaction and visual reasoning, Scout uses 16 experts per forward pass and features a context length of 10 million tokens, with a training corpus of ~40 trillion tokens.\n\nBuilt for high efficiency and local or commercial deployment, Llama 4 Scout incorporates early fusion for seamless modality integration. It is instruction-tuned for use in multilingual chat, captioning, and image understanding tasks. Released under the Llama 4 Community License, it was last trained on data up to August 2024 and launched publicly on April 5, 2025.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 512000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "mistral", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "[INST]", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-nemo", + "permaslug": "meta-llama/llama-4-scout-17b-16e-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "d35a6a62-895e-4130-8ab6-3a7dcca5edeb", - "name": "Chutes | mistralai/mistral-nemo:free", - "contextLength": 128000, + "id": "f0b6a83b-9818-4f8c-9941-20b3ead09379", + "name": "Chutes | meta-llama/llama-4-scout-17b-16e-instruct:free", + "contextLength": 512000, "model": { - "slug": "mistralai/mistral-nemo", - "hfSlug": "mistralai/Mistral-Nemo-Instruct-2407", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-07-19T00:00:00+00:00", + "slug": "meta-llama/llama-4-scout", + "hfSlug": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "updatedAt": "2025-04-05T19:34:05.019062+00:00", + "createdAt": "2025-04-05T19:31:59.735804+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral Nemo", - "shortName": "Mistral Nemo", - "author": "mistralai", - "description": "A 12B parameter model with a 128k token context length built by Mistral in collaboration with NVIDIA.\n\nThe model is multilingual, supporting English, French, German, Spanish, Italian, Portuguese, Chinese, Japanese, Korean, Arabic, and Hindi.\n\nIt supports function calling and is released under the Apache 2.0 license.", + "name": "Meta: Llama 4 Scout", + "shortName": "Llama 4 Scout", + "author": "meta-llama", + "description": "Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimodal input (text and image) and multilingual output (text and code) across 12 supported languages. Designed for assistant-style interaction and visual reasoning, Scout uses 16 experts per forward pass and features a context length of 10 million tokens, with a training corpus of ~40 trillion tokens.\n\nBuilt for high efficiency and local or commercial deployment, Llama 4 Scout incorporates early fusion for seamless modality integration. It is instruction-tuned for use in multilingual chat, captioning, and image understanding tasks. Released under the Llama 4 Community License, it was last trained on data up to August 2024 and launched publicly on April 5, 2025.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 10000000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "mistral", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "[INST]", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-nemo", + "permaslug": "meta-llama/llama-4-scout-17b-16e-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "mistralai/mistral-nemo:free", - "modelVariantPermaslug": "mistralai/mistral-nemo:free", + "modelVariantSlug": "meta-llama/llama-4-scout:free", + "modelVariantPermaslug": "meta-llama/llama-4-scout-17b-16e-instruct:free", "providerName": "Chutes", "providerInfo": { "name": "Chutes", @@ -24612,20 +25985,22 @@ export default { } }, "providerDisplayName": "Chutes", - "providerModelId": "unsloth/Mistral-Nemo-Instruct-2407", + "providerModelId": "chutesai/Llama-4-Scout-17B-16E-Instruct", "providerGroup": "Chutes", - "quantization": null, + "quantization": "bf16", "variant": "free", "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 128000, - "maxPromptImages": null, + "maxCompletionTokens": null, + "maxPromptImages": 8, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", + "structured_outputs", + "response_format", "stop", "frequency_penalty", "presence_penalty", @@ -24669,29 +26044,35 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": { + "responseFormat": true, + "structuredOutputs": true + }, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 15, - "newest": 234, - "throughputHighToLow": 155, - "latencyLowToHigh": 290, - "pricingLowToHigh": 68, - "pricingHighToLow": 233 + "topWeekly": 42, + "newest": 63, + "throughputHighToLow": 76, + "latencyLowToHigh": 43, + "pricingLowToHigh": 27, + "pricingHighToLow": 201 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { - "name": "Enfer", - "slug": "enfer", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 65536, - "providerModelId": "mistralai/mistral-nemo", + "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", + "slug": "lambda", + "quantization": "fp8", + "context": 1048576, + "maxCompletionTokens": 1048576, + "providerModelId": "llama-4-scout-17b-16e-instruct", "pricing": { - "prompt": "0.00000003", - "completion": "0.00000007", + "prompt": "0.00000008", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -24699,31 +26080,35 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", + "seed", "logit_bias", - "logprobs" + "logprobs", + "top_logprobs", + "response_format" ], - "inputCost": 0.03, - "outputCost": 0.07 + "inputCost": 0.08, + "outputCost": 0.3, + "throughput": 99.061, + "latency": 792 }, { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "bf16", - "context": 131072, + "context": 327680, "maxCompletionTokens": 16384, - "providerModelId": "mistralai/Mistral-Nemo-Instruct-2407", + "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", "pricing": { - "prompt": "0.000000035", - "completion": "0.00000008", - "image": "0", + "prompt": "0.00000008", + "completion": "0.0000003", + "image": "0.0003342", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -24742,20 +26127,23 @@ export default { "seed", "min_p" ], - "inputCost": 0.04, - "outputCost": 0.08 + "inputCost": 0.08, + "outputCost": 0.3, + "throughput": 44.971, + "latency": 839 }, { - "name": "inference.net", - "slug": "inferenceNet", - "quantization": "fp8", - "context": 16384, - "maxCompletionTokens": 16384, - "providerModelId": "mistralai/mistral-nemo-12b-instruct/fp-8", + "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", + "slug": "klusterAi", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", "pricing": { - "prompt": "0.0000000375", - "completion": "0.0000001", - "image": "0", + "prompt": "0.00000008", + "completion": "0.00000045", + "image": "0.0005013", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -24768,25 +26156,61 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", "top_k", - "min_p", + "repetition_penalty", "logit_bias", - "top_logprobs" + "logprobs", + "top_logprobs", + "min_p", + "seed" ], - "inputCost": 0.04, - "outputCost": 0.1 + "inputCost": 0.08, + "outputCost": 0.45, + "throughput": 59.7695, + "latency": 586 }, { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", "quantization": "fp8", - "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "mistralai/Mistral-Nemo-Instruct-2407", + "context": 158000, + "maxCompletionTokens": 158000, + "providerModelId": "parasail-llama-4-scout-instruct", "pricing": { - "prompt": "0.00000004", - "completion": "0.00000012", + "prompt": "0.00000009", + "completion": "0.00000048", + "image": "0.00046788", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "presence_penalty", + "frequency_penalty", + "repetition_penalty", + "top_k" + ], + "inputCost": 0.09, + "outputCost": 0.48, + "throughput": 95.9025, + "latency": 595 + }, + { + "name": "CentML", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://centml.ai/&size=256", + "slug": "centMl", + "quantization": "bf16", + "context": 1048576, + "maxCompletionTokens": 1048576, + "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000001", "image": "0", "request": "0", "webSearch": "0", @@ -24802,32 +26226,32 @@ export default { "presence_penalty", "seed", "top_k", - "logit_bias", - "logprobs", - "top_logprobs" + "min_p", + "repetition_penalty" ], - "inputCost": 0.04, - "outputCost": 0.12 + "inputCost": 0.1, + "outputCost": 0.1, + "throughput": 81.311, + "latency": 403 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", - "quantization": "unknown", + "quantization": null, "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "mistralai/mistral-nemo", + "maxCompletionTokens": 131072, + "providerModelId": "meta-llama/llama-4-scout-17b-16e-instruct", "pricing": { - "prompt": "0.00000004", - "completion": "0.00000017", - "image": "0", + "prompt": "0.0000001", + "completion": "0.0000005", + "image": "0.0003342", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -24840,26 +26264,31 @@ export default { "repetition_penalty", "logit_bias" ], - "inputCost": 0.04, - "outputCost": 0.17 + "inputCost": 0.1, + "outputCost": 0.5, + "throughput": 33.7425, + "latency": 1281 }, { - "name": "NextBit", - "slug": "nextBit", - "quantization": "bf16", - "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "mistral:nemo", + "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", + "slug": "groq", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 8192, + "providerModelId": "meta-llama/llama-4-scout-17b-16e-instruct", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", - "image": "0", + "prompt": "0.00000011", + "completion": "0.00000034", + "image": "0.00036762", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", @@ -24867,21 +26296,27 @@ export default { "frequency_penalty", "presence_penalty", "response_format", - "structured_outputs" + "top_logprobs", + "logprobs", + "logit_bias", + "seed" ], - "inputCost": 0.1, - "outputCost": 0.1 + "inputCost": 0.11, + "outputCost": 0.34, + "throughput": 707.6425, + "latency": 628 }, { - "name": "Atoma", - "slug": "atoma", + "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", + "slug": "fireworks", "quantization": null, - "context": 128000, - "maxCompletionTokens": 80000, - "providerModelId": "mistralai/Mistral-Nemo-Instruct-2407", + "context": 1048576, + "maxCompletionTokens": null, + "providerModelId": "accounts/fireworks/models/llama4-scout-instruct-basic", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", + "prompt": "0.00000015", + "completion": "0.0000006", "image": "0", "request": "0", "webSearch": "0", @@ -24889,23 +26324,38 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", - "top_p" + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "response_format", + "structured_outputs", + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 0.1, - "outputCost": 0.1 + "inputCost": 0.15, + "outputCost": 0.6, + "throughput": 97.1715, + "latency": 651 }, { - "name": "Parasail", - "slug": "parasail", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "parasail-mistral-nemo", + "name": "GMICloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://gmicloud.ai/&size=256", + "slug": "gmiCloud", + "quantization": "bf16", + "context": 1048576, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", "pricing": { - "prompt": "0.00000011", - "completion": "0.00000011", + "prompt": "0.00000015", + "completion": "0.0000006", "image": "0", "request": "0", "webSearch": "0", @@ -24916,25 +26366,25 @@ export default { "max_tokens", "temperature", "top_p", - "presence_penalty", - "frequency_penalty", - "repetition_penalty", - "top_k" + "seed" ], - "inputCost": 0.11, - "outputCost": 0.11 + "inputCost": 0.15, + "outputCost": 0.6, + "throughput": 100.051, + "latency": 901 }, { - "name": "Mistral", - "slug": "mistral", - "quantization": "unknown", - "context": 131072, + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": null, + "context": 1048576, "maxCompletionTokens": null, - "providerModelId": "open-mistral-nemo-2407", + "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", "pricing": { - "prompt": "0.00000015", - "completion": "0.00000015", - "image": "0", + "prompt": "0.00000018", + "completion": "0.00000059", + "image": "0.00090234", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -24949,23 +26399,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "response_format", - "structured_outputs", - "seed" + "top_k", + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" ], - "inputCost": 0.15, - "outputCost": 0.15 + "inputCost": 0.18, + "outputCost": 0.59, + "throughput": 98.8715, + "latency": 492 }, { - "name": "Azure", - "slug": "azure", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "mistral-nemo", + "name": "Cerebras", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cerebras.ai/&size=256", + "slug": "cerebras", + "quantization": null, + "context": 32000, + "maxCompletionTokens": 32000, + "providerModelId": "llama-4-scout-17b-16e-instruct", "pricing": { - "prompt": "0.0000003", - "completion": "0.0000003", + "prompt": "0.00000065", + "completion": "0.00000085", "image": "0", "request": "0", "webSearch": "0", @@ -24978,156 +26433,161 @@ export default { "max_tokens", "temperature", "top_p", + "structured_outputs", "response_format", "stop", - "seed" + "seed", + "logprobs", + "top_logprobs" ], - "inputCost": 0.3, - "outputCost": 0.3 + "inputCost": 0.65, + "outputCost": 0.85, + "throughput": 7090.909, + "latency": 325 } ] }, { - "slug": "anthropic/claude-3.5-haiku-20241022", - "hfSlug": null, + "slug": "mistralai/mistral-nemo", + "hfSlug": "mistralai/Mistral-Nemo-Instruct-2407", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-04T00:00:00+00:00", + "createdAt": "2024-07-19T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3.5 Haiku (2024-10-22)", - "shortName": "Claude 3.5 Haiku (2024-10-22)", - "author": "anthropic", - "description": "Claude 3.5 Haiku features enhancements across all skill sets including coding, tool use, and reasoning. As the fastest model in the Anthropic lineup, it offers rapid response times suitable for applications that require high interactivity and low latency, such as user-facing chatbots and on-the-fly code completions. It also excels in specialized tasks like data extraction and real-time content moderation, making it a versatile tool for a broad range of industries.\n\nIt does not support image inputs.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/3-5-models-and-computer-use)", - "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", - "contextLength": 200000, + "name": "Mistral: Mistral Nemo (free)", + "shortName": "Mistral Nemo (free)", + "author": "mistralai", + "description": "A 12B parameter model with a 128k token context length built by Mistral in collaboration with NVIDIA.\n\nThe model is multilingual, supporting English, French, German, Spanish, Italian, Portuguese, Chinese, Japanese, Korean, Arabic, and Hindi.\n\nIt supports function calling and is released under the Apache 2.0 license.", + "modelVersionGroupId": null, + "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", - "instructType": null, + "group": "Mistral", + "instructType": "mistral", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "[INST]", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-5-haiku-20241022", + "permaslug": "mistralai/mistral-nemo", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "81dc58ef-b297-4540-ae41-b232d2bf5c3b", - "name": "Anthropic | anthropic/claude-3-5-haiku-20241022", - "contextLength": 200000, + "id": "d35a6a62-895e-4130-8ab6-3a7dcca5edeb", + "name": "Chutes | mistralai/mistral-nemo:free", + "contextLength": 128000, "model": { - "slug": "anthropic/claude-3.5-haiku-20241022", - "hfSlug": null, + "slug": "mistralai/mistral-nemo", + "hfSlug": "mistralai/Mistral-Nemo-Instruct-2407", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-04T00:00:00+00:00", + "createdAt": "2024-07-19T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3.5 Haiku (2024-10-22)", - "shortName": "Claude 3.5 Haiku (2024-10-22)", - "author": "anthropic", - "description": "Claude 3.5 Haiku features enhancements across all skill sets including coding, tool use, and reasoning. As the fastest model in the Anthropic lineup, it offers rapid response times suitable for applications that require high interactivity and low latency, such as user-facing chatbots and on-the-fly code completions. It also excels in specialized tasks like data extraction and real-time content moderation, making it a versatile tool for a broad range of industries.\n\nIt does not support image inputs.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/3-5-models-and-computer-use)", - "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", - "contextLength": 200000, + "name": "Mistral: Mistral Nemo", + "shortName": "Mistral Nemo", + "author": "mistralai", + "description": "A 12B parameter model with a 128k token context length built by Mistral in collaboration with NVIDIA.\n\nThe model is multilingual, supporting English, French, German, Spanish, Italian, Portuguese, Chinese, Japanese, Korean, Arabic, and Hindi.\n\nIt supports function calling and is released under the Apache 2.0 license.", + "modelVersionGroupId": null, + "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", - "instructType": null, + "group": "Mistral", + "instructType": "mistral", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "[INST]", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-5-haiku-20241022", + "permaslug": "mistralai/mistral-nemo", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "anthropic/claude-3.5-haiku-20241022", - "modelVariantPermaslug": "anthropic/claude-3-5-haiku-20241022", - "providerName": "Anthropic", + "modelVariantSlug": "mistralai/mistral-nemo:free", + "modelVariantPermaslug": "mistralai/mistral-nemo:free", + "providerName": "Chutes", "providerInfo": { - "name": "Anthropic", - "displayName": "Anthropic", - "slug": "anthropic", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "training": true, + "retainsPrompts": true + } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "Anthropic", + "moderationRequired": false, + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.anthropic.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Anthropic.svg" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "Anthropic", - "providerModelId": "claude-3-5-haiku-20241022", - "providerGroup": "Anthropic", - "quantization": "unknown", - "variant": "standard", - "isFree": false, + "providerDisplayName": "Chutes", + "providerModelId": "unsloth/Mistral-Nemo-Instruct-2407", + "providerGroup": "Chutes", + "quantization": null, + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 8192, + "maxCompletionTokens": 128000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", "top_k", - "stop" + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.0000008", - "completion": "0.000004", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", - "inputCacheRead": "0.00000008", - "inputCacheWrite": "0.000001", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -25136,7 +26596,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -25149,58 +26609,65 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 67, - "newest": 177, - "throughputHighToLow": 167, - "latencyLowToHigh": 206, - "pricingLowToHigh": 219, - "pricingHighToLow": 95 + "topWeekly": 15, + "newest": 234, + "throughputHighToLow": 57, + "latencyLowToHigh": 61, + "pricingLowToHigh": 68, + "pricingHighToLow": 236 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { - "name": "Anthropic", - "slug": "anthropic", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 8192, - "providerModelId": "claude-3-5-haiku-20241022", + "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", + "slug": "klusterAi", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "mistralai/Mistral-Nemo-Instruct-2407", "pricing": { - "prompt": "0.0000008", - "completion": "0.000004", + "prompt": "0.000000025", + "completion": "0.00000007", "image": "0", "request": "0", - "inputCacheRead": "0.00000008", - "inputCacheWrite": "0.000001", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "top_k", - "stop" + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", + "seed" ], - "inputCost": 0.8, - "outputCost": 4 + "inputCost": 0.02, + "outputCost": 0.07, + "throughput": 119.605, + "latency": 470 }, { - "name": "Google Vertex", - "slug": "vertex", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 8192, - "providerModelId": "claude-3-5-haiku@20241022", + "name": "Enfer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://enfer.ai/&size=256", + "slug": "enfer", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 65536, + "providerModelId": "mistralai/mistral-nemo", "pricing": { - "prompt": "0.0000008", - "completion": "0.000004", + "prompt": "0.00000003", + "completion": "0.00000007", "image": "0", "request": "0", - "inputCacheRead": "0.00000008", - "inputCacheWrite": "0.000001", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -25211,19 +26678,323 @@ export default { "max_tokens", "temperature", "top_p", - "top_k", - "stop" + "stop", + "frequency_penalty", + "presence_penalty", + "logit_bias", + "logprobs" ], - "inputCost": 0.8, - "outputCost": 4 - } - ] - }, - { - "slug": "google/gemma-3-12b-it", - "hfSlug": "google/gemma-3-12b-it", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-13T21:50:25.140801+00:00", + "inputCost": 0.03, + "outputCost": 0.07, + "throughput": 53.943, + "latency": 6449.5 + }, + { + "name": "NextBit", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nextbit256.com/&size=256", + "slug": "nextBit", + "quantization": "bf16", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "mistral:nemo", + "pricing": { + "prompt": "0.000000033", + "completion": "0.00000007", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "structured_outputs" + ], + "inputCost": 0.03, + "outputCost": 0.07, + "throughput": 51.1245, + "latency": 1610 + }, + { + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": 16384, + "providerModelId": "mistralai/Mistral-Nemo-Instruct-2407", + "pricing": { + "prompt": "0.000000035", + "completion": "0.00000008", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" + ], + "inputCost": 0.04, + "outputCost": 0.08, + "throughput": 55.12, + "latency": 453 + }, + { + "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", + "slug": "inferenceNet", + "quantization": "fp8", + "context": 16384, + "maxCompletionTokens": 16384, + "providerModelId": "mistralai/mistral-nemo-12b-instruct/fp-8", + "pricing": { + "prompt": "0.0000000375", + "completion": "0.0000001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "logit_bias", + "top_logprobs" + ], + "inputCost": 0.04, + "outputCost": 0.1, + "throughput": 61.076, + "latency": 1211 + }, + { + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "mistralai/Mistral-Nemo-Instruct-2407", + "pricing": { + "prompt": "0.00000004", + "completion": "0.00000012", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.04, + "outputCost": 0.12, + "throughput": 34.408, + "latency": 657 + }, + { + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": "unknown", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "mistralai/mistral-nemo", + "pricing": { + "prompt": "0.00000004", + "completion": "0.00000017", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" + ], + "inputCost": 0.04, + "outputCost": 0.17, + "throughput": 36.1535, + "latency": 1354 + }, + { + "name": "Atoma", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://atoma.network/&size=256", + "slug": "atoma", + "quantization": null, + "context": 128000, + "maxCompletionTokens": 80000, + "providerModelId": "mistralai/Mistral-Nemo-Instruct-2407", + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p" + ], + "inputCost": 0.1, + "outputCost": 0.1, + "throughput": 52.1375, + "latency": 636 + }, + { + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "parasail-mistral-nemo", + "pricing": { + "prompt": "0.00000011", + "completion": "0.00000011", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "presence_penalty", + "frequency_penalty", + "repetition_penalty", + "top_k" + ], + "inputCost": 0.11, + "outputCost": 0.11, + "throughput": 142.857, + "latency": 523 + }, + { + "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", + "slug": "mistral", + "quantization": "unknown", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "open-mistral-nemo-2407", + "pricing": { + "prompt": "0.00000015", + "completion": "0.00000015", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "structured_outputs", + "seed" + ], + "inputCost": 0.15, + "outputCost": 0.15, + "throughput": 143.466, + "latency": 212 + }, + { + "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", + "slug": "azure", + "quantization": "unknown", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "mistral-nemo", + "pricing": { + "prompt": "0.0000003", + "completion": "0.0000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "response_format", + "stop", + "seed" + ], + "inputCost": 0.3, + "outputCost": 0.3, + "throughput": 98.261, + "latency": 1152 + } + ] + }, + { + "slug": "google/gemma-3-12b-it", + "hfSlug": "google/gemma-3-12b-it", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-03-13T21:50:25.140801+00:00", "hfUpdatedAt": null, "name": "Google: Gemma 3 12B (free)", "shortName": "Gemma 3 12B (free)", @@ -25387,14 +27158,16 @@ export default { "sorting": { "topWeekly": 77, "newest": 87, - "throughputHighToLow": 176, - "latencyLowToHigh": 100, + "throughputHighToLow": 170, + "latencyLowToHigh": 196, "pricingLowToHigh": 39, "pricingHighToLow": 220 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "bf16", "context": 131072, @@ -25423,10 +27196,13 @@ export default { "min_p" ], "inputCost": 0.05, - "outputCost": 0.1 + "outputCost": 0.1, + "throughput": 23.922, + "latency": 1180 }, { "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", "slug": "cloudflare", "quantization": null, "context": 80000, @@ -25452,22 +27228,24 @@ export default { "presence_penalty" ], "inputCost": 0.35, - "outputCost": 0.56 + "outputCost": 0.56, + "throughput": 68.1855, + "latency": 595 } ] }, { - "slug": "meta-llama/llama-4-scout", - "hfSlug": "meta-llama/Llama-4-Scout-17B-16E-Instruct", - "updatedAt": "2025-04-05T19:34:05.019062+00:00", - "createdAt": "2025-04-05T19:31:59.735804+00:00", + "slug": "openai/chatgpt-4o-latest", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-08-14T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 4 Scout (free)", - "shortName": "Llama 4 Scout (free)", - "author": "meta-llama", - "description": "Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimodal input (text and image) and multilingual output (text and code) across 12 supported languages. Designed for assistant-style interaction and visual reasoning, Scout uses 16 experts per forward pass and features a context length of 10 million tokens, with a training corpus of ~40 trillion tokens.\n\nBuilt for high efficiency and local or commercial deployment, Llama 4 Scout incorporates early fusion for seamless modality integration. It is instruction-tuned for use in multilingual chat, captioning, and image understanding tasks. Released under the Llama 4 Community License, it was last trained on data up to August 2024 and launched publicly on April 5, 2025.", + "name": "OpenAI: ChatGPT-4o", + "shortName": "ChatGPT-4o", + "author": "openai", + "description": "OpenAI ChatGPT 4o is continually updated by OpenAI to point to the current version of GPT-4o used by ChatGPT. It therefore differs slightly from the API version of [GPT-4o](/models/openai/gpt-4o) in that it has additional RLHF. It is intended for research and evaluation.\n\nOpenAI notes that this model is not suited for production use-cases as it may be removed or redirected to another model in the future.", "modelVersionGroupId": null, - "contextLength": 512000, + "contextLength": 128000, "inputModalities": [ "text", "image" @@ -25476,32 +27254,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-4-scout-17b-16e-instruct", + "permaslug": "openai/chatgpt-4o-latest", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "f0b6a83b-9818-4f8c-9941-20b3ead09379", - "name": "Chutes | meta-llama/llama-4-scout-17b-16e-instruct:free", - "contextLength": 512000, + "id": "aff4b825-af10-4633-9ab2-9ac68c547988", + "name": "OpenAI | openai/chatgpt-4o-latest", + "contextLength": 128000, "model": { - "slug": "meta-llama/llama-4-scout", - "hfSlug": "meta-llama/Llama-4-Scout-17B-16E-Instruct", - "updatedAt": "2025-04-05T19:34:05.019062+00:00", - "createdAt": "2025-04-05T19:31:59.735804+00:00", + "slug": "openai/chatgpt-4o-latest", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-08-14T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 4 Scout", - "shortName": "Llama 4 Scout", - "author": "meta-llama", - "description": "Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimodal input (text and image) and multilingual output (text and code) across 12 supported languages. Designed for assistant-style interaction and visual reasoning, Scout uses 16 experts per forward pass and features a context length of 10 million tokens, with a training corpus of ~40 trillion tokens.\n\nBuilt for high efficiency and local or commercial deployment, Llama 4 Scout incorporates early fusion for seamless modality integration. It is instruction-tuned for use in multilingual chat, captioning, and image understanding tasks. Released under the Llama 4 Community License, it was last trained on data up to August 2024 and launched publicly on April 5, 2025.", + "name": "OpenAI: ChatGPT-4o", + "shortName": "ChatGPT-4o", + "author": "openai", + "description": "OpenAI ChatGPT 4o is continually updated by OpenAI to point to the current version of GPT-4o used by ChatGPT. It therefore differs slightly from the API version of [GPT-4o](/models/openai/gpt-4o) in that it has additional RLHF. It is intended for research and evaluation.\n\nOpenAI notes that this model is not suited for production use-cases as it may be removed or redirected to another model in the future.", "modelVersionGroupId": null, - "contextLength": 10000000, + "contextLength": 128000, "inputModalities": [ "text", "image" @@ -25510,90 +27288,98 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-4-scout-17b-16e-instruct", + "permaslug": "openai/chatgpt-4o-latest", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "meta-llama/llama-4-scout:free", - "modelVariantPermaslug": "meta-llama/llama-4-scout-17b-16e-instruct:free", - "providerName": "Chutes", + "modelVariantSlug": "openai/chatgpt-4o-latest", + "modelVariantPermaslug": "openai/chatgpt-4o-latest", + "providerName": "OpenAI", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { - "training": true, - "retainsPrompts": true - } + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": false, - "group": "Chutes", + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Chutes", - "providerModelId": "chutesai/Llama-4-Scout-17B-16E-Instruct", - "providerGroup": "Chutes", - "quantization": "bf16", - "variant": "free", - "isFree": true, + "providerDisplayName": "OpenAI", + "providerModelId": "chatgpt-4o-latest", + "providerGroup": "OpenAI", + "quantization": "unknown", + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": 8, + "maxCompletionTokens": 16384, + "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "structured_outputs", - "response_format", "stop", "frequency_penalty", "presence_penalty", "seed", - "top_k", - "min_p", - "repetition_penalty", - "logprobs", "logit_bias", - "top_logprobs" + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "training": true, - "retainsPrompts": true + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0", - "completion": "0", - "image": "0", + "prompt": "0.000005", + "completion": "0.000015", + "image": "0.007225", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -25611,34 +27397,32 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": { - "responseFormat": true, - "structuredOutputs": true - }, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 44, - "newest": 63, - "throughputHighToLow": 72, - "latencyLowToHigh": 172, - "pricingLowToHigh": 27, - "pricingHighToLow": 201 + "topWeekly": 78, + "newest": 220, + "throughputHighToLow": 69, + "latencyLowToHigh": 79, + "pricingLowToHigh": 292, + "pricingHighToLow": 28 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "Lambda", - "slug": "lambda", - "quantization": "fp8", - "context": 1048576, - "maxCompletionTokens": 1048576, - "providerModelId": "llama-4-scout-17b-16e-instruct", + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": "unknown", + "context": 128000, + "maxCompletionTokens": 16384, + "providerModelId": "chatgpt-4o-latest", "pricing": { - "prompt": "0.00000008", - "completion": "0.0000003", - "image": "0", + "prompt": "0.000005", + "completion": "0.000015", + "image": "0.007225", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -25655,271 +27439,202 @@ export default { "logit_bias", "logprobs", "top_logprobs", - "response_format" - ], - "inputCost": 0.08, - "outputCost": 0.3 - }, - { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", - "context": 327680, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", - "pricing": { - "prompt": "0.00000008", - "completion": "0.0000003", - "image": "0.0003342", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", "response_format", - "top_k", - "seed", - "min_p" - ], - "inputCost": 0.08, - "outputCost": 0.3 - }, - { - "name": "kluster.ai", - "slug": "klusterAi", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", - "pricing": { - "prompt": "0.00000008", - "completion": "0.00000045", - "image": "0.0005013", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature" + "structured_outputs" ], - "inputCost": 0.08, - "outputCost": 0.45 - }, - { - "name": "Parasail", - "slug": "parasail", - "quantization": "fp8", - "context": 158000, - "maxCompletionTokens": 158000, - "providerModelId": "parasail-llama-4-scout-instruct", - "pricing": { - "prompt": "0.00000009", - "completion": "0.00000048", - "image": "0.00046788", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "presence_penalty", - "frequency_penalty", - "repetition_penalty", - "top_k" + "inputCost": 5, + "outputCost": 15, + "throughput": 93.5505, + "latency": 495 + } + ] + }, + { + "slug": "cohere/command-r7b-12-2024", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-12-14T06:35:52.905418+00:00", + "hfUpdatedAt": null, + "name": "Cohere: Command R7B (12-2024)", + "shortName": "Command R7B (12-2024)", + "author": "cohere", + "description": "Command R7B (12-2024) is a small, fast update of the Command R+ model, delivered in December 2024. It excels at RAG, tool use, agents, and similar tasks requiring complex reasoning and multiple steps.\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", + "modelVersionGroupId": null, + "contextLength": 128000, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Cohere", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "cohere/command-r7b-12-2024", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "d4a57205-8124-413a-ab4e-5ae94eb7aa9d", + "name": "Cohere | cohere/command-r7b-12-2024", + "contextLength": 128000, + "model": { + "slug": "cohere/command-r7b-12-2024", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-12-14T06:35:52.905418+00:00", + "hfUpdatedAt": null, + "name": "Cohere: Command R7B (12-2024)", + "shortName": "Command R7B (12-2024)", + "author": "cohere", + "description": "Command R7B (12-2024) is a small, fast update of the Command R+ model, delivered in December 2024. It excels at RAG, tool use, agents, and similar tasks requiring complex reasoning and multiple steps.\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", + "modelVersionGroupId": null, + "contextLength": 128000, + "inputModalities": [ + "text" ], - "inputCost": 0.09, - "outputCost": 0.48 - }, - { - "name": "CentML", - "slug": "centMl", - "quantization": "bf16", - "context": 1048576, - "maxCompletionTokens": 1048576, - "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", - "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty" + "outputModalities": [ + "text" ], - "inputCost": 0.1, - "outputCost": 0.1 + "hasTextOutput": true, + "group": "Cohere", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "cohere/command-r7b-12-2024", + "reasoningConfig": null, + "features": {} }, - { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "meta-llama/llama-4-scout-17b-16e-instruct", - "pricing": { - "prompt": "0.0000001", - "completion": "0.0000005", - "image": "0.0003342", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 + "modelVariantSlug": "cohere/command-r7b-12-2024", + "modelVariantPermaslug": "cohere/command-r7b-12-2024", + "providerName": "Cohere", + "providerInfo": { + "name": "Cohere", + "displayName": "Cohere", + "slug": "cohere", + "baseUrl": "url", + "dataPolicy": { + "privacyPolicyUrl": "https://cohere.com/privacy", + "termsOfServiceUrl": "https://cohere.com/terms-of-use", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + } }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" - ], - "inputCost": 0.1, - "outputCost": 0.5 + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": false, + "isAbortable": true, + "moderationRequired": false, + "group": "Cohere", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": "https://status.cohere.ai/", + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/Cohere.png" + } }, - { - "name": "Groq", - "slug": "groq", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 8192, - "providerModelId": "meta-llama/llama-4-scout-17b-16e-instruct", - "pricing": { - "prompt": "0.00000011", - "completion": "0.00000034", - "image": "0.00036762", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 + "providerDisplayName": "Cohere", + "providerModelId": "command-r7b-12-2024", + "providerGroup": "Cohere", + "quantization": null, + "variant": "standard", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": 4000, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "seed", + "response_format", + "structured_outputs" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "privacyPolicyUrl": "https://cohere.com/privacy", + "termsOfServiceUrl": "https://cohere.com/terms-of-use", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "top_logprobs", - "logprobs", - "logit_bias", - "seed" - ], - "inputCost": 0.11, - "outputCost": 0.34 + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - { - "name": "Fireworks", - "slug": "fireworks", - "quantization": null, - "context": 1048576, - "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/llama4-scout-instruct-basic", - "pricing": { - "prompt": "0.00000015", - "completion": "0.0000006", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "response_format", - "structured_outputs", - "logit_bias", - "logprobs", - "top_logprobs" - ], - "inputCost": 0.15, - "outputCost": 0.6 + "pricing": { + "prompt": "0.0000000375", + "completion": "0.00000015", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 }, - { - "name": "GMICloud", - "slug": "gmiCloud", - "quantization": "bf16", - "context": 1048576, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", - "pricing": { - "prompt": "0.00000015", - "completion": "0.0000006", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "seed" - ], - "inputCost": 0.15, - "outputCost": 0.6 + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": false, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 79, + "newest": 155, + "throughputHighToLow": 19, + "latencyLowToHigh": 5, + "pricingLowToHigh": 93, + "pricingHighToLow": 224 + }, + "authorIcon": "https://openrouter.ai/images/icons/Cohere.png", + "providers": [ { - "name": "Together", - "slug": "together", + "name": "Cohere", + "icon": "https://openrouter.ai/images/icons/Cohere.png", + "slug": "cohere", "quantization": null, - "context": 1048576, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "context": 128000, + "maxCompletionTokens": 4000, + "providerModelId": "command-r7b-12-2024", "pricing": { - "prompt": "0.00000018", - "completion": "0.00000059", - "image": "0.00090234", + "prompt": "0.0000000375", + "completion": "0.00000015", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -25927,60 +27642,29 @@ export default { "frequency_penalty", "presence_penalty", "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" - ], - "inputCost": 0.18, - "outputCost": 0.59 - }, - { - "name": "Cerebras", - "slug": "cerebras", - "quantization": null, - "context": 32000, - "maxCompletionTokens": 32000, - "providerModelId": "llama-4-scout-17b-16e-instruct", - "pricing": { - "prompt": "0.00000065", - "completion": "0.00000085", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "structured_outputs", - "response_format", - "stop", "seed", - "logprobs", - "top_logprobs" + "response_format", + "structured_outputs" ], - "inputCost": 0.65, - "outputCost": 0.85 + "inputCost": 0.04, + "outputCost": 0.15, + "throughput": 179.93, + "latency": 167 } ] }, { - "slug": "meta-llama/llama-3-8b-instruct", - "hfSlug": "meta-llama/Meta-Llama-3-8B-Instruct", + "slug": "sao10k/l3.1-euryale-70b", + "hfSlug": "Sao10K/L3.1-70B-Euryale-v2.2", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-04-18T00:00:00+00:00", + "createdAt": "2024-08-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3 8B Instruct", - "shortName": "Llama 3 8B Instruct", - "author": "meta-llama", - "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 8B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "modelVersionGroupId": "803c32ed-9861-4abf-b5da-7d9c9e6dcf04", - "contextLength": 8192, + "name": "Sao10K: Llama 3.1 Euryale 70B v2.2", + "shortName": "Llama 3.1 Euryale 70B v2.2", + "author": "sao10k", + "description": "Euryale L3.1 70B v2.2 is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k). It is the successor of [Euryale L3 70B v2.1](/models/sao10k/l3-euryale-70b).", + "modelVersionGroupId": null, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -25998,25 +27682,25 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3-8b-instruct", + "permaslug": "sao10k/l3.1-euryale-70b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "c6e34375-c207-4d60-9a43-38b2b730815a", - "name": "DeepInfra | meta-llama/llama-3-8b-instruct", - "contextLength": 8192, + "id": "7ab4ba43-98eb-4842-9046-f7f1822ab3a2", + "name": "DeepInfra | sao10k/l3.1-euryale-70b", + "contextLength": 131072, "model": { - "slug": "meta-llama/llama-3-8b-instruct", - "hfSlug": "meta-llama/Meta-Llama-3-8B-Instruct", + "slug": "sao10k/l3.1-euryale-70b", + "hfSlug": "Sao10K/L3.1-70B-Euryale-v2.2", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-04-18T00:00:00+00:00", + "createdAt": "2024-08-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3 8B Instruct", - "shortName": "Llama 3 8B Instruct", - "author": "meta-llama", - "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 8B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "modelVersionGroupId": "803c32ed-9861-4abf-b5da-7d9c9e6dcf04", - "contextLength": 8192, + "name": "Sao10K: Llama 3.1 Euryale 70B v2.2", + "shortName": "Llama 3.1 Euryale 70B v2.2", + "author": "sao10k", + "description": "Euryale L3.1 70B v2.2 is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k). It is the successor of [Euryale L3 70B v2.1](/models/sao10k/l3-euryale-70b).", + "modelVersionGroupId": null, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -26034,12 +27718,12 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3-8b-instruct", + "permaslug": "sao10k/l3.1-euryale-70b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "meta-llama/llama-3-8b-instruct", - "modelVariantPermaslug": "meta-llama/llama-3-8b-instruct", + "modelVariantSlug": "sao10k/l3.1-euryale-70b", + "modelVariantPermaslug": "sao10k/l3.1-euryale-70b", "providerName": "DeepInfra", "providerInfo": { "name": "DeepInfra", @@ -26072,9 +27756,9 @@ export default { } }, "providerDisplayName": "DeepInfra", - "providerModelId": "meta-llama/Meta-Llama-3-8B-Instruct", + "providerModelId": "Sao10K/L3.1-70B-Euryale-v2.2", "providerGroup": "DeepInfra", - "quantization": "bf16", + "quantization": "fp8", "variant": "standard", "isFree": false, "canAbort": true, @@ -26083,8 +27767,6 @@ export default { "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -26111,8 +27793,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000003", - "completion": "0.00000006", + "prompt": "0.0000007", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -26123,7 +27805,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -26136,24 +27818,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 79, - "newest": 266, - "throughputHighToLow": 69, - "latencyLowToHigh": 39, - "pricingLowToHigh": 84, - "pricingHighToLow": 234 + "topWeekly": 80, + "newest": 216, + "throughputHighToLow": 216, + "latencyLowToHigh": 40, + "pricingLowToHigh": 197, + "pricingHighToLow": 122 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", - "quantization": "bf16", - "context": 8192, + "quantization": "fp8", + "context": 131072, "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/Meta-Llama-3-8B-Instruct", + "providerModelId": "Sao10K/L3.1-70B-Euryale-v2.2", "pricing": { - "prompt": "0.00000003", - "completion": "0.00000006", + "prompt": "0.0000007", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -26161,8 +27845,6 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -26175,24 +27857,27 @@ export default { "seed", "min_p" ], - "inputCost": 0.03, - "outputCost": 0.06 + "inputCost": 0.7, + "outputCost": 0.8, + "throughput": 38.4885, + "latency": 481 }, { - "name": "Mancer", - "slug": "mancer", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", "quantization": "unknown", - "context": 16384, - "maxCompletionTokens": 2048, - "providerModelId": "llama3-8b", + "context": 8192, + "maxCompletionTokens": 8192, + "providerModelId": "sao10k/l31-70b-euryale-v2.2", "pricing": { - "prompt": "0.0000000375", - "completion": "0.0000001875", + "prompt": "0.00000148", + "completion": "0.00000148", "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", - "discount": 0.25 + "discount": 0 }, "supportedParameters": [ "max_tokens", @@ -26201,26 +27886,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "logit_bias", + "seed", "top_k", "min_p", - "seed", - "top_a" + "repetition_penalty", + "logit_bias" ], - "inputCost": 0.04, - "outputCost": 0.19 + "inputCost": 1.48, + "outputCost": 1.48, + "throughput": 34.1995, + "latency": 1462 }, { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": "unknown", - "context": 8192, - "maxCompletionTokens": 8192, - "providerModelId": "meta-llama/llama-3-8b-instruct", + "name": "Infermatic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://infermatic.ai/&size=256", + "slug": "infermatic", + "quantization": "fp8", + "context": 16000, + "maxCompletionTokens": null, + "providerModelId": "L3.1-70B-Euryale-v2.2-FP8-Dynamic", "pricing": { - "prompt": "0.00000004", - "completion": "0.00000004", + "prompt": "0.0000015", + "completion": "0.0000015", "image": "0", "request": "0", "webSearch": "0", @@ -26234,411 +27921,16 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", + "repetition_penalty", + "logit_bias", "top_k", "min_p", - "repetition_penalty", - "logit_bias" - ], - "inputCost": 0.04, - "outputCost": 0.04 - }, - { - "name": "Groq", - "slug": "groq", - "quantization": null, - "context": 8192, - "maxCompletionTokens": 8192, - "providerModelId": "llama3-8b-8192", - "pricing": { - "prompt": "0.00000005", - "completion": "0.00000008", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "top_logprobs", - "logprobs", - "logit_bias", - "seed" - ], - "inputCost": 0.05, - "outputCost": 0.08 - }, - { - "name": "Mancer (private)", - "slug": "mancer (private)", - "quantization": "unknown", - "context": 16384, - "maxCompletionTokens": 2048, - "providerModelId": "llama3-8b", - "pricing": { - "prompt": "0.00000005", - "completion": "0.00000025", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "logit_bias", - "top_k", - "min_p", - "seed", - "top_a" - ], - "inputCost": 0.05, - "outputCost": 0.25 - }, - { - "name": "Together", - "slug": "together", - "quantization": "int4", - "context": 8192, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Meta-Llama-3-8B-Instruct-Lite", - "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" - ], - "inputCost": 0.1, - "outputCost": 0.1 - }, - { - "name": "Cloudflare", - "slug": "cloudflare", - "quantization": null, - "context": 7968, - "maxCompletionTokens": null, - "providerModelId": "@cf/meta/llama-3-8b-instruct", - "pricing": { - "prompt": "0.00000028", - "completion": "0.00000083", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "top_k", - "seed", - "repetition_penalty", - "frequency_penalty", - "presence_penalty" - ], - "inputCost": 0.28, - "outputCost": 0.83 - } - ] - }, - { - "slug": "sao10k/l3.1-euryale-70b", - "hfSlug": "Sao10K/L3.1-70B-Euryale-v2.2", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-28T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Sao10K: Llama 3.1 Euryale 70B v2.2", - "shortName": "Llama 3.1 Euryale 70B v2.2", - "author": "sao10k", - "description": "Euryale L3.1 70B v2.2 is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k). It is the successor of [Euryale L3 70B v2.1](/models/sao10k/l3-euryale-70b).", - "modelVersionGroupId": null, - "contextLength": 131072, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", - "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "sao10k/l3.1-euryale-70b", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "7ab4ba43-98eb-4842-9046-f7f1822ab3a2", - "name": "DeepInfra | sao10k/l3.1-euryale-70b", - "contextLength": 131072, - "model": { - "slug": "sao10k/l3.1-euryale-70b", - "hfSlug": "Sao10K/L3.1-70B-Euryale-v2.2", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-28T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Sao10K: Llama 3.1 Euryale 70B v2.2", - "shortName": "Llama 3.1 Euryale 70B v2.2", - "author": "sao10k", - "description": "Euryale L3.1 70B v2.2 is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k). It is the successor of [Euryale L3 70B v2.1](/models/sao10k/l3-euryale-70b).", - "modelVersionGroupId": null, - "contextLength": 131072, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", - "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "sao10k/l3.1-euryale-70b", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "sao10k/l3.1-euryale-70b", - "modelVariantPermaslug": "sao10k/l3.1-euryale-70b", - "providerName": "DeepInfra", - "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", - "baseUrl": "url", - "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", - "paidModels": { - "training": false, - "retainsPrompts": false - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "DeepInfra", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/DeepInfra.webp" - } - }, - "providerDisplayName": "DeepInfra", - "providerModelId": "Sao10K/L3.1-70B-Euryale-v2.2", - "providerGroup": "DeepInfra", - "quantization": "fp8", - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 16384, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", - "paidModels": { - "training": false, - "retainsPrompts": false - }, - "training": false, - "retainsPrompts": false - }, - "pricing": { - "prompt": "0.0000007", - "completion": "0.0000008", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 80, - "newest": 216, - "throughputHighToLow": 218, - "latencyLowToHigh": 70, - "pricingLowToHigh": 197, - "pricingHighToLow": 122 - }, - "providers": [ - { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 16384, - "providerModelId": "Sao10K/L3.1-70B-Euryale-v2.2", - "pricing": { - "prompt": "0.0000007", - "completion": "0.0000008", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "inputCost": 0.7, - "outputCost": 0.8 - }, - { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": "unknown", - "context": 8192, - "maxCompletionTokens": 8192, - "providerModelId": "sao10k/l31-70b-euryale-v2.2", - "pricing": { - "prompt": "0.00000148", - "completion": "0.00000148", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" - ], - "inputCost": 1.48, - "outputCost": 1.48 - }, - { - "name": "Infermatic", - "slug": "infermatic", - "quantization": "fp8", - "context": 16000, - "maxCompletionTokens": null, - "providerModelId": "L3.1-70B-Euryale-v2.2-FP8-Dynamic", - "pricing": { - "prompt": "0.0000015", - "completion": "0.0000015", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "logit_bias", - "top_k", - "min_p", - "seed" + "seed" ], "inputCost": 1.5, - "outputCost": 1.5 + "outputCost": 1.5, + "throughput": 21.6565, + "latency": 1640.5 } ] }, @@ -26780,8 +28072,8 @@ export default { "retainsPrompts": true }, "pricing": { - "prompt": "0.00000009375", - "completion": "0.00000075", + "prompt": "0.00000015", + "completion": "0.0000009375", "image": "0", "request": "0", "webSearch": "0", @@ -26807,22 +28099,24 @@ export default { "sorting": { "topWeekly": 81, "newest": 263, - "throughputHighToLow": 148, - "latencyLowToHigh": 101, - "pricingLowToHigh": 136, - "pricingHighToLow": 182 + "throughputHighToLow": 144, + "latencyLowToHigh": 95, + "pricingLowToHigh": 152, + "pricingHighToLow": 166 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "Mancer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", "slug": "mancer", "quantization": "unknown", "context": 24576, "maxCompletionTokens": 2048, "providerModelId": "lumi-8b", "pricing": { - "prompt": "0.00000009375", - "completion": "0.00000075", + "prompt": "0.00000015", + "completion": "0.0000009375", "image": "0", "request": "0", "webSearch": "0", @@ -26843,19 +28137,22 @@ export default { "seed", "top_a" ], - "inputCost": 0.09, - "outputCost": 0.75 + "inputCost": 0.15, + "outputCost": 0.94, + "throughput": 58.175, + "latency": 642 }, { "name": "Mancer (private)", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", "slug": "mancer (private)", "quantization": "unknown", "context": 24576, "maxCompletionTokens": 2048, "providerModelId": "lumi-8b", "pricing": { - "prompt": "0.000000125", - "completion": "0.000001", + "prompt": "0.0000002", + "completion": "0.00000125", "image": "0", "request": "0", "webSearch": "0", @@ -26876,11 +28173,14 @@ export default { "seed", "top_a" ], - "inputCost": 0.13, - "outputCost": 1 + "inputCost": 0.2, + "outputCost": 1.25, + "throughput": 60.202, + "latency": 435 }, { "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", "slug": "featherless", "quantization": "fp8", "context": 8192, @@ -26908,22 +28208,24 @@ export default { "seed" ], "inputCost": 0.8, - "outputCost": 1.2 + "outputCost": 1.2, + "throughput": 25.6735, + "latency": 1299.5 } ] }, { - "slug": "qwen/qwen-2.5-coder-32b-instruct", - "hfSlug": "Qwen/Qwen2.5-Coder-32B-Instruct", + "slug": "meta-llama/llama-3-8b-instruct", + "hfSlug": "meta-llama/Meta-Llama-3-8B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-11T23:40:00.276653+00:00", + "createdAt": "2024-04-18T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen2.5 Coder 32B Instruct", - "shortName": "Qwen2.5 Coder 32B Instruct", - "author": "qwen", - "description": "Qwen2.5-Coder is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen). Qwen2.5-Coder brings the following improvements upon CodeQwen1.5:\n\n- Significantly improvements in **code generation**, **code reasoning** and **code fixing**. \n- A more comprehensive foundation for real-world applications such as **Code Agents**. Not only enhancing coding capabilities but also maintaining its strengths in mathematics and general competencies.\n\nTo read more about its evaluation results, check out [Qwen 2.5 Coder's blog](https://qwenlm.github.io/blog/qwen2.5-coder-family/).", - "modelVersionGroupId": null, - "contextLength": 32768, + "name": "Meta: Llama 3 8B Instruct", + "shortName": "Llama 3 8B Instruct", + "author": "meta-llama", + "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 8B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "modelVersionGroupId": "803c32ed-9861-4abf-b5da-7d9c9e6dcf04", + "contextLength": 8192, "inputModalities": [ "text" ], @@ -26931,36 +28233,35 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" + "<|eot_id|>", + "<|end_of_text|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-2.5-coder-32b-instruct", + "permaslug": "meta-llama/llama-3-8b-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "48e34779-1643-4717-8c51-9a9e02f4b993", - "name": "DeepInfra | qwen/qwen-2.5-coder-32b-instruct", - "contextLength": 32768, + "id": "c6e34375-c207-4d60-9a43-38b2b730815a", + "name": "DeepInfra | meta-llama/llama-3-8b-instruct", + "contextLength": 8192, "model": { - "slug": "qwen/qwen-2.5-coder-32b-instruct", - "hfSlug": "Qwen/Qwen2.5-Coder-32B-Instruct", + "slug": "meta-llama/llama-3-8b-instruct", + "hfSlug": "meta-llama/Meta-Llama-3-8B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-11T23:40:00.276653+00:00", + "createdAt": "2024-04-18T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen2.5 Coder 32B Instruct", - "shortName": "Qwen2.5 Coder 32B Instruct", - "author": "qwen", - "description": "Qwen2.5-Coder is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen). Qwen2.5-Coder brings the following improvements upon CodeQwen1.5:\n\n- Significantly improvements in **code generation**, **code reasoning** and **code fixing**. \n- A more comprehensive foundation for real-world applications such as **Code Agents**. Not only enhancing coding capabilities but also maintaining its strengths in mathematics and general competencies.\n\nTo read more about its evaluation results, check out [Qwen 2.5 Coder's blog](https://qwenlm.github.io/blog/qwen2.5-coder-family/).", - "modelVersionGroupId": null, - "contextLength": 128000, + "name": "Meta: Llama 3 8B Instruct", + "shortName": "Llama 3 8B Instruct", + "author": "meta-llama", + "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 8B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "modelVersionGroupId": "803c32ed-9861-4abf-b5da-7d9c9e6dcf04", + "contextLength": 8192, "inputModalities": [ "text" ], @@ -26968,23 +28269,22 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" + "<|eot_id|>", + "<|end_of_text|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-2.5-coder-32b-instruct", + "permaslug": "meta-llama/llama-3-8b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "qwen/qwen-2.5-coder-32b-instruct", - "modelVariantPermaslug": "qwen/qwen-2.5-coder-32b-instruct", + "modelVariantSlug": "meta-llama/llama-3-8b-instruct", + "modelVariantPermaslug": "meta-llama/llama-3-8b-instruct", "providerName": "DeepInfra", "providerInfo": { "name": "DeepInfra", @@ -27017,9 +28317,9 @@ export default { } }, "providerDisplayName": "DeepInfra", - "providerModelId": "Qwen/Qwen2.5-Coder-32B-Instruct", + "providerModelId": "meta-llama/Meta-Llama-3-8B-Instruct", "providerGroup": "DeepInfra", - "quantization": "fp8", + "quantization": "bf16", "variant": "standard", "isFree": false, "canAbort": true, @@ -27028,6 +28328,8 @@ export default { "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", @@ -27054,8 +28356,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000006", - "completion": "0.00000015", + "prompt": "0.00000003", + "completion": "0.00000006", "image": "0", "request": "0", "webSearch": "0", @@ -27066,7 +28368,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -27080,23 +28382,25 @@ export default { }, "sorting": { "topWeekly": 82, - "newest": 172, - "throughputHighToLow": 164, - "latencyLowToHigh": 31, - "pricingLowToHigh": 59, - "pricingHighToLow": 215 + "newest": 266, + "throughputHighToLow": 56, + "latencyLowToHigh": 47, + "pricingLowToHigh": 85, + "pricingHighToLow": 233 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", - "quantization": "fp8", - "context": 32768, + "quantization": "bf16", + "context": 8192, "maxCompletionTokens": 16384, - "providerModelId": "Qwen/Qwen2.5-Coder-32B-Instruct", + "providerModelId": "meta-llama/Meta-Llama-3-8B-Instruct", "pricing": { - "prompt": "0.00000006", - "completion": "0.00000015", + "prompt": "0.00000003", + "completion": "0.00000006", "image": "0", "request": "0", "webSearch": "0", @@ -27104,6 +28408,8 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", @@ -27116,24 +28422,27 @@ export default { "seed", "min_p" ], - "inputCost": 0.06, - "outputCost": 0.15 + "inputCost": 0.03, + "outputCost": 0.06, + "throughput": 118.1595, + "latency": 414 }, { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen2.5-Coder-32B-Instruct", + "name": "Mancer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", + "slug": "mancer", + "quantization": "unknown", + "context": 16384, + "maxCompletionTokens": 2048, + "providerModelId": "llama3-8b", "pricing": { - "prompt": "0.00000006", - "completion": "0.00000018", + "prompt": "0.0000000375", + "completion": "0.0000001875", "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", - "discount": 0 + "discount": 0.25 }, "supportedParameters": [ "max_tokens", @@ -27142,57 +28451,29 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", + "repetition_penalty", "logit_bias", - "logprobs", - "top_logprobs" - ], - "inputCost": 0.06, - "outputCost": 0.18 - }, - { - "name": "Lambda", - "slug": "lambda", - "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": 32768, - "providerModelId": "qwen25-coder-32b-instruct", - "pricing": { - "prompt": "0.00000007", - "completion": "0.00000016", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", + "top_k", + "min_p", "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format" + "top_a" ], - "inputCost": 0.07, - "outputCost": 0.16 + "inputCost": 0.04, + "outputCost": 0.19, + "throughput": 62.785, + "latency": 490 }, { - "name": "Hyperbolic", - "slug": "hyperbolic", - "quantization": "fp8", - "context": 32768, + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": "unknown", + "context": 8192, "maxCompletionTokens": 8192, - "providerModelId": "Qwen/Qwen2.5-Coder-32B-Instruct", + "providerModelId": "meta-llama/llama-3-8b-instruct", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", + "prompt": "0.00000004", + "completion": "0.00000004", "image": "0", "request": "0", "webSearch": "0", @@ -27206,60 +28487,33 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "logprobs", - "top_logprobs", "seed", - "logit_bias", "top_k", "min_p", - "repetition_penalty" - ], - "inputCost": 0.2, - "outputCost": 0.2 - }, - { - "name": "Parasail", - "slug": "parasail", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "parasail-qwen-coder32b-longcontext-128", - "pricing": { - "prompt": "0.00000035", - "completion": "0.0000005", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "presence_penalty", - "frequency_penalty", "repetition_penalty", - "top_k" + "logit_bias" ], - "inputCost": 0.35, - "outputCost": 0.5 + "inputCost": 0.04, + "outputCost": 0.04, + "throughput": 62.3095, + "latency": 853 }, { - "name": "Mancer", - "slug": "mancer", - "quantization": "fp8", - "context": 32768, - "maxCompletionTokens": 2048, - "providerModelId": "qwen2.5-coder-32b", + "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", + "slug": "groq", + "quantization": null, + "context": 8192, + "maxCompletionTokens": 8192, + "providerModelId": "llama3-8b-8192", "pricing": { - "prompt": "0.000000375", - "completion": "0.0000015", + "prompt": "0.00000005", + "completion": "0.00000008", "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", - "discount": 0.25 + "discount": 0 }, "supportedParameters": [ "max_tokens", @@ -27268,26 +28522,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", + "response_format", + "top_logprobs", + "logprobs", "logit_bias", - "top_k", - "min_p", - "seed", - "top_a" + "seed" ], - "inputCost": 0.38, - "outputCost": 1.5 + "inputCost": 0.05, + "outputCost": 0.08, + "throughput": 3259.259, + "latency": 448 }, { "name": "Mancer (private)", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", "slug": "mancer (private)", - "quantization": "fp8", - "context": 32768, + "quantization": "unknown", + "context": 16384, "maxCompletionTokens": 2048, - "providerModelId": "qwen2.5-coder-32b", + "providerModelId": "llama3-8b", "pricing": { - "prompt": "0.0000005", - "completion": "0.000002", + "prompt": "0.00000005", + "completion": "0.00000025", "image": "0", "request": "0", "webSearch": "0", @@ -27308,48 +28564,22 @@ export default { "seed", "top_a" ], - "inputCost": 0.5, - "outputCost": 2 - }, - { - "name": "Cloudflare", - "slug": "cloudflare", - "quantization": null, - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "@cf/qwen/qwen2.5-coder-32b-instruct", - "pricing": { - "prompt": "0.00000066", - "completion": "0.000001", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "top_k", - "seed", - "repetition_penalty", - "frequency_penalty", - "presence_penalty" - ], - "inputCost": 0.66, - "outputCost": 1 + "inputCost": 0.05, + "outputCost": 0.25, + "throughput": 59.815, + "latency": 483 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", - "quantization": "fp16", - "context": 16384, - "maxCompletionTokens": 2048, - "providerModelId": "Qwen/Qwen2.5-Coder-32B-Instruct", + "quantization": "int4", + "context": 8192, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Meta-Llama-3-8B-Instruct-Lite", "pricing": { - "prompt": "0.0000008", - "completion": "0.0000008", + "prompt": "0.0000001", + "completion": "0.0000001", "image": "0", "request": "0", "webSearch": "0", @@ -27369,19 +28599,22 @@ export default { "min_p", "response_format" ], - "inputCost": 0.8, - "outputCost": 0.8 + "inputCost": 0.1, + "outputCost": 0.1, + "throughput": 206.491, + "latency": 369 }, { - "name": "Featherless", - "slug": "featherless", - "quantization": "fp8", - "context": 16384, - "maxCompletionTokens": 4096, - "providerModelId": "Qwen/Qwen2.5-Coder-32B-Instruct", + "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", + "slug": "cloudflare", + "quantization": null, + "context": 7968, + "maxCompletionTokens": null, + "providerModelId": "@cf/meta/llama-3-8b-instruct", "pricing": { - "prompt": "0.0000026", - "completion": "0.0000034", + "prompt": "0.00000028", + "completion": "0.00000083", "image": "0", "request": "0", "webSearch": "0", @@ -27392,16 +28625,16 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", "top_k", - "min_p", - "seed" + "seed", + "repetition_penalty", + "frequency_penalty", + "presence_penalty" ], - "inputCost": 2.6, - "outputCost": 3.4 + "inputCost": 0.28, + "outputCost": 0.83, + "throughput": 15.618, + "latency": 835 } ] }, @@ -27556,14 +28789,16 @@ export default { "sorting": { "topWeekly": 83, "newest": 145, - "throughputHighToLow": 259, - "latencyLowToHigh": 211, + "throughputHighToLow": 254, + "latencyLowToHigh": 217, "pricingLowToHigh": 163, "pricingHighToLow": 155 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://minimaxi.com/\\u0026size=256", "providers": [ { "name": "Minimax", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://minimaxi.com/&size=256", "slug": "minimax", "quantization": null, "context": 1000192, @@ -27584,7 +28819,248 @@ export default { "top_p" ], "inputCost": 0.2, - "outputCost": 1.1 + "outputCost": 1.1, + "throughput": 27.0455, + "latency": 1540 + } + ] + }, + { + "slug": "anthropic/claude-3.5-haiku-20241022", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-11-04T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Anthropic: Claude 3.5 Haiku (2024-10-22)", + "shortName": "Claude 3.5 Haiku (2024-10-22)", + "author": "anthropic", + "description": "Claude 3.5 Haiku features enhancements across all skill sets including coding, tool use, and reasoning. As the fastest model in the Anthropic lineup, it offers rapid response times suitable for applications that require high interactivity and low latency, such as user-facing chatbots and on-the-fly code completions. It also excels in specialized tasks like data extraction and real-time content moderation, making it a versatile tool for a broad range of industries.\n\nIt does not support image inputs.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/3-5-models-and-computer-use)", + "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", + "contextLength": 200000, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Claude", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "anthropic/claude-3-5-haiku-20241022", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "81dc58ef-b297-4540-ae41-b232d2bf5c3b", + "name": "Anthropic | anthropic/claude-3-5-haiku-20241022", + "contextLength": 200000, + "model": { + "slug": "anthropic/claude-3.5-haiku-20241022", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-11-04T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Anthropic: Claude 3.5 Haiku (2024-10-22)", + "shortName": "Claude 3.5 Haiku (2024-10-22)", + "author": "anthropic", + "description": "Claude 3.5 Haiku features enhancements across all skill sets including coding, tool use, and reasoning. As the fastest model in the Anthropic lineup, it offers rapid response times suitable for applications that require high interactivity and low latency, such as user-facing chatbots and on-the-fly code completions. It also excels in specialized tasks like data extraction and real-time content moderation, making it a versatile tool for a broad range of industries.\n\nIt does not support image inputs.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/3-5-models-and-computer-use)", + "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", + "contextLength": 200000, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Claude", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "anthropic/claude-3-5-haiku-20241022", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "anthropic/claude-3.5-haiku-20241022", + "modelVariantPermaslug": "anthropic/claude-3-5-haiku-20241022", + "providerName": "Anthropic", + "providerInfo": { + "name": "Anthropic", + "displayName": "Anthropic", + "slug": "anthropic", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": true, + "group": "Anthropic", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": "https://status.anthropic.com/", + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/Anthropic.svg" + } + }, + "providerDisplayName": "Anthropic", + "providerModelId": "claude-3-5-haiku-20241022", + "providerGroup": "Anthropic", + "quantization": "unknown", + "variant": "standard", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": 8192, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "isByok": false, + "moderationRequired": true, + "dataPolicy": { + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "pricing": { + "prompt": "0.0000008", + "completion": "0.000004", + "image": "0", + "request": "0", + "inputCacheRead": "0.00000008", + "inputCacheWrite": "0.000001", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": true, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 67, + "newest": 179, + "throughputHighToLow": 163, + "latencyLowToHigh": 221, + "pricingLowToHigh": 219, + "pricingHighToLow": 95 + }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", + "providers": [ + { + "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", + "slug": "anthropic", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 8192, + "providerModelId": "claude-3-5-haiku-20241022", + "pricing": { + "prompt": "0.0000008", + "completion": "0.000004", + "image": "0", + "request": "0", + "inputCacheRead": "0.00000008", + "inputCacheWrite": "0.000001", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 0.8, + "outputCost": 4, + "throughput": 52.1635, + "latency": 1537 + }, + { + "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", + "slug": "vertex", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 8192, + "providerModelId": "claude-3-5-haiku@20241022", + "pricing": { + "prompt": "0.0000008", + "completion": "0.000004", + "image": "0", + "request": "0", + "inputCacheRead": "0.00000008", + "inputCacheWrite": "0.000001", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 0.8, + "outputCost": 4, + "throughput": 69.812, + "latency": 2450 } ] }, @@ -27749,16 +29225,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 84, + "topWeekly": 85, "newest": 140, - "throughputHighToLow": 278, - "latencyLowToHigh": 159, + "throughputHighToLow": 258, + "latencyLowToHigh": 137, "pricingLowToHigh": 75, "pricingHighToLow": 243 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://www.liquid.ai/\\u0026size=256", "providers": [ { "name": "Liquid", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.liquid.ai/&size=256", "slug": "liquid", "quantization": null, "context": 32768, @@ -27786,292 +29264,85 @@ export default { "repetition_penalty" ], "inputCost": 0.02, - "outputCost": 0.02 + "outputCost": 0.02, + "throughput": 23.455, + "latency": 860 } ] }, { - "slug": "openai/chatgpt-4o-latest", - "hfSlug": null, + "slug": "qwen/qwen-2.5-coder-32b-instruct", + "hfSlug": "Qwen/Qwen2.5-Coder-32B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-14T00:00:00+00:00", + "createdAt": "2024-11-11T23:40:00.276653+00:00", "hfUpdatedAt": null, - "name": "OpenAI: ChatGPT-4o", - "shortName": "ChatGPT-4o", - "author": "openai", - "description": "OpenAI ChatGPT 4o is continually updated by OpenAI to point to the current version of GPT-4o used by ChatGPT. It therefore differs slightly from the API version of [GPT-4o](/models/openai/gpt-4o) in that it has additional RLHF. It is intended for research and evaluation.\n\nOpenAI notes that this model is not suited for production use-cases as it may be removed or redirected to another model in the future.", + "name": "Qwen2.5 Coder 32B Instruct", + "shortName": "Qwen2.5 Coder 32B Instruct", + "author": "qwen", + "description": "Qwen2.5-Coder is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen). Qwen2.5-Coder brings the following improvements upon CodeQwen1.5:\n\n- Significantly improvements in **code generation**, **code reasoning** and **code fixing**. \n- A more comprehensive foundation for real-world applications such as **Code Agents**. Not only enhancing coding capabilities but also maintaining its strengths in mathematics and general competencies.\n\nTo read more about its evaluation results, check out [Qwen 2.5 Coder's blog](https://qwenlm.github.io/blog/qwen2.5-coder-family/).", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 32768, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, + "group": "Qwen", + "instructType": "chatml", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/chatgpt-4o-latest", + "permaslug": "qwen/qwen-2.5-coder-32b-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "aff4b825-af10-4633-9ab2-9ac68c547988", - "name": "OpenAI | openai/chatgpt-4o-latest", - "contextLength": 128000, + "id": "48e34779-1643-4717-8c51-9a9e02f4b993", + "name": "DeepInfra | qwen/qwen-2.5-coder-32b-instruct", + "contextLength": 32768, "model": { - "slug": "openai/chatgpt-4o-latest", - "hfSlug": null, + "slug": "qwen/qwen-2.5-coder-32b-instruct", + "hfSlug": "Qwen/Qwen2.5-Coder-32B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-14T00:00:00+00:00", + "createdAt": "2024-11-11T23:40:00.276653+00:00", "hfUpdatedAt": null, - "name": "OpenAI: ChatGPT-4o", - "shortName": "ChatGPT-4o", - "author": "openai", - "description": "OpenAI ChatGPT 4o is continually updated by OpenAI to point to the current version of GPT-4o used by ChatGPT. It therefore differs slightly from the API version of [GPT-4o](/models/openai/gpt-4o) in that it has additional RLHF. It is intended for research and evaluation.\n\nOpenAI notes that this model is not suited for production use-cases as it may be removed or redirected to another model in the future.", + "name": "Qwen2.5 Coder 32B Instruct", + "shortName": "Qwen2.5 Coder 32B Instruct", + "author": "qwen", + "description": "Qwen2.5-Coder is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen). Qwen2.5-Coder brings the following improvements upon CodeQwen1.5:\n\n- Significantly improvements in **code generation**, **code reasoning** and **code fixing**. \n- A more comprehensive foundation for real-world applications such as **Code Agents**. Not only enhancing coding capabilities but also maintaining its strengths in mathematics and general competencies.\n\nTo read more about its evaluation results, check out [Qwen 2.5 Coder's blog](https://qwenlm.github.io/blog/qwen2.5-coder-family/).", "modelVersionGroupId": null, "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "openai/chatgpt-4o-latest", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "openai/chatgpt-4o-latest", - "modelVariantPermaslug": "openai/chatgpt-4o-latest", - "providerName": "OpenAI", - "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true - } - }, - "providerDisplayName": "OpenAI", - "providerModelId": "chatgpt-4o-latest", - "providerGroup": "OpenAI", - "quantization": "unknown", - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 16384, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" - ], - "isByok": false, - "moderationRequired": true, - "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "pricing": { - "prompt": "0.000005", - "completion": "0.000015", - "image": "0.007225", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 85, - "newest": 220, - "throughputHighToLow": 89, - "latencyLowToHigh": 62, - "pricingLowToHigh": 292, - "pricingHighToLow": 28 - }, - "providers": [ - { - "name": "OpenAI", - "slug": "openAi", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 16384, - "providerModelId": "chatgpt-4o-latest", - "pricing": { - "prompt": "0.000005", - "completion": "0.000015", - "image": "0.007225", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" - ], - "inputCost": 5, - "outputCost": 15 - } - ] - }, - { - "slug": "meta-llama/llama-3.1-405b-instruct", - "hfSlug": "meta-llama/Meta-Llama-3.1-405B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-07-23T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Meta: Llama 3.1 405B Instruct", - "shortName": "Llama 3.1 405B Instruct", - "author": "meta-llama", - "description": "The highly anticipated 400B class of Llama3 is here! Clocking in at 128k context with impressive eval scores, the Meta AI team continues to push the frontier of open-source LLMs.\n\nMeta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 405B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models including GPT-4o and Claude 3.5 Sonnet in evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "modelVersionGroupId": "1fd9d06b-aa20-4c7d-a0b1-d3d9b5aae712", - "contextLength": 32768, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", - "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "meta-llama/llama-3.1-405b-instruct", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "ea74be37-05ac-4934-9734-709cef644954", - "name": "DeepInfra | meta-llama/llama-3.1-405b-instruct", - "contextLength": 32768, - "model": { - "slug": "meta-llama/llama-3.1-405b-instruct", - "hfSlug": "meta-llama/Meta-Llama-3.1-405B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-07-23T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Meta: Llama 3.1 405B Instruct", - "shortName": "Llama 3.1 405B Instruct", - "author": "meta-llama", - "description": "The highly anticipated 400B class of Llama3 is here! Clocking in at 128k context with impressive eval scores, the Meta AI team continues to push the frontier of open-source LLMs.\n\nMeta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 405B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models including GPT-4o and Claude 3.5 Sonnet in evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "modelVersionGroupId": "1fd9d06b-aa20-4c7d-a0b1-d3d9b5aae712", - "contextLength": 131072, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Qwen", + "instructType": "chatml", "defaultSystem": null, "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.1-405b-instruct", + "permaslug": "qwen/qwen-2.5-coder-32b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "meta-llama/llama-3.1-405b-instruct", - "modelVariantPermaslug": "meta-llama/llama-3.1-405b-instruct", + "modelVariantSlug": "qwen/qwen-2.5-coder-32b-instruct", + "modelVariantPermaslug": "qwen/qwen-2.5-coder-32b-instruct", "providerName": "DeepInfra", "providerInfo": { "name": "DeepInfra", @@ -28104,7 +29375,7 @@ export default { } }, "providerDisplayName": "DeepInfra", - "providerModelId": "meta-llama/Meta-Llama-3.1-405B-Instruct", + "providerModelId": "Qwen/Qwen2.5-Coder-32B-Instruct", "providerGroup": "DeepInfra", "quantization": "fp8", "variant": "standard", @@ -28115,8 +29386,6 @@ export default { "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -28143,8 +29412,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.0000008", - "completion": "0.0000008", + "prompt": "0.00000006", + "completion": "0.00000015", "image": "0", "request": "0", "webSearch": "0", @@ -28155,7 +29424,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -28169,23 +29438,25 @@ export default { }, "sorting": { "topWeekly": 86, - "newest": 232, - "throughputHighToLow": 260, - "latencyLowToHigh": 110, - "pricingLowToHigh": 201, - "pricingHighToLow": 117 + "newest": 172, + "throughputHighToLow": 159, + "latencyLowToHigh": 66, + "pricingLowToHigh": 59, + "pricingHighToLow": 215 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", "context": 32768, "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/Meta-Llama-3.1-405B-Instruct", + "providerModelId": "Qwen/Qwen2.5-Coder-32B-Instruct", "pricing": { - "prompt": "0.0000008", - "completion": "0.0000008", + "prompt": "0.00000006", + "completion": "0.00000015", "image": "0", "request": "0", "webSearch": "0", @@ -28193,8 +29464,6 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -28207,19 +29476,57 @@ export default { "seed", "min_p" ], - "inputCost": 0.8, - "outputCost": 0.8 + "inputCost": 0.06, + "outputCost": 0.15, + "throughput": 53.254, + "latency": 403 }, { - "name": "Lambda", - "slug": "lambda", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", "quantization": "fp8", "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "llama3.1-405b-instruct-fp8", + "maxCompletionTokens": null, + "providerModelId": "Qwen/Qwen2.5-Coder-32B-Instruct", "pricing": { - "prompt": "0.0000008", - "completion": "0.0000008", + "prompt": "0.00000006", + "completion": "0.00000018", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.06, + "outputCost": 0.18, + "throughput": 52.075, + "latency": 464 + }, + { + "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", + "slug": "lambda", + "quantization": "bf16", + "context": 32768, + "maxCompletionTokens": 32768, + "providerModelId": "qwen25-coder-32b-instruct", + "pricing": { + "prompt": "0.00000007", + "completion": "0.00000016", "image": "0", "request": "0", "webSearch": "0", @@ -28239,19 +29546,22 @@ export default { "top_logprobs", "response_format" ], - "inputCost": 0.8, - "outputCost": 0.8 + "inputCost": 0.07, + "outputCost": 0.16, + "throughput": 44.453, + "latency": 502 }, { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", + "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", + "slug": "hyperbolic", "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Meta-Llama-3.1-405B-Instruct", + "context": 32768, + "maxCompletionTokens": 8192, + "providerModelId": "Qwen/Qwen2.5-Coder-32B-Instruct", "pricing": { - "prompt": "0.000001", - "completion": "0.000003", + "prompt": "0.0000002", + "completion": "0.0000002", "image": "0", "request": "0", "webSearch": "0", @@ -28265,25 +29575,30 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "logprobs", + "top_logprobs", "seed", - "top_k", "logit_bias", - "logprobs", - "top_logprobs" + "top_k", + "min_p", + "repetition_penalty" ], - "inputCost": 1, - "outputCost": 3 + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 53.427, + "latency": 1119 }, { - "name": "Fireworks", - "slug": "fireworks", + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", "quantization": "fp8", "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/llama-v3p1-405b-instruct", + "maxCompletionTokens": 131072, + "providerModelId": "parasail-qwen-coder32b-longcontext-128", "pricing": { - "prompt": "0.000003", - "completion": "0.000003", + "prompt": "0.00000035", + "completion": "0.0000005", "image": "0", "request": "0", "webSearch": "0", @@ -28291,35 +29606,66 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", + "max_tokens", + "temperature", + "top_p", + "presence_penalty", + "frequency_penalty", + "repetition_penalty", + "top_k" + ], + "inputCost": 0.35, + "outputCost": 0.5, + "throughput": 56.904, + "latency": 569 + }, + { + "name": "Mancer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", + "slug": "mancer", + "quantization": "fp8", + "context": 32768, + "maxCompletionTokens": 2048, + "providerModelId": "qwen2.5-coder-32b", + "pricing": { + "prompt": "0.000000375", + "completion": "0.0000015", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0.25 + }, + "supportedParameters": [ "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "top_k", "repetition_penalty", - "response_format", - "structured_outputs", "logit_bias", - "logprobs", - "top_logprobs" + "top_k", + "min_p", + "seed", + "top_a" ], - "inputCost": 3, - "outputCost": 3 + "inputCost": 0.38, + "outputCost": 1.5, + "throughput": 31.575, + "latency": 807 }, { - "name": "Together", - "slug": "together", + "name": "Mancer (private)", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", + "slug": "mancer (private)", "quantization": "fp8", - "context": 130815, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo", + "context": 32768, + "maxCompletionTokens": 2048, + "providerModelId": "qwen2.5-coder-32b", "pricing": { - "prompt": "0.0000035", - "completion": "0.0000035", + "prompt": "0.0000005", + "completion": "0.000002", "image": "0", "request": "0", "webSearch": "0", @@ -28327,33 +29673,67 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "top_k", "repetition_penalty", "logit_bias", + "top_k", "min_p", - "response_format" + "seed", + "top_a" ], - "inputCost": 3.5, - "outputCost": 3.5 + "inputCost": 0.5, + "outputCost": 2, + "throughput": 31.802, + "latency": 1431 }, { - "name": "Hyperbolic", - "slug": "hyperbolic", - "quantization": "bf16", - "context": 131000, + "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", + "slug": "cloudflare", + "quantization": null, + "context": 32768, "maxCompletionTokens": null, - "providerModelId": "meta-llama/Meta-Llama-3.1-405B-Instruct", + "providerModelId": "@cf/qwen/qwen2.5-coder-32b-instruct", "pricing": { - "prompt": "0.000004", - "completion": "0.000004", + "prompt": "0.00000066", + "completion": "0.000001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "top_k", + "seed", + "repetition_penalty", + "frequency_penalty", + "presence_penalty" + ], + "inputCost": 0.66, + "outputCost": 1, + "throughput": 34.9445, + "latency": 1051 + }, + { + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": "fp16", + "context": 16384, + "maxCompletionTokens": 2048, + "providerModelId": "Qwen/Qwen2.5-Coder-32B-Instruct", + "pricing": { + "prompt": "0.0000008", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -28367,27 +29747,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "logprobs", - "top_logprobs", - "seed", - "logit_bias", "top_k", + "repetition_penalty", + "logit_bias", "min_p", - "repetition_penalty" + "response_format" ], - "inputCost": 4, - "outputCost": 4 + "inputCost": 0.8, + "outputCost": 0.8, + "throughput": 91.708, + "latency": 808 }, { - "name": "SambaNova", - "slug": "sambaNova", - "quantization": "bf16", + "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", + "slug": "featherless", + "quantization": "fp8", "context": 16384, "maxCompletionTokens": 4096, - "providerModelId": "Meta-Llama-3.1-405B-Instruct", + "providerModelId": "Qwen/Qwen2.5-Coder-32B-Instruct", "pricing": { - "prompt": "0.000005", - "completion": "0.00001", + "prompt": "0.0000026", + "completion": "0.0000034", "image": "0", "request": "0", "webSearch": "0", @@ -28398,11 +29779,16 @@ export default { "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", "top_k", - "stop" + "min_p", + "seed" ], - "inputCost": 5, - "outputCost": 10 + "inputCost": 2.6, + "outputCost": 3.4 } ] }, @@ -28571,15 +29957,17 @@ export default { }, "sorting": { "topWeekly": 87, - "newest": 195, - "throughputHighToLow": 102, - "latencyLowToHigh": 43, + "newest": 194, + "throughputHighToLow": 103, + "latencyLowToHigh": 42, "pricingLowToHigh": 162, "pricingHighToLow": 156 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "Infermatic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://infermatic.ai/&size=256", "slug": "infermatic", "quantization": "bf16", "context": 32768, @@ -28608,10 +29996,13 @@ export default { "seed" ], "inputCost": 0.25, - "outputCost": 0.5 + "outputCost": 0.5, + "throughput": 76.1635, + "latency": 385 }, { "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", "slug": "featherless", "quantization": "fp8", "context": 16384, @@ -28639,115 +30030,127 @@ export default { "seed" ], "inputCost": 0.8, - "outputCost": 1.2 + "outputCost": 1.2, + "throughput": 18.181, + "latency": 1292 } ] }, { - "slug": "cohere/command-r7b-12-2024", - "hfSlug": "", + "slug": "google/gemma-3-12b-it", + "hfSlug": "google/gemma-3-12b-it", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-14T06:35:52.905418+00:00", + "createdAt": "2025-03-13T21:50:25.140801+00:00", "hfUpdatedAt": null, - "name": "Cohere: Command R7B (12-2024)", - "shortName": "Command R7B (12-2024)", - "author": "cohere", - "description": "Command R7B (12-2024) is a small, fast update of the Command R+ model, delivered in December 2024. It excels at RAG, tool use, agents, and similar tasks requiring complex reasoning and multiple steps.\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", - "modelVersionGroupId": null, - "contextLength": 128000, + "name": "Google: Gemma 3 12B", + "shortName": "Gemma 3 12B", + "author": "google", + "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 12B is the second largest in the family of Gemma 3 models after [Gemma 3 27B](google/gemma-3-27b-it)", + "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", + "contextLength": 131072, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Cohere", - "instructType": null, + "group": "Gemini", + "instructType": "gemma", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "", + "", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cohere/command-r7b-12-2024", + "permaslug": "google/gemma-3-12b-it", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "d4a57205-8124-413a-ab4e-5ae94eb7aa9d", - "name": "Cohere | cohere/command-r7b-12-2024", - "contextLength": 128000, + "id": "eb06dc92-5a16-47ec-a776-6ef956457c47", + "name": "DeepInfra | google/gemma-3-12b-it", + "contextLength": 131072, "model": { - "slug": "cohere/command-r7b-12-2024", - "hfSlug": "", + "slug": "google/gemma-3-12b-it", + "hfSlug": "google/gemma-3-12b-it", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-14T06:35:52.905418+00:00", + "createdAt": "2025-03-13T21:50:25.140801+00:00", "hfUpdatedAt": null, - "name": "Cohere: Command R7B (12-2024)", - "shortName": "Command R7B (12-2024)", - "author": "cohere", - "description": "Command R7B (12-2024) is a small, fast update of the Command R+ model, delivered in December 2024. It excels at RAG, tool use, agents, and similar tasks requiring complex reasoning and multiple steps.\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", - "modelVersionGroupId": null, - "contextLength": 128000, + "name": "Google: Gemma 3 12B", + "shortName": "Gemma 3 12B", + "author": "google", + "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 12B is the second largest in the family of Gemma 3 models after [Gemma 3 27B](google/gemma-3-27b-it)", + "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", + "contextLength": 131072, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Cohere", - "instructType": null, + "group": "Gemini", + "instructType": "gemma", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "", + "", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cohere/command-r7b-12-2024", + "permaslug": "google/gemma-3-12b-it", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "cohere/command-r7b-12-2024", - "modelVariantPermaslug": "cohere/command-r7b-12-2024", - "providerName": "Cohere", + "modelVariantSlug": "google/gemma-3-12b-it", + "modelVariantPermaslug": "google/gemma-3-12b-it", + "providerName": "DeepInfra", "providerInfo": { - "name": "Cohere", - "displayName": "Cohere", - "slug": "cohere", + "name": "DeepInfra", + "displayName": "DeepInfra", + "slug": "deepinfra", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://cohere.com/privacy", - "termsOfServiceUrl": "https://cohere.com/terms-of-use", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, + "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Cohere", + "group": "DeepInfra", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.cohere.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Cohere.png" + "url": "/images/icons/DeepInfra.webp" } }, - "providerDisplayName": "Cohere", - "providerModelId": "command-r7b-12-2024", - "providerGroup": "Cohere", - "quantization": null, + "providerDisplayName": "DeepInfra", + "providerModelId": "google/gemma-3-12b-it", + "providerGroup": "DeepInfra", + "quantization": "bf16", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4000, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -28757,28 +30160,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "repetition_penalty", + "response_format", "top_k", "seed", - "response_format", - "structured_outputs" + "min_p" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://cohere.com/privacy", - "termsOfServiceUrl": "https://cohere.com/terms-of-use", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, "pricing": { - "prompt": "0.0000000375", - "completion": "0.00000015", + "prompt": "0.00000005", + "completion": "0.0000001", "image": "0", "request": "0", "webSearch": "0", @@ -28794,32 +30197,35 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 88, - "newest": 155, - "throughputHighToLow": 21, - "latencyLowToHigh": 6, - "pricingLowToHigh": 92, - "pricingHighToLow": 225 + "topWeekly": 77, + "newest": 87, + "throughputHighToLow": 170, + "latencyLowToHigh": 196, + "pricingLowToHigh": 39, + "pricingHighToLow": 220 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [ { - "name": "Cohere", - "slug": "cohere", - "quantization": null, - "context": 128000, - "maxCompletionTokens": 4000, - "providerModelId": "command-r7b-12-2024", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "google/gemma-3-12b-it", "pricing": { - "prompt": "0.0000000375", - "completion": "0.00000015", + "prompt": "0.00000005", + "completion": "0.0000001", "image": "0", "request": "0", "webSearch": "0", @@ -28833,28 +30239,63 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "repetition_penalty", + "response_format", "top_k", "seed", - "response_format", - "structured_outputs" + "min_p" ], - "inputCost": 0.04, - "outputCost": 0.15 + "inputCost": 0.05, + "outputCost": 0.1, + "throughput": 23.922, + "latency": 1180 + }, + { + "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", + "slug": "cloudflare", + "quantization": null, + "context": 80000, + "maxCompletionTokens": null, + "providerModelId": "@cf/google/gemma-3-12b-it", + "pricing": { + "prompt": "0.00000035", + "completion": "0.00000056", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "top_k", + "seed", + "repetition_penalty", + "frequency_penalty", + "presence_penalty" + ], + "inputCost": 0.35, + "outputCost": 0.56, + "throughput": 68.1855, + "latency": 595 } ] }, { - "slug": "qwen/qwen3-30b-a3b", - "hfSlug": "Qwen/Qwen3-30B-A3B", - "updatedAt": "2025-05-12T00:33:13.151167+00:00", - "createdAt": "2025-04-28T22:16:44.177326+00:00", + "slug": "nvidia/llama-3.1-nemotron-ultra-253b-v1", + "hfSlug": "nvidia/Llama-3_1-Nemotron-Ultra-253B-v1", + "updatedAt": "2025-04-08T16:16:48.985354+00:00", + "createdAt": "2025-04-08T12:24:19.697786+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 30B A3B (free)", - "shortName": "Qwen3 30B A3B (free)", - "author": "qwen", - "description": "Qwen3, the latest generation in the Qwen large language model series, features both dense and mixture-of-experts (MoE) architectures to excel in reasoning, multilingual support, and advanced agent tasks. Its unique ability to switch seamlessly between a thinking mode for complex reasoning and a non-thinking mode for efficient dialogue ensures versatile, high-quality performance.\n\nSignificantly outperforming prior models like QwQ and Qwen2.5, Qwen3 delivers superior mathematics, coding, commonsense reasoning, creative writing, and interactive dialogue capabilities. The Qwen3-30B-A3B variant includes 30.5 billion parameters (3.3 billion activated), 48 layers, 128 experts (8 activated per task), and supports up to 131K token contexts with YaRN, setting a new standard among open-source models.", + "name": "NVIDIA: Llama 3.1 Nemotron Ultra 253B v1 (free)", + "shortName": "Llama 3.1 Nemotron Ultra 253B v1 (free)", + "author": "nvidia", + "description": "Llama-3.1-Nemotron-Ultra-253B-v1 is a large language model (LLM) optimized for advanced reasoning, human-interactive chat, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta’s Llama-3.1-405B-Instruct, it has been significantly customized using Neural Architecture Search (NAS), resulting in enhanced efficiency, reduced memory usage, and improved inference latency. The model supports a context length of up to 128K tokens and can operate efficiently on an 8x NVIDIA H100 node.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.", "modelVersionGroupId": null, - "contextLength": 40960, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -28862,41 +30303,30 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Llama3", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen3-30b-a3b-04-28", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "permaslug": "nvidia/llama-3.1-nemotron-ultra-253b-v1", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "f31002e4-a626-4f70-9b11-f33e8102df39", - "name": "Chutes | qwen/qwen3-30b-a3b-04-28:free", - "contextLength": 40960, + "id": "2ce4225f-d3bf-47ed-9d92-d6a5c37c1725", + "name": "Chutes | nvidia/llama-3.1-nemotron-ultra-253b-v1:free", + "contextLength": 131072, "model": { - "slug": "qwen/qwen3-30b-a3b", - "hfSlug": "Qwen/Qwen3-30B-A3B", - "updatedAt": "2025-05-12T00:33:13.151167+00:00", - "createdAt": "2025-04-28T22:16:44.177326+00:00", + "slug": "nvidia/llama-3.1-nemotron-ultra-253b-v1", + "hfSlug": "nvidia/Llama-3_1-Nemotron-Ultra-253B-v1", + "updatedAt": "2025-04-08T16:16:48.985354+00:00", + "createdAt": "2025-04-08T12:24:19.697786+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 30B A3B", - "shortName": "Qwen3 30B A3B", - "author": "qwen", - "description": "Qwen3, the latest generation in the Qwen large language model series, features both dense and mixture-of-experts (MoE) architectures to excel in reasoning, multilingual support, and advanced agent tasks. Its unique ability to switch seamlessly between a thinking mode for complex reasoning and a non-thinking mode for efficient dialogue ensures versatile, high-quality performance.\n\nSignificantly outperforming prior models like QwQ and Qwen2.5, Qwen3 delivers superior mathematics, coding, commonsense reasoning, creative writing, and interactive dialogue capabilities. The Qwen3-30B-A3B variant includes 30.5 billion parameters (3.3 billion activated), 48 layers, 128 experts (8 activated per task), and supports up to 131K token contexts with YaRN, setting a new standard among open-source models.", + "name": "NVIDIA: Llama 3.1 Nemotron Ultra 253B v1", + "shortName": "Llama 3.1 Nemotron Ultra 253B v1", + "author": "nvidia", + "description": "Llama-3.1-Nemotron-Ultra-253B-v1 is a large language model (LLM) optimized for advanced reasoning, human-interactive chat, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta’s Llama-3.1-405B-Instruct, it has been significantly customized using Neural Architecture Search (NAS), resulting in enhanced efficiency, reduced memory usage, and improved inference latency. The model supports a context length of up to 128K tokens and can operate efficiently on an 8x NVIDIA H100 node.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.", "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ @@ -28906,30 +30336,19 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Llama3", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen3-30b-a3b-04-28", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "permaslug": "nvidia/llama-3.1-nemotron-ultra-253b-v1", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "qwen/qwen3-30b-a3b:free", - "modelVariantPermaslug": "qwen/qwen3-30b-a3b-04-28:free", + "modelVariantSlug": "nvidia/llama-3.1-nemotron-ultra-253b-v1:free", + "modelVariantPermaslug": "nvidia/llama-3.1-nemotron-ultra-253b-v1:free", "providerName": "Chutes", "providerInfo": { "name": "Chutes", @@ -28959,9 +30378,9 @@ export default { } }, "providerDisplayName": "Chutes", - "providerModelId": "Qwen/Qwen3-30B-A3B", + "providerModelId": "nvidia/Llama-3_1-Nemotron-Ultra-253B-v1", "providerGroup": "Chutes", - "quantization": null, + "quantization": "bf16", "variant": "free", "isFree": true, "canAbort": true, @@ -28973,8 +30392,6 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -29011,7 +30428,7 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, @@ -29025,23 +30442,202 @@ export default { }, "sorting": { "topWeekly": 89, - "newest": 22, - "throughputHighToLow": 23, - "latencyLowToHigh": 127, - "pricingLowToHigh": 9, - "pricingHighToLow": 193 + "newest": 60, + "throughputHighToLow": 230, + "latencyLowToHigh": 238, + "pricingLowToHigh": 25, + "pricingHighToLow": 273 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nvidia.com/\\u0026size=256", + "providers": [] + }, + { + "slug": "meta-llama/llama-3.1-405b-instruct", + "hfSlug": "meta-llama/Meta-Llama-3.1-405B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-07-23T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Meta: Llama 3.1 405B Instruct", + "shortName": "Llama 3.1 405B Instruct", + "author": "meta-llama", + "description": "The highly anticipated 400B class of Llama3 is here! Clocking in at 128k context with impressive eval scores, the Meta AI team continues to push the frontier of open-source LLMs.\n\nMeta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 405B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models including GPT-4o and Claude 3.5 Sonnet in evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "modelVersionGroupId": "1fd9d06b-aa20-4c7d-a0b1-d3d9b5aae712", + "contextLength": 32768, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Llama3", + "instructType": "llama3", + "defaultSystem": null, + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "meta-llama/llama-3.1-405b-instruct", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "ea74be37-05ac-4934-9734-709cef644954", + "name": "DeepInfra | meta-llama/llama-3.1-405b-instruct", + "contextLength": 32768, + "model": { + "slug": "meta-llama/llama-3.1-405b-instruct", + "hfSlug": "meta-llama/Meta-Llama-3.1-405B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-07-23T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Meta: Llama 3.1 405B Instruct", + "shortName": "Llama 3.1 405B Instruct", + "author": "meta-llama", + "description": "The highly anticipated 400B class of Llama3 is here! Clocking in at 128k context with impressive eval scores, the Meta AI team continues to push the frontier of open-source LLMs.\n\nMeta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This 405B instruct-tuned version is optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models including GPT-4o and Claude 3.5 Sonnet in evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3-1/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "modelVersionGroupId": "1fd9d06b-aa20-4c7d-a0b1-d3d9b5aae712", + "contextLength": 131072, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Llama3", + "instructType": "llama3", + "defaultSystem": null, + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "meta-llama/llama-3.1-405b-instruct", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "meta-llama/llama-3.1-405b-instruct", + "modelVariantPermaslug": "meta-llama/llama-3.1-405b-instruct", + "providerName": "DeepInfra", + "providerInfo": { + "name": "DeepInfra", + "displayName": "DeepInfra", + "slug": "deepinfra", + "baseUrl": "url", + "dataPolicy": { + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", + "paidModels": { + "training": false, + "retainsPrompts": false + } + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "DeepInfra", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/DeepInfra.webp" + } + }, + "providerDisplayName": "DeepInfra", + "providerModelId": "meta-llama/Meta-Llama-3.1-405B-Instruct", + "providerGroup": "DeepInfra", + "quantization": "fp8", + "variant": "standard", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": 16384, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", + "paidModels": { + "training": false, + "retainsPrompts": false + }, + "training": false, + "retainsPrompts": false + }, + "pricing": { + "prompt": "0.0000008", + "completion": "0.0000008", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": true, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 90, + "newest": 231, + "throughputHighToLow": 278, + "latencyLowToHigh": 130, + "pricingLowToHigh": 201, + "pricingHighToLow": 117 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": 40960, - "providerModelId": "Qwen/Qwen3-30B-A3B", + "context": 32768, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/Meta-Llama-3.1-405B-Instruct", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", + "prompt": "0.0000008", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -29049,11 +30645,11 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -29063,19 +30659,22 @@ export default { "seed", "min_p" ], - "inputCost": 0.1, - "outputCost": 0.3 + "inputCost": 0.8, + "outputCost": 0.8, + "throughput": 22.463, + "latency": 788 }, { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", + "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", + "slug": "lambda", "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen3-30B-A3B", + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "llama3.1-405b-instruct-fp8", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", + "prompt": "0.0000008", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -29086,30 +30685,31 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", "seed", - "top_k", "logit_bias", "logprobs", - "top_logprobs" + "top_logprobs", + "response_format" ], - "inputCost": 0.1, - "outputCost": 0.3 + "inputCost": 0.8, + "outputCost": 0.8, + "throughput": 34.9805, + "latency": 563 }, { - "name": "NovitaAI", - "slug": "novitaAi", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", "quantization": "fp8", - "context": 128000, + "context": 131072, "maxCompletionTokens": null, - "providerModelId": "qwen/qwen3-30b-a3b-fp8", + "providerModelId": "meta-llama/Meta-Llama-3.1-405B-Instruct", "pricing": { - "prompt": "0.0000001", - "completion": "0.00000045", + "prompt": "0.000001", + "completion": "0.000003", "image": "0", "request": "0", "webSearch": "0", @@ -29120,30 +30720,31 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", "seed", "top_k", - "min_p", - "repetition_penalty", - "logit_bias" + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 0.1, - "outputCost": 0.45 + "inputCost": 1, + "outputCost": 3, + "throughput": 33.742, + "latency": 637 }, { - "name": "Parasail", - "slug": "parasail", + "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", + "slug": "fireworks", "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": 40960, - "providerModelId": "parasail-qwen3-30b-a3b", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "accounts/fireworks/models/llama-v3p1-405b-instruct", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000005", + "prompt": "0.000003", + "completion": "0.000003", "image": "0", "request": "0", "webSearch": "0", @@ -29151,29 +30752,38 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", - "presence_penalty", + "stop", "frequency_penalty", + "presence_penalty", + "top_k", "repetition_penalty", - "top_k" + "response_format", + "structured_outputs", + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 0.1, - "outputCost": 0.5 + "inputCost": 3, + "outputCost": 3, + "throughput": 70.087, + "latency": 595 }, { - "name": "Fireworks", - "slug": "fireworks", - "quantization": null, - "context": 40000, + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": "fp8", + "context": 130815, "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/qwen3-30b-a3b", + "providerModelId": "meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo", "pricing": { - "prompt": "0.00000015", - "completion": "0.0000006", + "prompt": "0.0000035", + "completion": "0.0000035", "image": "0", "request": "0", "webSearch": "0", @@ -29186,109 +30796,196 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", "top_k", "repetition_penalty", - "response_format", - "structured_outputs", "logit_bias", + "min_p", + "response_format" + ], + "inputCost": 3.5, + "outputCost": 3.5, + "throughput": 53.2975, + "latency": 458.5 + }, + { + "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", + "slug": "hyperbolic", + "quantization": "bf16", + "context": 131000, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Meta-Llama-3.1-405B-Instruct", + "pricing": { + "prompt": "0.000004", + "completion": "0.000004", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "logprobs", - "top_logprobs" + "top_logprobs", + "seed", + "logit_bias", + "top_k", + "min_p", + "repetition_penalty" ], - "inputCost": 0.15, - "outputCost": 0.6 + "inputCost": 4, + "outputCost": 4, + "throughput": 72.215, + "latency": 1414 + }, + { + "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", + "slug": "sambaNova", + "quantization": "bf16", + "context": 16384, + "maxCompletionTokens": 4096, + "providerModelId": "Meta-Llama-3.1-405B-Instruct", + "pricing": { + "prompt": "0.000005", + "completion": "0.00001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 5, + "outputCost": 10, + "throughput": 111.627, + "latency": 2801 } ] }, { - "slug": "mistralai/mistral-small-3.1-24b-instruct", - "hfSlug": "mistralai/Mistral-Small-3.1-24B-Instruct-2503", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-17T19:15:37.00423+00:00", + "slug": "qwen/qwen3-14b", + "hfSlug": "Qwen/Qwen3-14B", + "updatedAt": "2025-05-12T00:36:13.09542+00:00", + "createdAt": "2025-04-28T21:41:18.320017+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral Small 3.1 24B (free)", - "shortName": "Mistral Small 3.1 24B (free)", - "author": "mistralai", - "description": "Mistral Small 3.1 24B Instruct is an upgraded variant of Mistral Small 3 (2501), featuring 24 billion parameters with advanced multimodal capabilities. It provides state-of-the-art performance in text-based reasoning and vision tasks, including image analysis, programming, mathematical reasoning, and multilingual support across dozens of languages. Equipped with an extensive 128k token context window and optimized for efficient local inference, it supports use cases such as conversational agents, function calling, long-document comprehension, and privacy-sensitive deployments.", + "name": "Qwen: Qwen3 14B", + "shortName": "Qwen3 14B", + "author": "qwen", + "description": "Qwen3-14B is a dense 14.8B parameter causal language model from the Qwen3 series, designed for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, programming, and logical inference, and a \"non-thinking\" mode for general-purpose conversation. The model is fine-tuned for instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling.", "modelVersionGroupId": null, - "contextLength": 96000, + "contextLength": 40960, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": null, + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-small-3.1-24b-instruct-2503", - "reasoningConfig": null, - "features": {}, + "permaslug": "qwen/qwen3-14b-04-28", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "6aed58f7-0b3a-4693-b8c9-57d5eadec99f", - "name": "Chutes | mistralai/mistral-small-3.1-24b-instruct-2503:free", - "contextLength": 96000, + "id": "778bdebf-d018-4a24-a34e-230fce0f2045", + "name": "DeepInfra | qwen/qwen3-14b-04-28", + "contextLength": 40960, "model": { - "slug": "mistralai/mistral-small-3.1-24b-instruct", - "hfSlug": "mistralai/Mistral-Small-3.1-24B-Instruct-2503", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-17T19:15:37.00423+00:00", + "slug": "qwen/qwen3-14b", + "hfSlug": "Qwen/Qwen3-14B", + "updatedAt": "2025-05-12T00:36:13.09542+00:00", + "createdAt": "2025-04-28T21:41:18.320017+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral Small 3.1 24B", - "shortName": "Mistral Small 3.1 24B", - "author": "mistralai", - "description": "Mistral Small 3.1 24B Instruct is an upgraded variant of Mistral Small 3 (2501), featuring 24 billion parameters with advanced multimodal capabilities. It provides state-of-the-art performance in text-based reasoning and vision tasks, including image analysis, programming, mathematical reasoning, and multilingual support across dozens of languages. Equipped with an extensive 128k token context window and optimized for efficient local inference, it supports use cases such as conversational agents, function calling, long-document comprehension, and privacy-sensitive deployments.", + "name": "Qwen: Qwen3 14B", + "shortName": "Qwen3 14B", + "author": "qwen", + "description": "Qwen3-14B is a dense 14.8B parameter causal language model from the Qwen3 series, designed for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, programming, and logical inference, and a \"non-thinking\" mode for general-purpose conversation. The model is fine-tuned for instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 131702, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": null, + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-small-3.1-24b-instruct-2503", - "reasoningConfig": null, - "features": {} + "permaslug": "qwen/qwen3-14b-04-28", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "mistralai/mistral-small-3.1-24b-instruct:free", - "modelVariantPermaslug": "mistralai/mistral-small-3.1-24b-instruct-2503:free", - "providerName": "Chutes", + "modelVariantSlug": "qwen/qwen3-14b", + "modelVariantPermaslug": "qwen/qwen3-14b-04-28", + "providerName": "DeepInfra", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "DeepInfra", + "displayName": "DeepInfra", + "slug": "deepinfra", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false } }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Chutes", + "group": "DeepInfra", "editors": [], "owners": [], "isMultipartSupported": true, @@ -29296,51 +30993,51 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "/images/icons/DeepInfra.webp" } }, - "providerDisplayName": "Chutes", - "providerModelId": "chutesai/Mistral-Small-3.1-24B-Instruct-2503", - "providerGroup": "Chutes", - "quantization": null, - "variant": "free", - "isFree": true, + "providerDisplayName": "DeepInfra", + "providerModelId": "Qwen/Qwen3-14B", + "providerGroup": "DeepInfra", + "quantization": "fp8", + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 96000, + "maxCompletionTokens": 40960, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "min_p", "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" + "response_format", + "top_k", + "seed", + "min_p" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.00000007", + "completion": "0.00000024", "image": "0", "request": "0", "webSearch": "0", @@ -29351,8 +31048,8 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, + "supportsToolParameters": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, @@ -29365,24 +31062,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 72, - "newest": 79, - "throughputHighToLow": 96, - "latencyLowToHigh": 42, - "pricingLowToHigh": 35, - "pricingHighToLow": 219 + "topWeekly": 91, + "newest": 26, + "throughputHighToLow": 64, + "latencyLowToHigh": 168, + "pricingLowToHigh": 11, + "pricingHighToLow": 212 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "mistralai/Mistral-Small-3.1-24B-Instruct-2503", + "context": 40960, + "maxCompletionTokens": 40960, + "providerModelId": "Qwen/Qwen3-14B", "pricing": { - "prompt": "0.00000005", - "completion": "0.00000015", + "prompt": "0.00000007", + "completion": "0.00000024", "image": "0", "request": "0", "webSearch": "0", @@ -29393,88 +31092,70 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "seed", + "repetition_penalty", + "response_format", "top_k", - "logit_bias", - "logprobs", - "top_logprobs" + "seed", + "min_p" ], - "inputCost": 0.05, - "outputCost": 0.15 + "inputCost": 0.07, + "outputCost": 0.24, + "throughput": 75.656, + "latency": 1075 }, { - "name": "Parasail", - "slug": "parasail", - "quantization": "bf16", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": "fp8", "context": 128000, - "maxCompletionTokens": 128000, - "providerModelId": "parasail-mistral-small-31-24b-instruct", - "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", - "image": "0.000926", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "presence_penalty", - "frequency_penalty", - "repetition_penalty", - "top_k" - ], - "inputCost": 0.1, - "outputCost": 0.3 - }, - { - "name": "Mistral", - "slug": "mistral", - "quantization": null, - "context": 131072, "maxCompletionTokens": null, - "providerModelId": "mistral-small-2503", + "providerModelId": "qwen/qwen3-14b-fp8", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", - "image": "0.0009264", + "prompt": "0.00000007", + "completion": "0.000000275", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "response_format", - "structured_outputs", - "seed" + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" ], - "inputCost": 0.1, - "outputCost": 0.3 + "inputCost": 0.07, + "outputCost": 0.28, + "throughput": 57.333, + "latency": 787 }, { - "name": "Cloudflare", - "slug": "cloudflare", - "quantization": null, - "context": 128000, + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 40960, "maxCompletionTokens": null, - "providerModelId": "@cf/mistralai/mistral-small-3.1-24b-instruct", + "providerModelId": "Qwen/Qwen3-14B", "pricing": { - "prompt": "0.00000035", - "completion": "0.00000056", + "prompt": "0.00000008", + "completion": "0.00000024", "image": "0", "request": "0", "webSearch": "0", @@ -29485,27 +31166,34 @@ export default { "max_tokens", "temperature", "top_p", - "top_k", - "seed", - "repetition_penalty", + "reasoning", + "include_reasoning", + "stop", "frequency_penalty", - "presence_penalty" + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 0.35, - "outputCost": 0.56 + "inputCost": 0.08, + "outputCost": 0.24, + "throughput": 93.2925, + "latency": 273 } ] }, { - "slug": "nvidia/llama-3.1-nemotron-ultra-253b-v1", - "hfSlug": "nvidia/Llama-3_1-Nemotron-Ultra-253B-v1", - "updatedAt": "2025-04-08T16:16:48.985354+00:00", - "createdAt": "2025-04-08T12:24:19.697786+00:00", + "slug": "deepseek/deepseek-prover-v2", + "hfSlug": "deepseek-ai/DeepSeek-Prover-V2-671B", + "updatedAt": "2025-04-30T13:01:50.865169+00:00", + "createdAt": "2025-04-30T11:38:14.302503+00:00", "hfUpdatedAt": null, - "name": "NVIDIA: Llama 3.1 Nemotron Ultra 253B v1 (free)", - "shortName": "Llama 3.1 Nemotron Ultra 253B v1 (free)", - "author": "nvidia", - "description": "Llama-3.1-Nemotron-Ultra-253B-v1 is a large language model (LLM) optimized for advanced reasoning, human-interactive chat, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta’s Llama-3.1-405B-Instruct, it has been significantly customized using Neural Architecture Search (NAS), resulting in enhanced efficiency, reduced memory usage, and improved inference latency. The model supports a context length of up to 128K tokens and can operate efficiently on an 8x NVIDIA H100 node.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.", + "name": "DeepSeek: DeepSeek Prover V2", + "shortName": "DeepSeek Prover V2", + "author": "deepseek", + "description": "DeepSeek Prover V2 is a 671B parameter model, speculated to be geared towards logic and mathematics. Likely an upgrade from [DeepSeek-Prover-V1.5](https://huggingface.co/deepseek-ai/DeepSeek-Prover-V1.5-RL) Not much is known about the model yet, as DeepSeek released it on Hugging Face without an announcement or description.", "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ @@ -29515,32 +31203,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", + "group": "DeepSeek", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "nvidia/llama-3.1-nemotron-ultra-253b-v1", + "permaslug": "deepseek/deepseek-prover-v2", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "2ce4225f-d3bf-47ed-9d92-d6a5c37c1725", - "name": "Chutes | nvidia/llama-3.1-nemotron-ultra-253b-v1:free", + "id": "c70b9a14-2531-4ed0-ab12-b8935873c6c7", + "name": "GMICloud | deepseek/deepseek-prover-v2", "contextLength": 131072, "model": { - "slug": "nvidia/llama-3.1-nemotron-ultra-253b-v1", - "hfSlug": "nvidia/Llama-3_1-Nemotron-Ultra-253B-v1", - "updatedAt": "2025-04-08T16:16:48.985354+00:00", - "createdAt": "2025-04-08T12:24:19.697786+00:00", + "slug": "deepseek/deepseek-prover-v2", + "hfSlug": "deepseek-ai/DeepSeek-Prover-V2-671B", + "updatedAt": "2025-04-30T13:01:50.865169+00:00", + "createdAt": "2025-04-30T11:38:14.302503+00:00", "hfUpdatedAt": null, - "name": "NVIDIA: Llama 3.1 Nemotron Ultra 253B v1", - "shortName": "Llama 3.1 Nemotron Ultra 253B v1", - "author": "nvidia", - "description": "Llama-3.1-Nemotron-Ultra-253B-v1 is a large language model (LLM) optimized for advanced reasoning, human-interactive chat, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta’s Llama-3.1-405B-Instruct, it has been significantly customized using Neural Architecture Search (NAS), resulting in enhanced efficiency, reduced memory usage, and improved inference latency. The model supports a context length of up to 128K tokens and can operate efficiently on an 8x NVIDIA H100 node.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.", + "name": "DeepSeek: DeepSeek Prover V2", + "shortName": "DeepSeek Prover V2", + "author": "deepseek", + "description": "DeepSeek Prover V2 is a 671B parameter model, speculated to be geared towards logic and mathematics. Likely an upgrade from [DeepSeek-Prover-V1.5](https://huggingface.co/deepseek-ai/DeepSeek-Prover-V1.5-RL) Not much is known about the model yet, as DeepSeek released it on Hugging Face without an announcement or description.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 163840, "inputModalities": [ "text" ], @@ -29548,37 +31236,36 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", + "group": "DeepSeek", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "nvidia/llama-3.1-nemotron-ultra-253b-v1", + "permaslug": "deepseek/deepseek-prover-v2", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "nvidia/llama-3.1-nemotron-ultra-253b-v1:free", - "modelVariantPermaslug": "nvidia/llama-3.1-nemotron-ultra-253b-v1:free", - "providerName": "Chutes", + "modelVariantSlug": "deepseek/deepseek-prover-v2", + "modelVariantPermaslug": "deepseek/deepseek-prover-v2", + "providerName": "GMICloud", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "GMICloud", + "displayName": "GMICloud", + "slug": "gmicloud", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "privacyPolicyUrl": "https://docs.gmicloud.ai/privacy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false } }, "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Chutes", + "group": "GMICloud", "editors": [], "owners": [], "isMultipartSupported": true, @@ -29586,15 +31273,16 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://gmicloud.ai/&size=256", + "invertRequired": true } }, - "providerDisplayName": "Chutes", - "providerModelId": "nvidia/Llama-3_1-Nemotron-Ultra-253B-v1", - "providerGroup": "Chutes", - "quantization": "bf16", - "variant": "free", - "isFree": true, + "providerDisplayName": "GMICloud", + "providerModelId": "deepseek-ai/DeepSeek-Prover-V2-671B", + "providerGroup": "GMICloud", + "quantization": "fp8", + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, @@ -29604,31 +31292,20 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "privacyPolicyUrl": "https://docs.gmicloud.ai/privacy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false }, - "training": true, - "retainsPrompts": true + "training": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.0000005", + "completion": "0.00000218", "image": "0", "request": "0", "webSearch": "0", @@ -29653,25 +31330,125 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 91, - "newest": 60, - "throughputHighToLow": 226, - "latencyLowToHigh": 223, - "pricingLowToHigh": 25, - "pricingHighToLow": 273 + "topWeekly": 74, + "newest": 19, + "throughputHighToLow": 110, + "latencyLowToHigh": 197, + "pricingLowToHigh": 8, + "pricingHighToLow": 123 }, - "providers": [] - }, - { - "slug": "meta-llama/llama-3-70b-instruct", - "hfSlug": "meta-llama/Meta-Llama-3-70B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-04-18T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Meta: Llama 3 70B Instruct", - "shortName": "Llama 3 70B Instruct", - "author": "meta-llama", - "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 70B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", + "providers": [ + { + "name": "GMICloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://gmicloud.ai/&size=256", + "slug": "gmiCloud", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "deepseek-ai/DeepSeek-Prover-V2-671B", + "pricing": { + "prompt": "0.0000005", + "completion": "0.00000218", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "seed" + ], + "inputCost": 0.5, + "outputCost": 2.18, + "throughput": 62.1385, + "latency": 1209 + }, + { + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "fp8", + "context": 163840, + "maxCompletionTokens": null, + "providerModelId": "deepseek-ai/DeepSeek-Prover-V2-671B", + "pricing": { + "prompt": "0.0000007", + "completion": "0.00000218", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" + ], + "inputCost": 0.7, + "outputCost": 2.18, + "throughput": 52.849, + "latency": 1226 + }, + { + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": "fp8", + "context": 160000, + "maxCompletionTokens": null, + "providerModelId": "deepseek/deepseek-prover-v2-671b", + "pricing": { + "prompt": "0.0000007", + "completion": "0.0000025", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" + ], + "inputCost": 0.7, + "outputCost": 2.5, + "throughput": 24.435, + "latency": 1327 + } + ] + }, + { + "slug": "meta-llama/llama-3-70b-instruct", + "hfSlug": "meta-llama/Meta-Llama-3-70B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-04-18T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Meta: Llama 3 70B Instruct", + "shortName": "Llama 3 70B Instruct", + "author": "meta-llama", + "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This 70B instruct-tuned version was optimized for high quality dialogue usecases.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", "modelVersionGroupId": "397604e2-45fa-454e-a85d-9921f5138747", "contextLength": 8192, "inputModalities": [ @@ -29829,16 +31606,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 92, + "topWeekly": 93, "newest": 267, - "throughputHighToLow": 252, - "latencyLowToHigh": 92, + "throughputHighToLow": 240, + "latencyLowToHigh": 117, "pricingLowToHigh": 165, "pricingHighToLow": 153 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "bf16", "context": 8192, @@ -29869,10 +31648,13 @@ export default { "min_p" ], "inputCost": 0.3, - "outputCost": 0.4 + "outputCost": 0.4, + "throughput": 31.989, + "latency": 677 }, { "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", "slug": "hyperbolic", "quantization": null, "context": 8192, @@ -29903,10 +31685,13 @@ export default { "repetition_penalty" ], "inputCost": 0.4, - "outputCost": 0.4 + "outputCost": 0.4, + "throughput": 21.79, + "latency": 1526 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "unknown", "context": 8192, @@ -29935,10 +31720,13 @@ export default { "logit_bias" ], "inputCost": 0.51, - "outputCost": 0.74 + "outputCost": 0.74, + "throughput": 20.4, + "latency": 1112 }, { "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", "slug": "groq", "quantization": null, "context": 8192, @@ -29967,10 +31755,13 @@ export default { "seed" ], "inputCost": 0.59, - "outputCost": 0.79 + "outputCost": 0.79, + "throughput": 428.1315, + "latency": 295 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "fp8", "context": 8192, @@ -29999,22 +31790,24 @@ export default { "response_format" ], "inputCost": 0.88, - "outputCost": 0.88 + "outputCost": 0.88, + "throughput": 70.498, + "latency": 1003 } ] }, { - "slug": "mistralai/codestral-2501", - "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-01-14T22:58:42.43034+00:00", + "slug": "qwen/qwen3-30b-a3b", + "hfSlug": "Qwen/Qwen3-30B-A3B", + "updatedAt": "2025-05-12T00:33:13.151167+00:00", + "createdAt": "2025-04-28T22:16:44.177326+00:00", "hfUpdatedAt": null, - "name": "Mistral: Codestral 2501", - "shortName": "Codestral 2501", - "author": "mistralai", - "description": "[Mistral](/mistralai)'s cutting-edge language model for coding. Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM), code correction and test generation. \n\nLearn more on their blog post: https://mistral.ai/news/codestral-2501/", + "name": "Qwen: Qwen3 30B A3B (free)", + "shortName": "Qwen3 30B A3B (free)", + "author": "qwen", + "description": "Qwen3, the latest generation in the Qwen large language model series, features both dense and mixture-of-experts (MoE) architectures to excel in reasoning, multilingual support, and advanced agent tasks. Its unique ability to switch seamlessly between a thinking mode for complex reasoning and a non-thinking mode for efficient dialogue ensures versatile, high-quality performance.\n\nSignificantly outperforming prior models like QwQ and Qwen2.5, Qwen3 delivers superior mathematics, coding, commonsense reasoning, creative writing, and interactive dialogue capabilities. The Qwen3-30B-A3B variant includes 30.5 billion parameters (3.3 billion activated), 48 layers, 128 experts (8 activated per task), and supports up to 131K token contexts with YaRN, setting a new standard among open-source models.", "modelVersionGroupId": null, - "contextLength": 262144, + "contextLength": 40960, "inputModalities": [ "text" ], @@ -30022,32 +31815,43 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": null, + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>" + ], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "mistralai/codestral-2501", - "reasoningConfig": null, - "features": {}, + "warningMessage": null, + "permaslug": "qwen/qwen3-30b-a3b-04-28", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "7df94044-ae68-4b4f-b3eb-748ed262b778", - "name": "Mistral | mistralai/codestral-2501", - "contextLength": 262144, + "id": "f31002e4-a626-4f70-9b11-f33e8102df39", + "name": "Chutes | qwen/qwen3-30b-a3b-04-28:free", + "contextLength": 40960, "model": { - "slug": "mistralai/codestral-2501", - "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-01-14T22:58:42.43034+00:00", + "slug": "qwen/qwen3-30b-a3b", + "hfSlug": "Qwen/Qwen3-30B-A3B", + "updatedAt": "2025-05-12T00:33:13.151167+00:00", + "createdAt": "2025-04-28T22:16:44.177326+00:00", "hfUpdatedAt": null, - "name": "Mistral: Codestral 2501", - "shortName": "Codestral 2501", - "author": "mistralai", - "description": "[Mistral](/mistralai)'s cutting-edge language model for coding. Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM), code correction and test generation. \n\nLearn more on their blog post: https://mistral.ai/news/codestral-2501/", + "name": "Qwen: Qwen3 30B A3B", + "shortName": "Qwen3 30B A3B", + "author": "qwen", + "description": "Qwen3, the latest generation in the Qwen large language model series, features both dense and mixture-of-experts (MoE) architectures to excel in reasoning, multilingual support, and advanced agent tasks. Its unique ability to switch seamlessly between a thinking mode for complex reasoning and a non-thinking mode for efficient dialogue ensures versatile, high-quality performance.\n\nSignificantly outperforming prior models like QwQ and Qwen2.5, Qwen3 delivers superior mathematics, coding, commonsense reasoning, creative writing, and interactive dialogue capabilities. The Qwen3-30B-A3B variant includes 30.5 billion parameters (3.3 billion activated), 48 layers, 128 experts (8 activated per task), and supports up to 131K token contexts with YaRN, setting a new standard among open-source models.", "modelVersionGroupId": null, - "contextLength": 256000, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -30055,40 +31859,48 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": null, + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>" + ], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "mistralai/codestral-2501", - "reasoningConfig": null, - "features": {} + "warningMessage": null, + "permaslug": "qwen/qwen3-30b-a3b-04-28", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "mistralai/codestral-2501", - "modelVariantPermaslug": "mistralai/codestral-2501", - "providerName": "Mistral", + "modelVariantSlug": "qwen/qwen3-30b-a3b:free", + "modelVariantPermaslug": "qwen/qwen3-30b-a3b-04-28:free", + "providerName": "Chutes", "providerInfo": { - "name": "Mistral", - "displayName": "Mistral", - "slug": "mistral", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true } }, - "headquarters": "FR", "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, + "hasCompletions": true, + "isAbortable": true, "moderationRequired": false, - "group": "Mistral", + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, @@ -30096,50 +31908,51 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Mistral.png" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "Mistral", - "providerModelId": "codestral-2501", - "providerGroup": "Mistral", + "providerDisplayName": "Chutes", + "providerModelId": "Qwen/Qwen3-30B-A3B", + "providerGroup": "Chutes", "quantization": null, - "variant": "standard", - "isFree": false, - "canAbort": false, + "variant": "free", + "isFree": true, + "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "response_format", - "structured_outputs", - "seed" + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.0000003", - "completion": "0.0000009", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -30150,8 +31963,8 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, + "supportsToolParameters": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, @@ -30164,24 +31977,170 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 93, - "newest": 146, - "throughputHighToLow": 43, - "latencyLowToHigh": 37, - "pricingLowToHigh": 171, - "pricingHighToLow": 147 + "topWeekly": 94, + "newest": 22, + "throughputHighToLow": 20, + "latencyLowToHigh": 127, + "pricingLowToHigh": 9, + "pricingHighToLow": 193 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { - "name": "Mistral", - "slug": "mistral", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "fp8", + "context": 40960, + "maxCompletionTokens": 40960, + "providerModelId": "Qwen/Qwen3-30B-A3B", + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" + ], + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 45.735, + "latency": 762 + }, + { + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 40960, + "maxCompletionTokens": null, + "providerModelId": "Qwen/Qwen3-30B-A3B", + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 97.242, + "latency": 349 + }, + { + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": "fp8", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "qwen/qwen3-30b-a3b-fp8", + "pricing": { + "prompt": "0.0000001", + "completion": "0.00000045", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" + ], + "inputCost": 0.1, + "outputCost": 0.45, + "throughput": 98.0455, + "latency": 900 + }, + { + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", + "quantization": "fp8", + "context": 40960, + "maxCompletionTokens": 40960, + "providerModelId": "parasail-qwen3-30b-a3b", + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000005", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "presence_penalty", + "frequency_penalty", + "repetition_penalty", + "top_k" + ], + "inputCost": 0.1, + "outputCost": 0.5, + "throughput": 119.9285, + "latency": 779 + }, + { + "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", + "slug": "fireworks", "quantization": null, - "context": 262144, + "context": 40000, "maxCompletionTokens": null, - "providerModelId": "codestral-2501", + "providerModelId": "accounts/fireworks/models/qwen3-30b-a3b", "pricing": { - "prompt": "0.0000003", - "completion": "0.0000009", + "prompt": "0.00000015", + "completion": "0.0000006", "image": "0", "request": "0", "webSearch": "0", @@ -30194,30 +32153,38 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", + "top_k", + "repetition_penalty", "response_format", "structured_outputs", - "seed" + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 0.3, - "outputCost": 0.9 + "inputCost": 0.15, + "outputCost": 0.6, + "throughput": 117.839, + "latency": 725 } ] }, { - "slug": "google/gemma-3-12b-it", - "hfSlug": "google/gemma-3-12b-it", + "slug": "mistralai/mistral-small-3.1-24b-instruct", + "hfSlug": "mistralai/Mistral-Small-3.1-24B-Instruct-2503", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-13T21:50:25.140801+00:00", + "createdAt": "2025-03-17T19:15:37.00423+00:00", "hfUpdatedAt": null, - "name": "Google: Gemma 3 12B", - "shortName": "Gemma 3 12B", - "author": "google", - "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 12B is the second largest in the family of Gemma 3 models after [Gemma 3 27B](google/gemma-3-27b-it)", - "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", - "contextLength": 131072, + "name": "Mistral: Mistral Small 3.1 24B (free)", + "shortName": "Mistral Small 3.1 24B (free)", + "author": "mistralai", + "description": "Mistral Small 3.1 24B Instruct is an upgraded variant of Mistral Small 3 (2501), featuring 24 billion parameters with advanced multimodal capabilities. It provides state-of-the-art performance in text-based reasoning and vision tasks, including image analysis, programming, mathematical reasoning, and multilingual support across dozens of languages. Equipped with an extensive 128k token context window and optimized for efficient local inference, it supports use cases such as conversational agents, function calling, long-document comprehension, and privacy-sensitive deployments.", + "modelVersionGroupId": null, + "contextLength": 96000, "inputModalities": [ "text", "image" @@ -30226,36 +32193,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Gemini", - "instructType": "gemma", + "group": "Mistral", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "", - "", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "google/gemma-3-12b-it", + "permaslug": "mistralai/mistral-small-3.1-24b-instruct-2503", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "eb06dc92-5a16-47ec-a776-6ef956457c47", - "name": "DeepInfra | google/gemma-3-12b-it", - "contextLength": 131072, + "id": "6aed58f7-0b3a-4693-b8c9-57d5eadec99f", + "name": "Chutes | mistralai/mistral-small-3.1-24b-instruct-2503:free", + "contextLength": 96000, "model": { - "slug": "google/gemma-3-12b-it", - "hfSlug": "google/gemma-3-12b-it", + "slug": "mistralai/mistral-small-3.1-24b-instruct", + "hfSlug": "mistralai/Mistral-Small-3.1-24B-Instruct-2503", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-13T21:50:25.140801+00:00", + "createdAt": "2025-03-17T19:15:37.00423+00:00", "hfUpdatedAt": null, - "name": "Google: Gemma 3 12B", - "shortName": "Gemma 3 12B", - "author": "google", - "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Gemma 3 12B is the second largest in the family of Gemma 3 models after [Gemma 3 27B](google/gemma-3-27b-it)", - "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", - "contextLength": 131072, + "name": "Mistral: Mistral Small 3.1 24B", + "shortName": "Mistral Small 3.1 24B", + "author": "mistralai", + "description": "Mistral Small 3.1 24B Instruct is an upgraded variant of Mistral Small 3 (2501), featuring 24 billion parameters with advanced multimodal capabilities. It provides state-of-the-art performance in text-based reasoning and vision tasks, including image analysis, programming, mathematical reasoning, and multilingual support across dozens of languages. Equipped with an extensive 128k token context window and optimized for efficient local inference, it supports use cases such as conversational agents, function calling, long-document comprehension, and privacy-sensitive deployments.", + "modelVersionGroupId": null, + "contextLength": 128000, "inputModalities": [ "text", "image" @@ -30264,44 +32227,37 @@ export default { "text" ], "hasTextOutput": true, - "group": "Gemini", - "instructType": "gemma", + "group": "Mistral", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "", - "", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "google/gemma-3-12b-it", + "permaslug": "mistralai/mistral-small-3.1-24b-instruct-2503", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "google/gemma-3-12b-it", - "modelVariantPermaslug": "google/gemma-3-12b-it", - "providerName": "DeepInfra", + "modelVariantSlug": "mistralai/mistral-small-3.1-24b-instruct:free", + "modelVariantPermaslug": "mistralai/mistral-small-3.1-24b-instruct-2503:free", + "providerName": "Chutes", "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", - "baseUrl": "url", - "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "DeepInfra", + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, @@ -30309,49 +32265,51 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/DeepInfra.webp" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "DeepInfra", - "providerModelId": "google/gemma-3-12b-it", - "providerGroup": "DeepInfra", - "quantization": "bf16", - "variant": "standard", - "isFree": false, + "providerDisplayName": "Chutes", + "providerModelId": "chutesai/Mistral-Small-3.1-24B-Instruct-2503", + "providerGroup": "Chutes", + "quantization": null, + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 96000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", "seed", - "min_p" + "top_k", + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true }, - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.00000005", - "completion": "0.0000001", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -30362,7 +32320,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -30376,24 +32334,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 77, - "newest": 87, - "throughputHighToLow": 176, - "latencyLowToHigh": 100, - "pricingLowToHigh": 39, - "pricingHighToLow": 220 + "topWeekly": 63, + "newest": 79, + "throughputHighToLow": 104, + "latencyLowToHigh": 33, + "pricingLowToHigh": 35, + "pricingHighToLow": 219 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", "context": 131072, "maxCompletionTokens": null, - "providerModelId": "google/gemma-3-12b-it", + "providerModelId": "mistralai/Mistral-Small-3.1-24B-Instruct-2503", "pricing": { "prompt": "0.00000005", - "completion": "0.0000001", + "completion": "0.00000015", "image": "0", "request": "0", "webSearch": "0", @@ -30407,22 +32367,91 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", "seed", - "min_p" + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" ], "inputCost": 0.05, - "outputCost": 0.1 + "outputCost": 0.15, + "throughput": 58.5635, + "latency": 635 + }, + { + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", + "quantization": "bf16", + "context": 128000, + "maxCompletionTokens": 128000, + "providerModelId": "parasail-mistral-small-31-24b-instruct", + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000003", + "image": "0.000926", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "presence_penalty", + "frequency_penalty", + "repetition_penalty", + "top_k" + ], + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 67.322, + "latency": 1379 + }, + { + "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", + "slug": "mistral", + "quantization": null, + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "mistral-small-2503", + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000003", + "image": "0.0009264", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "structured_outputs", + "seed" + ], + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 140.218, + "latency": 220 }, { "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", "slug": "cloudflare", "quantization": null, - "context": 80000, + "context": 128000, "maxCompletionTokens": null, - "providerModelId": "@cf/google/gemma-3-12b-it", + "providerModelId": "@cf/mistralai/mistral-small-3.1-24b-instruct", "pricing": { "prompt": "0.00000035", "completion": "0.00000056", @@ -30443,22 +32472,24 @@ export default { "presence_penalty" ], "inputCost": 0.35, - "outputCost": 0.56 + "outputCost": 0.56, + "throughput": 43.0765, + "latency": 390.5 } ] }, { - "slug": "deepseek/deepseek-r1-distill-llama-70b", - "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Llama-70B", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-01-23T20:12:49.780212+00:00", + "slug": "mistralai/codestral-2501", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-01-14T22:58:42.43034+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: R1 Distill Llama 70B (free)", - "shortName": "R1 Distill Llama 70B (free)", - "author": "deepseek", - "description": "DeepSeek R1 Distill Llama 70B is a distilled large language model based on [Llama-3.3-70B-Instruct](/meta-llama/llama-3.3-70b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 70.0\n- MATH-500 pass@1: 94.5\n- CodeForces Rating: 1633\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", - "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", - "contextLength": 8192, + "name": "Mistral: Codestral 2501", + "shortName": "Codestral 2501", + "author": "mistralai", + "description": "[Mistral](/mistralai)'s cutting-edge language model for coding. Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM), code correction and test generation. \n\nLearn more on their blog post: https://mistral.ai/news/codestral-2501/", + "modelVersionGroupId": null, + "contextLength": 262144, "inputModalities": [ "text" ], @@ -30466,43 +32497,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "deepseek-r1", + "group": "Mistral", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "deepseek/deepseek-r1-distill-llama-70b", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "warningMessage": "", + "permaslug": "mistralai/codestral-2501", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "c4e973c5-aae4-4830-8e81-9f950e1e3e6c", - "name": "Together | deepseek/deepseek-r1-distill-llama-70b:free", - "contextLength": 8192, + "id": "7df94044-ae68-4b4f-b3eb-748ed262b778", + "name": "Mistral | mistralai/codestral-2501", + "contextLength": 262144, "model": { - "slug": "deepseek/deepseek-r1-distill-llama-70b", - "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Llama-70B", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-01-23T20:12:49.780212+00:00", + "slug": "mistralai/codestral-2501", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-01-14T22:58:42.43034+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: R1 Distill Llama 70B", - "shortName": "R1 Distill Llama 70B", - "author": "deepseek", - "description": "DeepSeek R1 Distill Llama 70B is a distilled large language model based on [Llama-3.3-70B-Instruct](/meta-llama/llama-3.3-70b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 70.0\n- MATH-500 pass@1: 94.5\n- CodeForces Rating: 1633\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", - "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", - "contextLength": 128000, + "name": "Mistral: Codestral 2501", + "shortName": "Codestral 2501", + "author": "mistralai", + "description": "[Mistral](/mistralai)'s cutting-edge language model for coding. Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM), code correction and test generation. \n\nLearn more on their blog post: https://mistral.ai/news/codestral-2501/", + "modelVersionGroupId": null, + "contextLength": 256000, "inputModalities": [ "text" ], @@ -30510,50 +32530,40 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "deepseek-r1", + "group": "Mistral", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "deepseek/deepseek-r1-distill-llama-70b", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "warningMessage": "", + "permaslug": "mistralai/codestral-2501", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "deepseek/deepseek-r1-distill-llama-70b:free", - "modelVariantPermaslug": "deepseek/deepseek-r1-distill-llama-70b:free", - "providerName": "Together", + "modelVariantSlug": "mistralai/codestral-2501", + "modelVariantPermaslug": "mistralai/codestral-2501", + "providerName": "Mistral", "providerInfo": { - "name": "Together", - "displayName": "Together", - "slug": "together", + "name": "Mistral", + "displayName": "Mistral", + "slug": "mistral", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 } }, - "headquarters": "US", + "headquarters": "FR", "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, + "hasCompletions": false, + "isAbortable": false, "moderationRequired": false, - "group": "Together", + "group": "Mistral", "editors": [], "owners": [], "isMultipartSupported": true, @@ -30561,50 +32571,50 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" + "url": "/images/icons/Mistral.png" } }, - "providerDisplayName": "Together", - "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Llama-70B-free", - "providerGroup": "Together", + "providerDisplayName": "Mistral", + "providerModelId": "codestral-2501", + "providerGroup": "Mistral", "quantization": null, - "variant": "free", - "isFree": true, - "canAbort": true, + "variant": "standard", + "isFree": false, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "response_format", + "structured_outputs", + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.0000003", + "completion": "0.0000009", "image": "0", "request": "0", "webSearch": "0", @@ -30615,71 +32625,40 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": true, + "supportsToolParameters": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 21, - "newest": 141, - "throughputHighToLow": 170, - "latencyLowToHigh": 90, - "pricingLowToHigh": 52, - "pricingHighToLow": 191 + "topWeekly": 96, + "newest": 146, + "throughputHighToLow": 74, + "latencyLowToHigh": 32, + "pricingLowToHigh": 171, + "pricingHighToLow": 147 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 16384, - "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Llama-70B", - "pricing": { - "prompt": "0.0000001", - "completion": "0.0000004", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "inputCost": 0.1, - "outputCost": 0.4 - }, - { - "name": "inference.net", - "slug": "inferenceNet", + "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", + "slug": "mistral", "quantization": null, - "context": 128000, - "maxCompletionTokens": 16384, - "providerModelId": "deepseek/r1-distill-llama-70b/fp-8", + "context": 262144, + "maxCompletionTokens": null, + "providerModelId": "codestral-2501", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000004", + "prompt": "0.0000003", + "completion": "0.0000009", "image": "0", "request": "0", "webSearch": "0", @@ -30687,102 +32666,204 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "min_p", - "logit_bias", - "top_logprobs", "response_format", - "structured_outputs" + "structured_outputs", + "seed" ], - "inputCost": 0.1, - "outputCost": 0.4 - }, - { - "name": "Lambda", - "slug": "lambda", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "deepseek-llama3.3-70b", - "pricing": { - "prompt": "0.0000002", - "completion": "0.0000006", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format" + "inputCost": 0.3, + "outputCost": 0.9, + "throughput": 104.0225, + "latency": 327 + } + ] + }, + { + "slug": "qwen/qwen2.5-coder-7b-instruct", + "hfSlug": "Qwen/Qwen2.5-Coder-7B-Instruct", + "updatedAt": "2025-04-15T16:36:45.293973+00:00", + "createdAt": "2025-04-15T16:34:47.067646+00:00", + "hfUpdatedAt": null, + "name": "Qwen: Qwen2.5 Coder 7B Instruct", + "shortName": "Qwen2.5 Coder 7B Instruct", + "author": "qwen", + "description": "Qwen2.5-Coder-7B-Instruct is a 7B parameter instruction-tuned language model optimized for code-related tasks such as code generation, reasoning, and bug fixing. Based on the Qwen2.5 architecture, it incorporates enhancements like RoPE, SwiGLU, RMSNorm, and GQA attention with support for up to 128K tokens using YaRN-based extrapolation. It is trained on a large corpus of source code, synthetic data, and text-code grounding, providing robust performance across programming languages and agentic coding workflows.\n\nThis model is part of the Qwen2.5-Coder family and offers strong compatibility with tools like vLLM for efficient deployment. Released under the Apache 2.0 license.", + "modelVersionGroupId": null, + "contextLength": 32768, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Qwen", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "qwen/qwen2.5-coder-7b-instruct", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "f8e0be71-53c3-42d5-b579-7f8d7e7b55a7", + "name": "Nebius | qwen/qwen2.5-coder-7b-instruct", + "contextLength": 32768, + "model": { + "slug": "qwen/qwen2.5-coder-7b-instruct", + "hfSlug": "Qwen/Qwen2.5-Coder-7B-Instruct", + "updatedAt": "2025-04-15T16:36:45.293973+00:00", + "createdAt": "2025-04-15T16:34:47.067646+00:00", + "hfUpdatedAt": null, + "name": "Qwen: Qwen2.5 Coder 7B Instruct", + "shortName": "Qwen2.5 Coder 7B Instruct", + "author": "qwen", + "description": "Qwen2.5-Coder-7B-Instruct is a 7B parameter instruction-tuned language model optimized for code-related tasks such as code generation, reasoning, and bug fixing. Based on the Qwen2.5 architecture, it incorporates enhancements like RoPE, SwiGLU, RMSNorm, and GQA attention with support for up to 128K tokens using YaRN-based extrapolation. It is trained on a large corpus of source code, synthetic data, and text-code grounding, providing robust performance across programming languages and agentic coding workflows.\n\nThis model is part of the Qwen2.5-Coder family and offers strong compatibility with tools like vLLM for efficient deployment. Released under the Apache 2.0 license.", + "modelVersionGroupId": null, + "contextLength": 131072, + "inputModalities": [ + "text" ], - "inputCost": 0.2, - "outputCost": 0.6 - }, - { - "name": "Phala", - "slug": "phala", - "quantization": null, - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "phala/deepseek-r1-70b", - "pricing": { - "prompt": "0.0000002", - "completion": "0.0000007", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty" + "outputModalities": [ + "text" ], - "inputCost": 0.2, - "outputCost": 0.7 + "hasTextOutput": true, + "group": "Qwen", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "qwen/qwen2.5-coder-7b-instruct", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "qwen/qwen2.5-coder-7b-instruct", + "modelVariantPermaslug": "qwen/qwen2.5-coder-7b-instruct", + "providerName": "Nebius", + "providerInfo": { + "name": "Nebius", + "displayName": "Nebius AI Studio", + "slug": "nebius", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "paidModels": { + "training": false, + "retainsPrompts": false + } + }, + "headquarters": "DE", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": false, + "moderationRequired": false, + "group": "Nebius", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "invertRequired": true + } + }, + "providerDisplayName": "Nebius AI Studio", + "providerModelId": "Qwen/Qwen2.5-Coder-7B-Instruct", + "providerGroup": "Nebius", + "quantization": "fp8", + "variant": "standard", + "isFree": false, + "canAbort": false, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "paidModels": { + "training": false, + "retainsPrompts": false + }, + "training": false, + "retainsPrompts": false + }, + "pricing": { + "prompt": "0.00000001", + "completion": "0.00000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 97, + "newest": 47, + "throughputHighToLow": 10, + "latencyLowToHigh": 8, + "pricingLowToHigh": 74, + "pricingHighToLow": 244 + }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", + "providers": [ { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", - "context": 131072, + "context": 32768, "maxCompletionTokens": null, - "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Llama-70B", + "providerModelId": "Qwen/Qwen2.5-Coder-7B-Instruct", "pricing": { - "prompt": "0.00000025", - "completion": "0.00000075", + "prompt": "0.00000001", + "completion": "0.00000003", "image": "0", "request": "0", "webSearch": "0", @@ -30793,8 +32874,6 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -30804,47 +32883,199 @@ export default { "logprobs", "top_logprobs" ], - "inputCost": 0.25, - "outputCost": 0.75 + "inputCost": 0.01, + "outputCost": 0.03, + "throughput": 224.5315, + "latency": 202 + } + ] + }, + { + "slug": "sao10k/l3-lunaris-8b", + "hfSlug": "Sao10K/L3-8B-Lunaris-v1", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-08-13T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Sao10K: Llama 3 8B Lunaris", + "shortName": "Llama 3 8B Lunaris", + "author": "sao10k", + "description": "Lunaris 8B is a versatile generalist and roleplaying model based on Llama 3. It's a strategic merge of multiple models, designed to balance creativity with improved logic and general knowledge.\n\nCreated by [Sao10k](https://huggingface.co/Sao10k), this model aims to offer an improved experience over Stheno v3.2, with enhanced creativity and logical reasoning.\n\nFor best results, use with Llama 3 Instruct context template, temperature 1.4, and min_p 0.1.", + "modelVersionGroupId": null, + "contextLength": 8192, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Llama3", + "instructType": "llama3", + "defaultSystem": null, + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "sao10k/l3-lunaris-8b", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "cc4c8dc5-8b3f-4d54-84e2-8381184ff841", + "name": "DeepInfra | sao10k/l3-lunaris-8b", + "contextLength": 8192, + "model": { + "slug": "sao10k/l3-lunaris-8b", + "hfSlug": "Sao10K/L3-8B-Lunaris-v1", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-08-13T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Sao10K: Llama 3 8B Lunaris", + "shortName": "Llama 3 8B Lunaris", + "author": "sao10k", + "description": "Lunaris 8B is a versatile generalist and roleplaying model based on Llama 3. It's a strategic merge of multiple models, designed to balance creativity with improved logic and general knowledge.\n\nCreated by [Sao10k](https://huggingface.co/Sao10k), this model aims to offer an improved experience over Stheno v3.2, with enhanced creativity and logical reasoning.\n\nFor best results, use with Llama 3 Instruct context template, temperature 1.4, and min_p 0.1.", + "modelVersionGroupId": null, + "contextLength": 8192, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Llama3", + "instructType": "llama3", + "defaultSystem": null, + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "sao10k/l3-lunaris-8b", + "reasoningConfig": null, + "features": {} }, - { - "name": "SambaNova", - "slug": "sambaNova", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": 4096, - "providerModelId": "DeepSeek-R1-Distill-Llama-70B", - "pricing": { - "prompt": "0.0000007", - "completion": "0.0000014", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 + "modelVariantSlug": "sao10k/l3-lunaris-8b", + "modelVariantPermaslug": "sao10k/l3-lunaris-8b", + "providerName": "DeepInfra", + "providerInfo": { + "name": "DeepInfra", + "displayName": "DeepInfra", + "slug": "deepinfra", + "baseUrl": "url", + "dataPolicy": { + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", + "paidModels": { + "training": false, + "retainsPrompts": false + } }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "top_k", - "stop" - ], - "inputCost": 0.7, - "outputCost": 1.4 + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "DeepInfra", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/DeepInfra.webp" + } }, + "providerDisplayName": "DeepInfra", + "providerModelId": "Sao10K/L3-8B-Lunaris-v1-Turbo", + "providerGroup": "DeepInfra", + "quantization": "fp8", + "variant": "standard", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", + "paidModels": { + "training": false, + "retainsPrompts": false + }, + "training": false, + "retainsPrompts": false + }, + "pricing": { + "prompt": "0.00000002", + "completion": "0.00000005", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 98, + "newest": 221, + "throughputHighToLow": 115, + "latencyLowToHigh": 18, + "pricingLowToHigh": 78, + "pricingHighToLow": 240 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [ { - "name": "Groq", - "slug": "groq", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "deepseek-r1-distill-llama-70b", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "fp8", + "context": 8192, + "maxCompletionTokens": null, + "providerModelId": "Sao10K/L3-8B-Lunaris-v1-Turbo", "pricing": { - "prompt": "0.00000075", - "completion": "0.00000099", + "prompt": "0.00000002", + "completion": "0.00000005", "image": "0", "request": "0", "webSearch": "0", @@ -30852,35 +33083,34 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", + "repetition_penalty", "response_format", - "top_logprobs", - "logprobs", - "logit_bias", - "seed" + "top_k", + "seed", + "min_p" ], - "inputCost": 0.75, - "outputCost": 0.99 + "inputCost": 0.02, + "outputCost": 0.05, + "throughput": 65.3165, + "latency": 263 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": null, - "context": 32000, - "maxCompletionTokens": 32000, - "providerModelId": "deepseek/deepseek-r1-distill-llama-70b", + "context": 8192, + "maxCompletionTokens": null, + "providerModelId": "sao10k/l3-8b-lunaris", "pricing": { - "prompt": "0.0000008", - "completion": "0.0000008", + "prompt": "0.00000005", + "completion": "0.00000005", "image": "0", "request": "0", "webSearch": "0", @@ -30891,8 +33121,6 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -30902,76 +33130,10 @@ export default { "repetition_penalty", "logit_bias" ], - "inputCost": 0.8, - "outputCost": 0.8 - }, - { - "name": "Together", - "slug": "together", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 32768, - "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Llama-70B", - "pricing": { - "prompt": "0.000002", - "completion": "0.000002", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" - ], - "inputCost": 2, - "outputCost": 2 - }, - { - "name": "Cerebras", - "slug": "cerebras", - "quantization": null, - "context": 32000, - "maxCompletionTokens": 32000, - "providerModelId": "deepseek-r1-distill-llama-70b", - "pricing": { - "prompt": "0.0000022", - "completion": "0.0000025", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "structured_outputs", - "response_format", - "stop", - "seed", - "logprobs", - "top_logprobs" - ], - "inputCost": 2.2, - "outputCost": 2.5 + "inputCost": 0.05, + "outputCost": 0.05, + "throughput": 75.846, + "latency": 737 } ] }, @@ -31155,16 +33317,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 96, + "topWeekly": 99, "newest": 24, "throughputHighToLow": 27, - "latencyLowToHigh": 116, + "latencyLowToHigh": 132, "pricingLowToHigh": 10, "pricingHighToLow": 229 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "fp8", "context": 128000, @@ -31195,22 +33359,24 @@ export default { "logit_bias" ], "inputCost": 0.04, - "outputCost": 0.14 + "outputCost": 0.14, + "throughput": 89.547, + "latency": 751 } ] }, { - "slug": "mistralai/mistral-7b-instruct", - "hfSlug": "mistralai/Mistral-7B-Instruct-v0.3", + "slug": "microsoft/phi-4", + "hfSlug": "microsoft/phi-4", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-05-27T00:00:00+00:00", + "createdAt": "2025-01-10T06:17:52.16346+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral 7B Instruct", - "shortName": "Mistral 7B Instruct", - "author": "mistralai", - "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*", - "modelVersionGroupId": "1d07cc56-c54d-4587-b785-5093496397a4", - "contextLength": 32768, + "name": "Microsoft: Phi 4", + "shortName": "Phi 4", + "author": "microsoft", + "description": "[Microsoft Research](/microsoft) Phi-4 is designed to perform well in complex reasoning tasks and can operate efficiently in situations with limited memory or where quick responses are needed. \n\nAt 14 billion parameters, it was trained on a mix of high-quality synthetic datasets, data from curated websites, and academic materials. It has undergone careful improvement to follow instructions accurately and maintain strong safety standards. It works best with English language inputs.\n\nFor more information, please see [Phi-4 Technical Report](https://arxiv.org/pdf/2412.08905)\n", + "modelVersionGroupId": null, + "contextLength": 16384, "inputModalities": [ "text" ], @@ -31218,35 +33384,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "mistral", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "[INST]", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-7b-instruct", + "permaslug": "microsoft/phi-4", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "54a41d98-d92a-4ec1-a68c-a1da232e0946", - "name": "Enfer | mistralai/mistral-7b-instruct", - "contextLength": 32768, + "id": "8e48942d-17fc-4d00-a234-4723034fd971", + "name": "DeepInfra | microsoft/phi-4", + "contextLength": 16384, "model": { - "slug": "mistralai/mistral-7b-instruct", - "hfSlug": "mistralai/Mistral-7B-Instruct-v0.3", + "slug": "microsoft/phi-4", + "hfSlug": "microsoft/phi-4", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-05-27T00:00:00+00:00", + "createdAt": "2025-01-10T06:17:52.16346+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral 7B Instruct", - "shortName": "Mistral 7B Instruct", - "author": "mistralai", - "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*", - "modelVersionGroupId": "1d07cc56-c54d-4587-b785-5093496397a4", - "contextLength": 32768, + "name": "Microsoft: Phi 4", + "shortName": "Phi 4", + "author": "microsoft", + "description": "[Microsoft Research](/microsoft) Phi-4 is designed to perform well in complex reasoning tasks and can operate efficiently in situations with limited memory or where quick responses are needed. \n\nAt 14 billion parameters, it was trained on a mix of high-quality synthetic datasets, data from curated websites, and academic materials. It has undergone careful improvement to follow instructions accurately and maintain strong safety standards. It works best with English language inputs.\n\nFor more information, please see [Phi-4 Technical Report](https://arxiv.org/pdf/2412.08905)\n", + "modelVersionGroupId": null, + "contextLength": 16384, "inputModalities": [ "text" ], @@ -31254,40 +33417,40 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "mistral", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "[INST]", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-7b-instruct", + "permaslug": "microsoft/phi-4", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "mistralai/mistral-7b-instruct", - "modelVariantPermaslug": "mistralai/mistral-7b-instruct", - "providerName": "Enfer", - "providerInfo": { - "name": "Enfer", - "displayName": "Enfer", - "slug": "enfer", + "modelVariantSlug": "microsoft/phi-4", + "modelVariantPermaslug": "microsoft/phi-4", + "providerName": "DeepInfra", + "providerInfo": { + "name": "DeepInfra", + "displayName": "DeepInfra", + "slug": "deepinfra", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://enfer.ai/privacy-policy", - "privacyPolicyUrl": "https://enfer.ai/privacy-policy", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false } }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Enfer", + "group": "DeepInfra", "editors": [], "owners": [], "isMultipartSupported": true, @@ -31295,17 +33458,17 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://enfer.ai/&size=256" + "url": "/images/icons/DeepInfra.webp" } }, - "providerDisplayName": "Enfer", - "providerModelId": "mistralai/mistral-7b-instruct-v0-3", - "providerGroup": "Enfer", - "quantization": null, + "providerDisplayName": "DeepInfra", + "providerModelId": "microsoft/phi-4", + "providerGroup": "DeepInfra", + "quantization": "bf16", "variant": "standard", "isFree": false, "canAbort": true, - "maxPromptTokens": null, + "maxPromptTokens": 4096, "maxCompletionTokens": 16384, "maxPromptImages": null, "maxTokensPerImage": null, @@ -31316,22 +33479,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "logit_bias", - "logprobs" + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://enfer.ai/privacy-policy", - "privacyPolicyUrl": "https://enfer.ai/privacy-policy", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false }, - "training": false + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0.000000028", - "completion": "0.000000054", + "prompt": "0.00000007", + "completion": "0.00000014", "image": "0", "request": "0", "webSearch": "0", @@ -31350,30 +33519,31 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 97, - "newest": 251, - "throughputHighToLow": 66, - "latencyLowToHigh": 143, - "pricingLowToHigh": 70, - "pricingHighToLow": 235 + "topWeekly": 100, + "newest": 147, + "throughputHighToLow": 215, + "latencyLowToHigh": 114, + "pricingLowToHigh": 104, + "pricingHighToLow": 213 }, + "authorIcon": "https://openrouter.ai/images/icons/Microsoft.svg", "providers": [ { - "name": "Enfer", - "slug": "enfer", - "quantization": null, - "context": 32768, + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 16384, "maxCompletionTokens": 16384, - "providerModelId": "mistralai/mistral-7b-instruct-v0-3", + "providerModelId": "microsoft/phi-4", "pricing": { - "prompt": "0.000000028", - "completion": "0.000000054", + "prompt": "0.00000007", + "completion": "0.00000014", "image": "0", "request": "0", "webSearch": "0", @@ -31387,22 +33557,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "logit_bias", - "logprobs" + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" ], - "inputCost": 0.03, - "outputCost": 0.05 + "inputCost": 0.07, + "outputCost": 0.14, + "throughput": 39.4525, + "latency": 735.5 }, { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": 16384, - "providerModelId": "mistralai/Mistral-7B-Instruct-v0.3", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 16384, + "maxCompletionTokens": null, + "providerModelId": "microsoft/phi-4", "pricing": { - "prompt": "0.000000029", - "completion": "0.000000055", + "prompt": "0.0000001", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -31410,33 +33586,212 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", "seed", - "min_p" + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 0.03, - "outputCost": 0.06 + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 114.82, + "latency": 335 + } + ] + }, + { + "slug": "google/gemma-2-9b-it", + "hfSlug": "google/gemma-2-9b-it", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-06-28T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Google: Gemma 2 9B", + "shortName": "Gemma 2 9B", + "author": "google", + "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).", + "modelVersionGroupId": null, + "contextLength": 8192, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Gemini", + "instructType": "gemma", + "defaultSystem": null, + "defaultStops": [ + "", + "", + "" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "google/gemma-2-9b-it", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "01c4c94a-14b7-4db3-a04b-c68e41a8df38", + "name": "Nebius | google/gemma-2-9b-it", + "contextLength": 8192, + "model": { + "slug": "google/gemma-2-9b-it", + "hfSlug": "google/gemma-2-9b-it", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-06-28T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Google: Gemma 2 9B", + "shortName": "Gemma 2 9B", + "author": "google", + "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).", + "modelVersionGroupId": null, + "contextLength": 8192, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Gemini", + "instructType": "gemma", + "defaultSystem": null, + "defaultStops": [ + "", + "", + "" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "google/gemma-2-9b-it", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "google/gemma-2-9b-it", + "modelVariantPermaslug": "google/gemma-2-9b-it", + "providerName": "Nebius", + "providerInfo": { + "name": "Nebius", + "displayName": "Nebius AI Studio", + "slug": "nebius", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "paidModels": { + "training": false, + "retainsPrompts": false + } + }, + "headquarters": "DE", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": false, + "moderationRequired": false, + "group": "Nebius", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "invertRequired": true + } + }, + "providerDisplayName": "Nebius AI Studio", + "providerModelId": "google/gemma-2-9b-it", + "providerGroup": "Nebius", + "quantization": "fp8", + "variant": "standard", + "isFree": false, + "canAbort": false, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "paidModels": { + "training": false, + "retainsPrompts": false + }, + "training": false, + "retainsPrompts": false + }, + "pricing": { + "prompt": "0.00000002", + "completion": "0.00000006", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 101, + "newest": 240, + "throughputHighToLow": 40, + "latencyLowToHigh": 6, + "pricingLowToHigh": 69, + "pricingHighToLow": 239 + }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", + "providers": [ { - "name": "NextBit", - "slug": "nextBit", - "quantization": "bf16", - "context": 32768, + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 8192, "maxCompletionTokens": null, - "providerModelId": "mistral:7b", + "providerModelId": "google/gemma-2-9b-it", "pricing": { - "prompt": "0.000000029", - "completion": "0.000000058", + "prompt": "0.00000002", + "completion": "0.00000006", "image": "0", "request": "0", "webSearch": "0", @@ -31450,22 +33805,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "response_format", - "structured_outputs" + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 0.03, - "outputCost": 0.06 + "inputCost": 0.02, + "outputCost": 0.06, + "throughput": 142.233, + "latency": 181 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "unknown", - "context": 32768, + "context": 8192, "maxCompletionTokens": null, - "providerModelId": "mistralai/mistral-7b-instruct", + "providerModelId": "google/gemma-2-9b-it", "pricing": { - "prompt": "0.000000029", - "completion": "0.000000059", + "prompt": "0.00000008", + "completion": "0.00000008", "image": "0", "request": "0", "webSearch": "0", @@ -31485,19 +33846,22 @@ export default { "repetition_penalty", "logit_bias" ], - "inputCost": 0.03, - "outputCost": 0.06 + "inputCost": 0.08, + "outputCost": 0.08, + "throughput": 32.0005, + "latency": 661 }, { - "name": "Parasail", - "slug": "parasail", - "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": 32768, - "providerModelId": "parasail-mistral-7b-instruct-03", + "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", + "slug": "groq", + "quantization": null, + "context": 8192, + "maxCompletionTokens": 8192, + "providerModelId": "gemma2-9b-it", "pricing": { - "prompt": "0.00000011", - "completion": "0.00000011", + "prompt": "0.0000002", + "completion": "0.0000002", "image": "0", "request": "0", "webSearch": "0", @@ -31508,24 +33872,31 @@ export default { "max_tokens", "temperature", "top_p", - "presence_penalty", + "stop", "frequency_penalty", - "repetition_penalty", - "top_k" + "presence_penalty", + "response_format", + "top_logprobs", + "logprobs", + "logit_bias", + "seed" ], - "inputCost": 0.11, - "outputCost": 0.11 + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 849.364, + "latency": 330 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", - "quantization": "unknown", - "context": 32768, - "maxCompletionTokens": 4096, - "providerModelId": "mistralai/Mistral-7B-Instruct-v0.3", + "quantization": null, + "context": 8192, + "maxCompletionTokens": 8192, + "providerModelId": "google/gemma-2-9b-it", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", + "prompt": "0.0000003", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -31545,23 +33916,25 @@ export default { "min_p", "response_format" ], - "inputCost": 0.2, - "outputCost": 0.2 + "inputCost": 0.3, + "outputCost": 0.3, + "throughput": 116.9945, + "latency": 345 } ] }, { - "slug": "sao10k/l3-lunaris-8b", - "hfSlug": "Sao10K/L3-8B-Lunaris-v1", + "slug": "sao10k/l3.3-euryale-70b", + "hfSlug": "Sao10K/L3.3-70B-Euryale-v2.3", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-13T00:00:00+00:00", + "createdAt": "2024-12-18T15:32:08.468786+00:00", "hfUpdatedAt": null, - "name": "Sao10K: Llama 3 8B Lunaris", - "shortName": "Llama 3 8B Lunaris", + "name": "Sao10K: Llama 3.3 Euryale 70B", + "shortName": "Llama 3.3 Euryale 70B", "author": "sao10k", - "description": "Lunaris 8B is a versatile generalist and roleplaying model based on Llama 3. It's a strategic merge of multiple models, designed to balance creativity with improved logic and general knowledge.\n\nCreated by [Sao10k](https://huggingface.co/Sao10k), this model aims to offer an improved experience over Stheno v3.2, with enhanced creativity and logical reasoning.\n\nFor best results, use with Llama 3 Instruct context template, temperature 1.4, and min_p 0.1.", + "description": "Euryale L3.3 70B is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k). It is the successor of [Euryale L3 70B v2.2](/models/sao10k/l3-euryale-70b).", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -31579,23 +33952,23 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "sao10k/l3-lunaris-8b", + "permaslug": "sao10k/l3.3-euryale-70b-v2.3", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "cc4c8dc5-8b3f-4d54-84e2-8381184ff841", - "name": "DeepInfra | sao10k/l3-lunaris-8b", - "contextLength": 8192, + "id": "6e3850b8-2305-4bda-990a-7aa06427bc83", + "name": "DeepInfra | sao10k/l3.3-euryale-70b-v2.3", + "contextLength": 131072, "model": { - "slug": "sao10k/l3-lunaris-8b", - "hfSlug": "Sao10K/L3-8B-Lunaris-v1", + "slug": "sao10k/l3.3-euryale-70b", + "hfSlug": "Sao10K/L3.3-70B-Euryale-v2.3", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-13T00:00:00+00:00", + "createdAt": "2024-12-18T15:32:08.468786+00:00", "hfUpdatedAt": null, - "name": "Sao10K: Llama 3 8B Lunaris", - "shortName": "Llama 3 8B Lunaris", + "name": "Sao10K: Llama 3.3 Euryale 70B", + "shortName": "Llama 3.3 Euryale 70B", "author": "sao10k", - "description": "Lunaris 8B is a versatile generalist and roleplaying model based on Llama 3. It's a strategic merge of multiple models, designed to balance creativity with improved logic and general knowledge.\n\nCreated by [Sao10k](https://huggingface.co/Sao10k), this model aims to offer an improved experience over Stheno v3.2, with enhanced creativity and logical reasoning.\n\nFor best results, use with Llama 3 Instruct context template, temperature 1.4, and min_p 0.1.", + "description": "Euryale L3.3 70B is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k). It is the successor of [Euryale L3 70B v2.2](/models/sao10k/l3-euryale-70b).", "modelVersionGroupId": null, "contextLength": 8192, "inputModalities": [ @@ -31615,12 +33988,12 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "sao10k/l3-lunaris-8b", + "permaslug": "sao10k/l3.3-euryale-70b-v2.3", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "sao10k/l3-lunaris-8b", - "modelVariantPermaslug": "sao10k/l3-lunaris-8b", + "modelVariantSlug": "sao10k/l3.3-euryale-70b", + "modelVariantPermaslug": "sao10k/l3.3-euryale-70b-v2.3", "providerName": "DeepInfra", "providerInfo": { "name": "DeepInfra", @@ -31653,14 +34026,14 @@ export default { } }, "providerDisplayName": "DeepInfra", - "providerModelId": "Sao10K/L3-8B-Lunaris-v1-Turbo", + "providerModelId": "Sao10K/L3.3-70B-Euryale-v2.3", "providerGroup": "DeepInfra", "quantization": "fp8", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 16384, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -31690,8 +34063,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000002", - "completion": "0.00000005", + "prompt": "0.0000007", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -31710,30 +34083,31 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 98, - "newest": 221, - "throughputHighToLow": 133, - "latencyLowToHigh": 22, - "pricingLowToHigh": 78, - "pricingHighToLow": 240 + "topWeekly": 102, + "newest": 150, + "throughputHighToLow": 211, + "latencyLowToHigh": 14, + "pricingLowToHigh": 196, + "pricingHighToLow": 121 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", - "context": 8192, - "maxCompletionTokens": null, - "providerModelId": "Sao10K/L3-8B-Lunaris-v1-Turbo", + "context": 131072, + "maxCompletionTokens": 16384, + "providerModelId": "Sao10K/L3.3-70B-Euryale-v2.3", "pricing": { - "prompt": "0.00000002", - "completion": "0.00000005", + "prompt": "0.0000007", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -31753,19 +34127,22 @@ export default { "seed", "min_p" ], - "inputCost": 0.02, - "outputCost": 0.05 + "inputCost": 0.7, + "outputCost": 0.8, + "throughput": 40.024, + "latency": 229 }, { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": null, - "context": 8192, + "name": "Infermatic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://infermatic.ai/&size=256", + "slug": "infermatic", + "quantization": "fp8", + "context": 16384, "maxCompletionTokens": null, - "providerModelId": "sao10k/l3-8b-lunaris", + "providerModelId": "Sao10K-L3.3-70B-Euryale-v2.3-FP8-Dynamic", "pricing": { - "prompt": "0.00000005", - "completion": "0.00000005", + "prompt": "0.0000015", + "completion": "0.0000015", "image": "0", "request": "0", "webSearch": "0", @@ -31779,102 +34156,113 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", + "repetition_penalty", + "logit_bias", "top_k", "min_p", - "repetition_penalty", - "logit_bias" + "seed" ], - "inputCost": 0.05, - "outputCost": 0.05 + "inputCost": 1.5, + "outputCost": 1.5, + "throughput": 49.592, + "latency": 554 } ] }, { - "slug": "qwen/qwen2.5-coder-7b-instruct", - "hfSlug": "Qwen/Qwen2.5-Coder-7B-Instruct", - "updatedAt": "2025-04-15T16:36:45.293973+00:00", - "createdAt": "2025-04-15T16:34:47.067646+00:00", + "slug": "meta-llama/llama-3.2-11b-vision-instruct", + "hfSlug": "meta-llama/Llama-3.2-11B-Vision-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-09-25T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen2.5 Coder 7B Instruct", - "shortName": "Qwen2.5 Coder 7B Instruct", - "author": "qwen", - "description": "Qwen2.5-Coder-7B-Instruct is a 7B parameter instruction-tuned language model optimized for code-related tasks such as code generation, reasoning, and bug fixing. Based on the Qwen2.5 architecture, it incorporates enhancements like RoPE, SwiGLU, RMSNorm, and GQA attention with support for up to 128K tokens using YaRN-based extrapolation. It is trained on a large corpus of source code, synthetic data, and text-code grounding, providing robust performance across programming languages and agentic coding workflows.\n\nThis model is part of the Qwen2.5-Coder family and offers strong compatibility with tools like vLLM for efficient deployment. Released under the Apache 2.0 license.", + "name": "Meta: Llama 3.2 11B Vision Instruct", + "shortName": "Llama 3.2 11B Vision Instruct", + "author": "meta-llama", + "description": "Llama 3.2 11B Vision is a multimodal model with 11 billion parameters, designed to handle tasks combining visual and textual data. It excels in tasks such as image captioning and visual question answering, bridging the gap between language generation and visual reasoning. Pre-trained on a massive dataset of image-text pairs, it performs well in complex, high-accuracy image analysis.\n\nIts ability to integrate visual understanding with language processing makes it an ideal solution for industries requiring comprehensive visual-linguistic AI applications, such as content creation, AI-driven customer service, and research.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 131072, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": null, + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen2.5-coder-7b-instruct", + "permaslug": "meta-llama/llama-3.2-11b-vision-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "f8e0be71-53c3-42d5-b579-7f8d7e7b55a7", - "name": "Nebius | qwen/qwen2.5-coder-7b-instruct", - "contextLength": 32768, + "id": "4a07b512-e030-412d-b1d6-39773a8b8dcf", + "name": "DeepInfra | meta-llama/llama-3.2-11b-vision-instruct", + "contextLength": 131072, "model": { - "slug": "qwen/qwen2.5-coder-7b-instruct", - "hfSlug": "Qwen/Qwen2.5-Coder-7B-Instruct", - "updatedAt": "2025-04-15T16:36:45.293973+00:00", - "createdAt": "2025-04-15T16:34:47.067646+00:00", + "slug": "meta-llama/llama-3.2-11b-vision-instruct", + "hfSlug": "meta-llama/Llama-3.2-11B-Vision-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-09-25T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen2.5 Coder 7B Instruct", - "shortName": "Qwen2.5 Coder 7B Instruct", - "author": "qwen", - "description": "Qwen2.5-Coder-7B-Instruct is a 7B parameter instruction-tuned language model optimized for code-related tasks such as code generation, reasoning, and bug fixing. Based on the Qwen2.5 architecture, it incorporates enhancements like RoPE, SwiGLU, RMSNorm, and GQA attention with support for up to 128K tokens using YaRN-based extrapolation. It is trained on a large corpus of source code, synthetic data, and text-code grounding, providing robust performance across programming languages and agentic coding workflows.\n\nThis model is part of the Qwen2.5-Coder family and offers strong compatibility with tools like vLLM for efficient deployment. Released under the Apache 2.0 license.", + "name": "Meta: Llama 3.2 11B Vision Instruct", + "shortName": "Llama 3.2 11B Vision Instruct", + "author": "meta-llama", + "description": "Llama 3.2 11B Vision is a multimodal model with 11 billion parameters, designed to handle tasks combining visual and textual data. It excels in tasks such as image captioning and visual question answering, bridging the gap between language generation and visual reasoning. Pre-trained on a massive dataset of image-text pairs, it performs well in complex, high-accuracy image analysis.\n\nIts ability to integrate visual understanding with language processing makes it an ideal solution for industries requiring comprehensive visual-linguistic AI applications, such as content creation, AI-driven customer service, and research.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": null, + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen2.5-coder-7b-instruct", + "permaslug": "meta-llama/llama-3.2-11b-vision-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "qwen/qwen2.5-coder-7b-instruct", - "modelVariantPermaslug": "qwen/qwen2.5-coder-7b-instruct", - "providerName": "Nebius", + "modelVariantSlug": "meta-llama/llama-3.2-11b-vision-instruct", + "modelVariantPermaslug": "meta-llama/llama-3.2-11b-vision-instruct", + "providerName": "DeepInfra", "providerInfo": { - "name": "Nebius", - "displayName": "Nebius AI Studio", - "slug": "nebius", + "name": "DeepInfra", + "displayName": "DeepInfra", + "slug": "deepinfra", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { "training": false, "retainsPrompts": false } }, - "headquarters": "DE", + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, + "isAbortable": true, "moderationRequired": false, - "group": "Nebius", + "group": "DeepInfra", "editors": [], "owners": [], "isMultipartSupported": true, @@ -31882,19 +34270,18 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", - "invertRequired": true + "url": "/images/icons/DeepInfra.webp" } }, - "providerDisplayName": "Nebius AI Studio", - "providerModelId": "Qwen/Qwen2.5-Coder-7B-Instruct", - "providerGroup": "Nebius", - "quantization": "fp8", + "providerDisplayName": "DeepInfra", + "providerModelId": "meta-llama/Llama-3.2-11B-Vision-Instruct", + "providerGroup": "DeepInfra", + "quantization": "bf16", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 16384, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -31904,17 +34291,18 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", + "repetition_penalty", + "response_format", "top_k", - "logit_bias", - "logprobs", - "top_logprobs" + "seed", + "min_p" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { "training": false, "retainsPrompts": false @@ -31923,9 +34311,9 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000001", - "completion": "0.00000003", - "image": "0", + "prompt": "0.000000049", + "completion": "0.000000049", + "image": "0.00007948", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -31943,30 +34331,168 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 99, - "newest": 47, - "throughputHighToLow": 12, - "latencyLowToHigh": 13, - "pricingLowToHigh": 74, - "pricingHighToLow": 244 + "topWeekly": 103, + "newest": 202, + "throughputHighToLow": 54, + "latencyLowToHigh": 157, + "pricingLowToHigh": 62, + "pricingHighToLow": 223 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/Llama-3.2-11B-Vision-Instruct", + "pricing": { + "prompt": "0.000000049", + "completion": "0.000000049", + "image": "0.00007948", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" + ], + "inputCost": 0.05, + "outputCost": 0.05, + "throughput": 12.962, + "latency": 2799 + }, + { + "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", + "slug": "cloudflare", + "quantization": null, + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "@cf/meta/llama-3.2-11b-vision-instruct", + "pricing": { + "prompt": "0.000000049", + "completion": "0.00000068", + "image": "0.001281", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "top_k", + "seed", + "repetition_penalty", + "frequency_penalty", + "presence_penalty" + ], + "inputCost": 0.05, + "outputCost": 0.68, + "throughput": 36.1265, + "latency": 287 + }, + { + "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", + "slug": "lambda", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "llama3.2-11b-vision-instruct", + "pricing": { + "prompt": "0.00000005", + "completion": "0.00000005", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" + ], + "inputCost": 0.05, + "outputCost": 0.05, + "throughput": 119.7235, + "latency": 251.5 + }, + { + "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", + "slug": "inferenceNet", + "quantization": "fp16", + "context": 16384, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/llama-3.2-11b-instruct/fp-16", + "pricing": { + "prompt": "0.000000055", + "completion": "0.000000055", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "logit_bias", + "top_logprobs" + ], + "inputCost": 0.06, + "outputCost": 0.06, + "throughput": 35.252, + "latency": 1401.5 + }, + { + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", "quantization": "fp8", "context": 32768, "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen2.5-Coder-7B-Instruct", + "providerModelId": "meta-llama/llama-3.2-11b-vision-instruct", "pricing": { - "prompt": "0.00000001", - "completion": "0.00000003", + "prompt": "0.00000006", + "completion": "0.00000006", "image": "0", "request": "0", "webSearch": "0", @@ -31982,27 +34508,64 @@ export default { "presence_penalty", "seed", "top_k", + "min_p", + "repetition_penalty", + "logit_bias" + ], + "inputCost": 0.06, + "outputCost": 0.06, + "throughput": 60.813, + "latency": 701 + }, + { + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo", + "pricing": { + "prompt": "0.00000018", + "completion": "0.00000018", + "image": "0.001156", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", "logit_bias", - "logprobs", - "top_logprobs" + "min_p", + "response_format" ], - "inputCost": 0.01, - "outputCost": 0.03 + "inputCost": 0.18, + "outputCost": 0.18, + "throughput": 140.829, + "latency": 405 } ] }, { - "slug": "microsoft/phi-4", - "hfSlug": "microsoft/phi-4", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-01-10T06:17:52.16346+00:00", + "slug": "deepseek/deepseek-r1-distill-llama-70b", + "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Llama-70B", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-01-23T20:12:49.780212+00:00", "hfUpdatedAt": null, - "name": "Microsoft: Phi 4", - "shortName": "Phi 4", - "author": "microsoft", - "description": "[Microsoft Research](/microsoft) Phi-4 is designed to perform well in complex reasoning tasks and can operate efficiently in situations with limited memory or where quick responses are needed. \n\nAt 14 billion parameters, it was trained on a mix of high-quality synthetic datasets, data from curated websites, and academic materials. It has undergone careful improvement to follow instructions accurately and maintain strong safety standards. It works best with English language inputs.\n\nFor more information, please see [Phi-4 Technical Report](https://arxiv.org/pdf/2412.08905)\n", - "modelVersionGroupId": null, - "contextLength": 16384, + "name": "DeepSeek: R1 Distill Llama 70B (free)", + "shortName": "R1 Distill Llama 70B (free)", + "author": "deepseek", + "description": "DeepSeek R1 Distill Llama 70B is a distilled large language model based on [Llama-3.3-70B-Instruct](/meta-llama/llama-3.3-70b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 70.0\n- MATH-500 pass@1: 94.5\n- CodeForces Rating: 1633\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", + "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", + "contextLength": 8192, "inputModalities": [ "text" ], @@ -32010,32 +34573,43 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Llama3", + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "microsoft/phi-4", - "reasoningConfig": null, - "features": {}, + "permaslug": "deepseek/deepseek-r1-distill-llama-70b", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "8e48942d-17fc-4d00-a234-4723034fd971", - "name": "DeepInfra | microsoft/phi-4", - "contextLength": 16384, + "id": "c4e973c5-aae4-4830-8e81-9f950e1e3e6c", + "name": "Together | deepseek/deepseek-r1-distill-llama-70b:free", + "contextLength": 8192, "model": { - "slug": "microsoft/phi-4", - "hfSlug": "microsoft/phi-4", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-01-10T06:17:52.16346+00:00", + "slug": "deepseek/deepseek-r1-distill-llama-70b", + "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Llama-70B", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-01-23T20:12:49.780212+00:00", "hfUpdatedAt": null, - "name": "Microsoft: Phi 4", - "shortName": "Phi 4", - "author": "microsoft", - "description": "[Microsoft Research](/microsoft) Phi-4 is designed to perform well in complex reasoning tasks and can operate efficiently in situations with limited memory or where quick responses are needed. \n\nAt 14 billion parameters, it was trained on a mix of high-quality synthetic datasets, data from curated websites, and academic materials. It has undergone careful improvement to follow instructions accurately and maintain strong safety standards. It works best with English language inputs.\n\nFor more information, please see [Phi-4 Technical Report](https://arxiv.org/pdf/2412.08905)\n", - "modelVersionGroupId": null, - "contextLength": 16384, + "name": "DeepSeek: R1 Distill Llama 70B", + "shortName": "R1 Distill Llama 70B", + "author": "deepseek", + "description": "DeepSeek R1 Distill Llama 70B is a distilled large language model based on [Llama-3.3-70B-Instruct](/meta-llama/llama-3.3-70b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 70.0\n- MATH-500 pass@1: 94.5\n- CodeForces Rating: 1633\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", + "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", + "contextLength": 128000, "inputModalities": [ "text" ], @@ -32043,29 +34617,39 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Llama3", + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "microsoft/phi-4", - "reasoningConfig": null, - "features": {} + "permaslug": "deepseek/deepseek-r1-distill-llama-70b", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "microsoft/phi-4", - "modelVariantPermaslug": "microsoft/phi-4", - "providerName": "DeepInfra", + "modelVariantSlug": "deepseek/deepseek-r1-distill-llama-70b:free", + "modelVariantPermaslug": "deepseek/deepseek-r1-distill-llama-70b:free", + "providerName": "Together", "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", + "name": "Together", + "displayName": "Together", + "slug": "together", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -32076,7 +34660,7 @@ export default { "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "DeepInfra", + "group": "Together", "editors": [], "owners": [], "isMultipartSupported": true, @@ -32084,39 +34668,40 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/DeepInfra.webp" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" } }, - "providerDisplayName": "DeepInfra", - "providerModelId": "microsoft/phi-4", - "providerGroup": "DeepInfra", - "quantization": "bf16", - "variant": "standard", - "isFree": false, + "providerDisplayName": "Together", + "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Llama-70B-free", + "providerGroup": "Together", + "quantization": null, + "variant": "free", + "isFree": true, "canAbort": true, - "maxPromptTokens": 4096, - "maxCompletionTokens": 16384, + "maxPromptTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", "top_k", - "seed", - "min_p" + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -32125,8 +34710,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000007", - "completion": "0.00000014", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -32138,7 +34723,7 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, @@ -32150,24 +34735,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 100, - "newest": 147, - "throughputHighToLow": 211, - "latencyLowToHigh": 136, - "pricingLowToHigh": 104, - "pricingHighToLow": 213 + "topWeekly": 21, + "newest": 141, + "throughputHighToLow": 162, + "latencyLowToHigh": 39, + "pricingLowToHigh": 52, + "pricingHighToLow": 191 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", - "quantization": "bf16", - "context": 16384, + "quantization": "fp8", + "context": 131072, "maxCompletionTokens": 16384, - "providerModelId": "microsoft/phi-4", + "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Llama-70B", "pricing": { - "prompt": "0.00000007", - "completion": "0.00000014", + "prompt": "0.0000001", + "completion": "0.0000004", "image": "0", "request": "0", "webSearch": "0", @@ -32178,6 +34765,8 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -32187,19 +34776,22 @@ export default { "seed", "min_p" ], - "inputCost": 0.07, - "outputCost": 0.14 + "inputCost": 0.1, + "outputCost": 0.4, + "throughput": 34.986, + "latency": 459.5 }, { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 16384, - "maxCompletionTokens": null, - "providerModelId": "microsoft/phi-4", + "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", + "slug": "inferenceNet", + "quantization": null, + "context": 128000, + "maxCompletionTokens": 16384, + "providerModelId": "deepseek/r1-distill-llama-70b/fp-8", "pricing": { "prompt": "0.0000001", - "completion": "0.0000003", + "completion": "0.0000004", "image": "0", "request": "0", "webSearch": "0", @@ -32210,205 +34802,108 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", "seed", "top_k", + "min_p", "logit_bias", - "logprobs", - "top_logprobs" + "top_logprobs", + "response_format", + "structured_outputs" ], "inputCost": 0.1, - "outputCost": 0.3 - } - ] - }, - { - "slug": "google/gemma-2-9b-it", - "hfSlug": "google/gemma-2-9b-it", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-06-28T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Google: Gemma 2 9B", - "shortName": "Gemma 2 9B", - "author": "google", - "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).", - "modelVersionGroupId": null, - "contextLength": 8192, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Gemini", - "instructType": "gemma", - "defaultSystem": null, - "defaultStops": [ - "", - "", - "" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "google/gemma-2-9b-it", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "01c4c94a-14b7-4db3-a04b-c68e41a8df38", - "name": "Nebius | google/gemma-2-9b-it", - "contextLength": 8192, - "model": { - "slug": "google/gemma-2-9b-it", - "hfSlug": "google/gemma-2-9b-it", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-06-28T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Google: Gemma 2 9B", - "shortName": "Gemma 2 9B", - "author": "google", - "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).", - "modelVersionGroupId": null, - "contextLength": 8192, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Gemini", - "instructType": "gemma", - "defaultSystem": null, - "defaultStops": [ - "", - "", - "" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "google/gemma-2-9b-it", - "reasoningConfig": null, - "features": {} + "outputCost": 0.4, + "throughput": 37.947, + "latency": 754 }, - "modelVariantSlug": "google/gemma-2-9b-it", - "modelVariantPermaslug": "google/gemma-2-9b-it", - "providerName": "Nebius", - "providerInfo": { - "name": "Nebius", - "displayName": "Nebius AI Studio", - "slug": "nebius", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", - "paidModels": { - "training": false, - "retainsPrompts": false - } + { + "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", + "slug": "lambda", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "deepseek-llama3.3-70b", + "pricing": { + "prompt": "0.0000002", + "completion": "0.0000006", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 }, - "headquarters": "DE", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": false, - "moderationRequired": false, - "group": "Nebius", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", - "invertRequired": true - } + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" + ], + "inputCost": 0.2, + "outputCost": 0.6, + "throughput": 67.7255, + "latency": 444 }, - "providerDisplayName": "Nebius AI Studio", - "providerModelId": "google/gemma-2-9b-it", - "providerGroup": "Nebius", - "quantization": "fp8", - "variant": "standard", - "isFree": false, - "canAbort": false, - "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", - "paidModels": { - "training": false, - "retainsPrompts": false + { + "name": "Phala", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://phala.network/&size=256", + "slug": "phala", + "quantization": null, + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "phala/deepseek-r1-70b", + "pricing": { + "prompt": "0.0000002", + "completion": "0.0000007", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 }, - "training": false, - "retainsPrompts": false - }, - "pricing": { - "prompt": "0.00000002", - "completion": "0.00000006", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty" + ], + "inputCost": 0.2, + "outputCost": 0.7, + "throughput": 36.5945, + "latency": 608 }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 101, - "newest": 240, - "throughputHighToLow": 42, - "latencyLowToHigh": 8, - "pricingLowToHigh": 69, - "pricingHighToLow": 239 - }, - "providers": [ { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", - "context": 8192, + "context": 131072, "maxCompletionTokens": null, - "providerModelId": "google/gemma-2-9b-it", + "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Llama-70B", "pricing": { - "prompt": "0.00000002", - "completion": "0.00000006", + "prompt": "0.00000025", + "completion": "0.00000075", "image": "0", "request": "0", "webSearch": "0", @@ -32419,6 +34914,8 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -32428,19 +34925,22 @@ export default { "logprobs", "top_logprobs" ], - "inputCost": 0.02, - "outputCost": 0.06 + "inputCost": 0.25, + "outputCost": 0.75, + "throughput": 59.612, + "latency": 484 }, { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": "unknown", - "context": 8192, - "maxCompletionTokens": null, - "providerModelId": "google/gemma-2-9b-it", + "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", + "slug": "sambaNova", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": 4096, + "providerModelId": "DeepSeek-R1-Distill-Llama-70B", "pricing": { - "prompt": "0.00000008", - "completion": "0.00000008", + "prompt": "0.0000007", + "completion": "0.0000014", "image": "0", "request": "0", "webSearch": "0", @@ -32451,28 +34951,27 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", + "reasoning", + "include_reasoning", "top_k", - "min_p", - "repetition_penalty", - "logit_bias" + "stop" ], - "inputCost": 0.08, - "outputCost": 0.08 + "inputCost": 0.7, + "outputCost": 1.4, + "throughput": 232.5305, + "latency": 2045.5 }, { "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", "slug": "groq", "quantization": null, - "context": 8192, - "maxCompletionTokens": 8192, - "providerModelId": "gemma2-9b-it", + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "deepseek-r1-distill-llama-70b", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", + "prompt": "0.00000075", + "completion": "0.00000099", "image": "0", "request": "0", "webSearch": "0", @@ -32480,9 +34979,13 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -32492,19 +34995,59 @@ export default { "logit_bias", "seed" ], - "inputCost": 0.2, - "outputCost": 0.2 + "inputCost": 0.75, + "outputCost": 0.99, + "throughput": 293.145, + "latency": 881 + }, + { + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": null, + "context": 32000, + "maxCompletionTokens": 32000, + "providerModelId": "deepseek/deepseek-r1-distill-llama-70b", + "pricing": { + "prompt": "0.0000008", + "completion": "0.0000008", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" + ], + "inputCost": 0.8, + "outputCost": 0.8, + "throughput": 71.3745, + "latency": 891 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": null, - "context": 8192, - "maxCompletionTokens": 8192, - "providerModelId": "google/gemma-2-9b-it", + "context": 131072, + "maxCompletionTokens": 32768, + "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Llama-70B", "pricing": { - "prompt": "0.0000003", - "completion": "0.0000003", + "prompt": "0.000002", + "completion": "0.000002", "image": "0", "request": "0", "webSearch": "0", @@ -32515,6 +35058,8 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -32524,119 +35069,136 @@ export default { "min_p", "response_format" ], - "inputCost": 0.3, - "outputCost": 0.3 + "inputCost": 2, + "outputCost": 2, + "throughput": 101.26, + "latency": 690.5 + }, + { + "name": "Cerebras", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cerebras.ai/&size=256", + "slug": "cerebras", + "quantization": null, + "context": 32000, + "maxCompletionTokens": 32000, + "providerModelId": "deepseek-r1-distill-llama-70b", + "pricing": { + "prompt": "0.0000022", + "completion": "0.0000025", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "structured_outputs", + "response_format", + "stop", + "seed", + "logprobs", + "top_logprobs" + ], + "inputCost": 2.2, + "outputCost": 2.5, + "throughput": 2122.065, + "latency": 865 } ] }, { - "slug": "qwen/qwen3-14b", - "hfSlug": "Qwen/Qwen3-14B", - "updatedAt": "2025-05-12T00:36:13.09542+00:00", - "createdAt": "2025-04-28T21:41:18.320017+00:00", + "slug": "qwen/qwen2.5-vl-72b-instruct", + "hfSlug": "Qwen/Qwen2.5-VL-72B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-01T11:45:11.997326+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 14B", - "shortName": "Qwen3 14B", + "name": "Qwen: Qwen2.5 VL 72B Instruct (free)", + "shortName": "Qwen2.5 VL 72B Instruct (free)", "author": "qwen", - "description": "Qwen3-14B is a dense 14.8B parameter causal language model from the Qwen3 series, designed for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, programming, and logical inference, and a \"non-thinking\" mode for general-purpose conversation. The model is fine-tuned for instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling.", + "description": "Qwen2.5-VL is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.", "modelVersionGroupId": null, - "contextLength": 40960, + "contextLength": 131072, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Qwen", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen3-14b-04-28", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "permaslug": "qwen/qwen2.5-vl-72b-instruct", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "778bdebf-d018-4a24-a34e-230fce0f2045", - "name": "DeepInfra | qwen/qwen3-14b-04-28", - "contextLength": 40960, + "id": "75b071e0-0415-4543-a28f-e8d8e28d4421", + "name": "Alibaba | qwen/qwen2.5-vl-72b-instruct:free", + "contextLength": 131072, "model": { - "slug": "qwen/qwen3-14b", - "hfSlug": "Qwen/Qwen3-14B", - "updatedAt": "2025-05-12T00:36:13.09542+00:00", - "createdAt": "2025-04-28T21:41:18.320017+00:00", + "slug": "qwen/qwen2.5-vl-72b-instruct", + "hfSlug": "Qwen/Qwen2.5-VL-72B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-01T11:45:11.997326+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 14B", - "shortName": "Qwen3 14B", + "name": "Qwen: Qwen2.5 VL 72B Instruct", + "shortName": "Qwen2.5 VL 72B Instruct", "author": "qwen", - "description": "Qwen3-14B is a dense 14.8B parameter causal language model from the Qwen3 series, designed for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, programming, and logical inference, and a \"non-thinking\" mode for general-purpose conversation. The model is fine-tuned for instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling.", + "description": "Qwen2.5-VL is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.", "modelVersionGroupId": null, - "contextLength": 131702, + "contextLength": 131072, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Qwen", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen3-14b-04-28", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "permaslug": "qwen/qwen2.5-vl-72b-instruct", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "qwen/qwen3-14b", - "modelVariantPermaslug": "qwen/qwen3-14b-04-28", - "providerName": "DeepInfra", + "modelVariantSlug": "qwen/qwen2.5-vl-72b-instruct:free", + "modelVariantPermaslug": "qwen/qwen2.5-vl-72b-instruct:free", + "providerName": "Alibaba", "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", + "name": "Alibaba", + "displayName": "Alibaba", + "slug": "alibaba", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", + "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": false + "training": false } }, - "headquarters": "US", + "headquarters": "CN", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": true, + "isAbortable": false, "moderationRequired": false, - "group": "DeepInfra", + "group": "Alibaba", "editors": [], "owners": [], "isMultipartSupported": true, @@ -32644,51 +35206,41 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/DeepInfra.webp" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256" } }, - "providerDisplayName": "DeepInfra", - "providerModelId": "Qwen/Qwen3-14B", - "providerGroup": "DeepInfra", - "quantization": "fp8", - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 40960, + "providerDisplayName": "Alibaba", + "providerModelId": "qwen2.5-vl-72b-instruct", + "providerGroup": "Alibaba", + "quantization": null, + "variant": "free", + "isFree": true, + "canAbort": false, + "maxPromptTokens": 129024, + "maxCompletionTokens": 2048, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", "seed", - "min_p" + "response_format", + "presence_penalty" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", + "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": false + "training": false }, - "training": false, - "retainsPrompts": false + "training": false }, "pricing": { - "prompt": "0.00000007", - "completion": "0.00000024", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -32700,37 +35252,38 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 102, - "newest": 26, - "throughputHighToLow": 63, - "latencyLowToHigh": 171, - "pricingLowToHigh": 11, - "pricingHighToLow": 212 + "topWeekly": 72, + "newest": 125, + "throughputHighToLow": 249, + "latencyLowToHigh": 257, + "pricingLowToHigh": 48, + "pricingHighToLow": 154 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": 40960, - "providerModelId": "Qwen/Qwen3-14B", + "context": 32000, + "maxCompletionTokens": null, + "providerModelId": "Qwen/Qwen2.5-VL-72B-Instruct", "pricing": { - "prompt": "0.00000007", - "completion": "0.00000024", + "prompt": "0.00000025", + "completion": "0.00000075", "image": "0", "request": "0", "webSearch": "0", @@ -32741,30 +35294,31 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", "seed", - "min_p" + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 0.07, - "outputCost": 0.24 + "inputCost": 0.25, + "outputCost": 0.75, + "throughput": 13.064, + "latency": 2040.5 }, { - "name": "NovitaAI", - "slug": "novitaAi", + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", "quantization": "fp8", "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "qwen/qwen3-14b-fp8", + "maxCompletionTokens": 128000, + "providerModelId": "parasail-qwen25-vl-72b-instruct", "pricing": { - "prompt": "0.00000007", - "completion": "0.000000275", + "prompt": "0.0000007", + "completion": "0.0000007", "image": "0", "request": "0", "webSearch": "0", @@ -32775,30 +35329,27 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "min_p", + "frequency_penalty", "repetition_penalty", - "logit_bias" + "top_k" ], - "inputCost": 0.07, - "outputCost": 0.28 + "inputCost": 0.7, + "outputCost": 0.7, + "throughput": 33.3855, + "latency": 1933 }, { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 40960, + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": null, + "context": 96000, "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen3-14B", + "providerModelId": "qwen/qwen2.5-vl-72b-instruct", "pricing": { - "prompt": "0.00000008", - "completion": "0.00000024", + "prompt": "0.0000008", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -32809,243 +35360,31 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", "seed", "top_k", - "logit_bias", - "logprobs", - "top_logprobs" - ], - "inputCost": 0.08, - "outputCost": 0.24 - } - ] - }, - { - "slug": "deepseek/deepseek-prover-v2", - "hfSlug": "deepseek-ai/DeepSeek-Prover-V2-671B", - "updatedAt": "2025-04-30T13:01:50.865169+00:00", - "createdAt": "2025-04-30T11:38:14.302503+00:00", - "hfUpdatedAt": null, - "name": "DeepSeek: DeepSeek Prover V2", - "shortName": "DeepSeek Prover V2", - "author": "deepseek", - "description": "DeepSeek Prover V2 is a 671B parameter model, speculated to be geared towards logic and mathematics. Likely an upgrade from [DeepSeek-Prover-V1.5](https://huggingface.co/deepseek-ai/DeepSeek-Prover-V1.5-RL) Not much is known about the model yet, as DeepSeek released it on Hugging Face without an announcement or description.", - "modelVersionGroupId": null, - "contextLength": 131072, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "DeepSeek", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "deepseek/deepseek-prover-v2", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "c70b9a14-2531-4ed0-ab12-b8935873c6c7", - "name": "GMICloud | deepseek/deepseek-prover-v2", - "contextLength": 131072, - "model": { - "slug": "deepseek/deepseek-prover-v2", - "hfSlug": "deepseek-ai/DeepSeek-Prover-V2-671B", - "updatedAt": "2025-04-30T13:01:50.865169+00:00", - "createdAt": "2025-04-30T11:38:14.302503+00:00", - "hfUpdatedAt": null, - "name": "DeepSeek: DeepSeek Prover V2", - "shortName": "DeepSeek Prover V2", - "author": "deepseek", - "description": "DeepSeek Prover V2 is a 671B parameter model, speculated to be geared towards logic and mathematics. Likely an upgrade from [DeepSeek-Prover-V1.5](https://huggingface.co/deepseek-ai/DeepSeek-Prover-V1.5-RL) Not much is known about the model yet, as DeepSeek released it on Hugging Face without an announcement or description.", - "modelVersionGroupId": null, - "contextLength": 163840, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "DeepSeek", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "deepseek/deepseek-prover-v2", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "deepseek/deepseek-prover-v2", - "modelVariantPermaslug": "deepseek/deepseek-prover-v2", - "providerName": "GMICloud", - "providerInfo": { - "name": "GMICloud", - "displayName": "GMICloud", - "slug": "gmicloud", - "baseUrl": "url", - "dataPolicy": { - "privacyPolicyUrl": "https://docs.gmicloud.ai/privacy", - "paidModels": { - "training": false - } - }, - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "GMICloud", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://gmicloud.ai/&size=256", - "invertRequired": true - } - }, - "providerDisplayName": "GMICloud", - "providerModelId": "deepseek-ai/DeepSeek-Prover-V2-671B", - "providerGroup": "GMICloud", - "quantization": "fp8", - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "seed" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "privacyPolicyUrl": "https://docs.gmicloud.ai/privacy", - "paidModels": { - "training": false - }, - "training": false - }, - "pricing": { - "prompt": "0.0000005", - "completion": "0.00000218", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 74, - "newest": 19, - "throughputHighToLow": 105, - "latencyLowToHigh": 139, - "pricingLowToHigh": 8, - "pricingHighToLow": 123 - }, - "providers": [ - { - "name": "GMICloud", - "slug": "gmiCloud", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "deepseek-ai/DeepSeek-Prover-V2-671B", - "pricing": { - "prompt": "0.0000005", - "completion": "0.00000218", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "seed" - ], - "inputCost": 0.5, - "outputCost": 2.18 - }, - { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "fp8", - "context": 163840, - "maxCompletionTokens": null, - "providerModelId": "deepseek-ai/DeepSeek-Prover-V2-671B", - "pricing": { - "prompt": "0.0000007", - "completion": "0.00000218", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", + "min_p", "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" + "logit_bias" ], - "inputCost": 0.7, - "outputCost": 2.18 + "inputCost": 0.8, + "outputCost": 0.8, + "throughput": 17.542, + "latency": 4798 }, { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": "fp8", - "context": 160000, + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": null, + "context": 32768, "maxCompletionTokens": null, - "providerModelId": "deepseek/deepseek-prover-v2-671b", + "providerModelId": "Qwen/Qwen2.5-VL-72B-Instruct", "pricing": { - "prompt": "0.0000007", - "completion": "0.0000025", + "prompt": "0.00000195", + "completion": "0.000008", "image": "0", "request": "0", "webSearch": "0", @@ -33059,14 +35398,16 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", "top_k", - "min_p", "repetition_penalty", - "logit_bias" + "logit_bias", + "min_p", + "response_format" ], - "inputCost": 0.7, - "outputCost": 2.5 + "inputCost": 1.95, + "outputCost": 8, + "throughput": 31.478, + "latency": 1337 } ] }, @@ -33250,27 +35591,28 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 104, + "topWeekly": 106, "newest": 101, - "throughputHighToLow": 159, - "latencyLowToHigh": 194, + "throughputHighToLow": 132, + "latencyLowToHigh": 204, "pricingLowToHigh": 42, "pricingHighToLow": 290 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [] }, { - "slug": "sao10k/l3.3-euryale-70b", - "hfSlug": "Sao10K/L3.3-70B-Euryale-v2.3", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-18T15:32:08.468786+00:00", + "slug": "deepseek/deepseek-v3-base", + "hfSlug": "deepseek-ai/Deepseek-v3-base", + "updatedAt": "2025-03-29T18:19:47.595899+00:00", + "createdAt": "2025-03-29T18:13:43.339174+00:00", "hfUpdatedAt": null, - "name": "Sao10K: Llama 3.3 Euryale 70B", - "shortName": "Llama 3.3 Euryale 70B", - "author": "sao10k", - "description": "Euryale L3.3 70B is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k). It is the successor of [Euryale L3 70B v2.2](/models/sao10k/l3-euryale-70b).", + "name": "DeepSeek: DeepSeek V3 Base (free)", + "shortName": "DeepSeek V3 Base (free)", + "author": "deepseek", + "description": "Note that this is a base model mostly meant for testing, you need to provide detailed prompts for the model to return useful responses. \n\nDeepSeek-V3 Base is a 671B parameter open Mixture-of-Experts (MoE) language model with 37B active parameters per forward pass and a context length of 128K tokens. Trained on 14.8T tokens using FP8 mixed precision, it achieves high training efficiency and stability, with strong performance across language, reasoning, math, and coding tasks. \n\nDeepSeek-V3 Base is the pre-trained model behind [DeepSeek V3](/deepseek/deepseek-chat-v3)", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 163840, "inputModalities": [ "text" ], @@ -33278,35 +35620,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "DeepSeek", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "sao10k/l3.3-euryale-70b-v2.3", + "permaslug": "deepseek/deepseek-v3-base", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "6e3850b8-2305-4bda-990a-7aa06427bc83", - "name": "DeepInfra | sao10k/l3.3-euryale-70b-v2.3", - "contextLength": 131072, + "id": "c20782a8-abb2-4a2d-9519-d6bfa4b08d7e", + "name": "Chutes | deepseek/deepseek-v3-base:free", + "contextLength": 163840, "model": { - "slug": "sao10k/l3.3-euryale-70b", - "hfSlug": "Sao10K/L3.3-70B-Euryale-v2.3", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-18T15:32:08.468786+00:00", + "slug": "deepseek/deepseek-v3-base", + "hfSlug": "deepseek-ai/Deepseek-v3-base", + "updatedAt": "2025-03-29T18:19:47.595899+00:00", + "createdAt": "2025-03-29T18:13:43.339174+00:00", "hfUpdatedAt": null, - "name": "Sao10K: Llama 3.3 Euryale 70B", - "shortName": "Llama 3.3 Euryale 70B", - "author": "sao10k", - "description": "Euryale L3.3 70B is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k). It is the successor of [Euryale L3 70B v2.2](/models/sao10k/l3-euryale-70b).", + "name": "DeepSeek: DeepSeek V3 Base", + "shortName": "DeepSeek V3 Base", + "author": "deepseek", + "description": "Note that this is a base model mostly meant for testing, you need to provide detailed prompts for the model to return useful responses. \n\nDeepSeek-V3 Base is a 671B parameter open Mixture-of-Experts (MoE) language model with 37B active parameters per forward pass and a context length of 128K tokens. Trained on 14.8T tokens using FP8 mixed precision, it achieves high training efficiency and stability, with strong performance across language, reasoning, math, and coding tasks. \n\nDeepSeek-V3 Base is the pre-trained model behind [DeepSeek V3](/deepseek/deepseek-chat-v3)", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -33314,43 +35653,37 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "DeepSeek", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "sao10k/l3.3-euryale-70b-v2.3", + "permaslug": "deepseek/deepseek-v3-base", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "sao10k/l3.3-euryale-70b", - "modelVariantPermaslug": "sao10k/l3.3-euryale-70b-v2.3", - "providerName": "DeepInfra", + "modelVariantSlug": "deepseek/deepseek-v3-base:free", + "modelVariantPermaslug": "deepseek/deepseek-v3-base:free", + "providerName": "Chutes", "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "DeepInfra", + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, @@ -33358,18 +35691,18 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/DeepInfra.webp" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "DeepInfra", - "providerModelId": "Sao10K/L3.3-70B-Euryale-v2.3", - "providerGroup": "DeepInfra", - "quantization": "fp8", - "variant": "standard", - "isFree": false, + "providerDisplayName": "Chutes", + "providerModelId": "deepseek-ai/DeepSeek-V3-Base", + "providerGroup": "Chutes", + "quantization": null, + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 16384, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -33379,28 +35712,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", "seed", - "min_p" + "top_k", + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true }, - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.0000007", - "completion": "0.0000008", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -33419,97 +35752,34 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 105, - "newest": 150, - "throughputHighToLow": 210, - "latencyLowToHigh": 11, - "pricingLowToHigh": 196, - "pricingHighToLow": 121 + "topWeekly": 107, + "newest": 67, + "throughputHighToLow": 220, + "latencyLowToHigh": 205, + "pricingLowToHigh": 28, + "pricingHighToLow": 276 }, - "providers": [ - { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 16384, - "providerModelId": "Sao10K/L3.3-70B-Euryale-v2.3", - "pricing": { - "prompt": "0.0000007", - "completion": "0.0000008", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "inputCost": 0.7, - "outputCost": 0.8 - }, - { - "name": "Infermatic", - "slug": "infermatic", - "quantization": "fp8", - "context": 16384, - "maxCompletionTokens": null, - "providerModelId": "Sao10K-L3.3-70B-Euryale-v2.3-FP8-Dynamic", - "pricing": { - "prompt": "0.0000015", - "completion": "0.0000015", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "logit_bias", - "top_k", - "min_p", - "seed" - ], - "inputCost": 1.5, - "outputCost": 1.5 - } - ] + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", + "providers": [] }, { - "slug": "deepseek/deepseek-v3-base", - "hfSlug": "deepseek-ai/Deepseek-v3-base", - "updatedAt": "2025-03-29T18:19:47.595899+00:00", - "createdAt": "2025-03-29T18:13:43.339174+00:00", + "slug": "qwen/qwen3-30b-a3b", + "hfSlug": "Qwen/Qwen3-30B-A3B", + "updatedAt": "2025-05-12T00:33:13.151167+00:00", + "createdAt": "2025-04-28T22:16:44.177326+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: DeepSeek V3 Base (free)", - "shortName": "DeepSeek V3 Base (free)", - "author": "deepseek", - "description": "Note that this is a base model mostly meant for testing, you need to provide detailed prompts for the model to return useful responses. \n\nDeepSeek-V3 Base is a 671B parameter open Mixture-of-Experts (MoE) language model with 37B active parameters per forward pass and a context length of 128K tokens. Trained on 14.8T tokens using FP8 mixed precision, it achieves high training efficiency and stability, with strong performance across language, reasoning, math, and coding tasks. \n\nDeepSeek-V3 Base is the pre-trained model behind [DeepSeek V3](/deepseek/deepseek-chat-v3)", + "name": "Qwen: Qwen3 30B A3B", + "shortName": "Qwen3 30B A3B", + "author": "qwen", + "description": "Qwen3, the latest generation in the Qwen large language model series, features both dense and mixture-of-experts (MoE) architectures to excel in reasoning, multilingual support, and advanced agent tasks. Its unique ability to switch seamlessly between a thinking mode for complex reasoning and a non-thinking mode for efficient dialogue ensures versatile, high-quality performance.\n\nSignificantly outperforming prior models like QwQ and Qwen2.5, Qwen3 delivers superior mathematics, coding, commonsense reasoning, creative writing, and interactive dialogue capabilities. The Qwen3-30B-A3B variant includes 30.5 billion parameters (3.3 billion activated), 48 layers, 128 experts (8 activated per task), and supports up to 131K token contexts with YaRN, setting a new standard among open-source models.", "modelVersionGroupId": null, - "contextLength": 163840, + "contextLength": 40960, "inputModalities": [ "text" ], @@ -33517,226 +35787,74 @@ export default { "text" ], "hasTextOutput": true, - "group": "DeepSeek", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "deepseek/deepseek-v3-base", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "c20782a8-abb2-4a2d-9519-d6bfa4b08d7e", - "name": "Chutes | deepseek/deepseek-v3-base:free", - "contextLength": 163840, - "model": { - "slug": "deepseek/deepseek-v3-base", - "hfSlug": "deepseek-ai/Deepseek-v3-base", - "updatedAt": "2025-03-29T18:19:47.595899+00:00", - "createdAt": "2025-03-29T18:13:43.339174+00:00", - "hfUpdatedAt": null, - "name": "DeepSeek: DeepSeek V3 Base", - "shortName": "DeepSeek V3 Base", - "author": "deepseek", - "description": "Note that this is a base model mostly meant for testing, you need to provide detailed prompts for the model to return useful responses. \n\nDeepSeek-V3 Base is a 671B parameter open Mixture-of-Experts (MoE) language model with 37B active parameters per forward pass and a context length of 128K tokens. Trained on 14.8T tokens using FP8 mixed precision, it achieves high training efficiency and stability, with strong performance across language, reasoning, math, and coding tasks. \n\nDeepSeek-V3 Base is the pre-trained model behind [DeepSeek V3](/deepseek/deepseek-chat-v3)", - "modelVersionGroupId": null, - "contextLength": 131072, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "DeepSeek", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "deepseek/deepseek-v3-base", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "deepseek/deepseek-v3-base:free", - "modelVariantPermaslug": "deepseek/deepseek-v3-base:free", - "providerName": "Chutes", - "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", - "paidModels": { - "training": true, - "retainsPrompts": true - } - }, - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "Chutes", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" - } - }, - "providerDisplayName": "Chutes", - "providerModelId": "deepseek-ai/DeepSeek-V3-Base", - "providerGroup": "Chutes", - "quantization": null, - "variant": "free", - "isFree": true, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", - "paidModels": { - "training": true, - "retainsPrompts": true - }, - "training": true, - "retainsPrompts": true - }, - "pricing": { - "prompt": "0", - "completion": "0", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 106, - "newest": 67, - "throughputHighToLow": 220, - "latencyLowToHigh": 193, - "pricingLowToHigh": 28, - "pricingHighToLow": 276 - }, - "providers": [] - }, - { - "slug": "meta-llama/llama-3.2-11b-vision-instruct", - "hfSlug": "meta-llama/Llama-3.2-11B-Vision-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-25T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Meta: Llama 3.2 11B Vision Instruct", - "shortName": "Llama 3.2 11B Vision Instruct", - "author": "meta-llama", - "description": "Llama 3.2 11B Vision is a multimodal model with 11 billion parameters, designed to handle tasks combining visual and textual data. It excels in tasks such as image captioning and visual question answering, bridging the gap between language generation and visual reasoning. Pre-trained on a massive dataset of image-text pairs, it performs well in complex, high-accuracy image analysis.\n\nIts ability to integrate visual understanding with language processing makes it an ideal solution for industries requiring comprehensive visual-linguistic AI applications, such as content creation, AI-driven customer service, and research.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", - "modelVersionGroupId": null, - "contextLength": 131072, - "inputModalities": [ - "text", - "image" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" + "<|im_start|>", + "<|im_end|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.2-11b-vision-instruct", - "reasoningConfig": null, - "features": {}, + "permaslug": "qwen/qwen3-30b-a3b-04-28", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "4a07b512-e030-412d-b1d6-39773a8b8dcf", - "name": "DeepInfra | meta-llama/llama-3.2-11b-vision-instruct", - "contextLength": 131072, + "id": "efdfff77-9574-4695-8e08-32d968a43376", + "name": "DeepInfra | qwen/qwen3-30b-a3b-04-28", + "contextLength": 40960, "model": { - "slug": "meta-llama/llama-3.2-11b-vision-instruct", - "hfSlug": "meta-llama/Llama-3.2-11B-Vision-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-25T00:00:00+00:00", + "slug": "qwen/qwen3-30b-a3b", + "hfSlug": "Qwen/Qwen3-30B-A3B", + "updatedAt": "2025-05-12T00:33:13.151167+00:00", + "createdAt": "2025-04-28T22:16:44.177326+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.2 11B Vision Instruct", - "shortName": "Llama 3.2 11B Vision Instruct", - "author": "meta-llama", - "description": "Llama 3.2 11B Vision is a multimodal model with 11 billion parameters, designed to handle tasks combining visual and textual data. It excels in tasks such as image captioning and visual question answering, bridging the gap between language generation and visual reasoning. Pre-trained on a massive dataset of image-text pairs, it performs well in complex, high-accuracy image analysis.\n\nIts ability to integrate visual understanding with language processing makes it an ideal solution for industries requiring comprehensive visual-linguistic AI applications, such as content creation, AI-driven customer service, and research.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "name": "Qwen: Qwen3 30B A3B", + "shortName": "Qwen3 30B A3B", + "author": "qwen", + "description": "Qwen3, the latest generation in the Qwen large language model series, features both dense and mixture-of-experts (MoE) architectures to excel in reasoning, multilingual support, and advanced agent tasks. Its unique ability to switch seamlessly between a thinking mode for complex reasoning and a non-thinking mode for efficient dialogue ensures versatile, high-quality performance.\n\nSignificantly outperforming prior models like QwQ and Qwen2.5, Qwen3 delivers superior mathematics, coding, commonsense reasoning, creative writing, and interactive dialogue capabilities. The Qwen3-30B-A3B variant includes 30.5 billion parameters (3.3 billion activated), 48 layers, 128 experts (8 activated per task), and supports up to 131K token contexts with YaRN, setting a new standard among open-source models.", "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" + "<|im_start|>", + "<|im_end|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.2-11b-vision-instruct", - "reasoningConfig": null, - "features": {} + "permaslug": "qwen/qwen3-30b-a3b-04-28", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "meta-llama/llama-3.2-11b-vision-instruct", - "modelVariantPermaslug": "meta-llama/llama-3.2-11b-vision-instruct", + "modelVariantSlug": "qwen/qwen3-30b-a3b", + "modelVariantPermaslug": "qwen/qwen3-30b-a3b-04-28", "providerName": "DeepInfra", "providerInfo": { "name": "DeepInfra", @@ -33769,20 +35887,22 @@ export default { } }, "providerDisplayName": "DeepInfra", - "providerModelId": "meta-llama/Llama-3.2-11B-Vision-Instruct", + "providerModelId": "Qwen/Qwen3-30B-A3B", "providerGroup": "DeepInfra", - "quantization": "bf16", + "quantization": "fp8", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 16384, + "maxCompletionTokens": 40960, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -33806,9 +35926,9 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.000000049", - "completion": "0.000000049", - "image": "0.00007948", + "prompt": "0.0000001", + "completion": "0.0000003", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -33819,37 +35939,40 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 107, - "newest": 199, - "throughputHighToLow": 303, - "latencyLowToHigh": 272, - "pricingLowToHigh": 62, - "pricingHighToLow": 224 + "topWeekly": 94, + "newest": 22, + "throughputHighToLow": 20, + "latencyLowToHigh": 127, + "pricingLowToHigh": 9, + "pricingHighToLow": 193 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/Llama-3.2-11B-Vision-Instruct", + "quantization": "fp8", + "context": 40960, + "maxCompletionTokens": 40960, + "providerModelId": "Qwen/Qwen3-30B-A3B", "pricing": { - "prompt": "0.000000049", - "completion": "0.000000049", - "image": "0.00007948", + "prompt": "0.0000001", + "completion": "0.0000003", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -33859,6 +35982,8 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -33868,48 +35993,22 @@ export default { "seed", "min_p" ], - "inputCost": 0.05, - "outputCost": 0.05 + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 45.735, + "latency": 762 }, { - "name": "Cloudflare", - "slug": "cloudflare", - "quantization": null, - "context": 131072, + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 40960, "maxCompletionTokens": null, - "providerModelId": "@cf/meta/llama-3.2-11b-vision-instruct", - "pricing": { - "prompt": "0.000000049", - "completion": "0.00000068", - "image": "0.001281", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "top_k", - "seed", - "repetition_penalty", - "frequency_penalty", - "presence_penalty" - ], - "inputCost": 0.05, - "outputCost": 0.68 - }, - { - "name": "Lambda", - "slug": "lambda", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "llama3.2-11b-vision-instruct", + "providerModelId": "Qwen/Qwen3-30B-A3B", "pricing": { - "prompt": "0.00000005", - "completion": "0.00000005", + "prompt": "0.0000001", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -33920,28 +36019,33 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", "seed", + "top_k", "logit_bias", "logprobs", - "top_logprobs", - "response_format" + "top_logprobs" ], - "inputCost": 0.05, - "outputCost": 0.05 + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 97.242, + "latency": 349 }, { - "name": "inference.net", - "slug": "inferenceNet", - "quantization": "fp16", - "context": 16384, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/llama-3.2-11b-instruct/fp-16", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": "fp8", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "qwen/qwen3-30b-a3b-fp8", "pricing": { - "prompt": "0.000000055", - "completion": "0.000000055", + "prompt": "0.0000001", + "completion": "0.00000045", "image": "0", "request": "0", "webSearch": "0", @@ -33952,28 +36056,33 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", "seed", "top_k", "min_p", - "logit_bias", - "top_logprobs" + "repetition_penalty", + "logit_bias" ], - "inputCost": 0.06, - "outputCost": 0.06 + "inputCost": 0.1, + "outputCost": 0.45, + "throughput": 98.0455, + "latency": 900 }, { - "name": "NovitaAI", - "slug": "novitaAi", + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", "quantization": "fp8", - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/llama-3.2-11b-vision-instruct", + "context": 40960, + "maxCompletionTokens": 40960, + "providerModelId": "parasail-qwen3-30b-a3b", "pricing": { - "prompt": "0.00000006", - "completion": "0.00000006", + "prompt": "0.0000001", + "completion": "0.0000005", "image": "0", "request": "0", "webSearch": "0", @@ -33984,138 +36093,147 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", + "reasoning", + "include_reasoning", "presence_penalty", - "seed", - "top_k", - "min_p", + "frequency_penalty", "repetition_penalty", - "logit_bias" + "top_k" ], - "inputCost": 0.06, - "outputCost": 0.06 + "inputCost": 0.1, + "outputCost": 0.5, + "throughput": 119.9285, + "latency": 779 }, { - "name": "Together", - "slug": "together", - "quantization": "fp8", - "context": 131072, + "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", + "slug": "fireworks", + "quantization": null, + "context": 40000, "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo", + "providerModelId": "accounts/fireworks/models/qwen3-30b-a3b", "pricing": { - "prompt": "0.00000018", - "completion": "0.00000018", - "image": "0.001156", + "prompt": "0.00000015", + "completion": "0.0000006", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", "top_k", "repetition_penalty", + "response_format", + "structured_outputs", "logit_bias", - "min_p", - "response_format" + "logprobs", + "top_logprobs" ], - "inputCost": 0.18, - "outputCost": 0.18 + "inputCost": 0.15, + "outputCost": 0.6, + "throughput": 117.839, + "latency": 725 } ] }, { - "slug": "qwen/qwen2.5-vl-72b-instruct", - "hfSlug": "Qwen/Qwen2.5-VL-72B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-01T11:45:11.997326+00:00", + "slug": "mistral/ministral-8b", + "hfSlug": "mistralai/Ministral-8B-Instruct-2410", + "updatedAt": "2025-04-05T20:00:23.966407+00:00", + "createdAt": "2025-03-31T14:07:01.87388+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen2.5 VL 72B Instruct (free)", - "shortName": "Qwen2.5 VL 72B Instruct (free)", - "author": "qwen", - "description": "Qwen2.5-VL is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.", + "name": "Mistral: Ministral 8B", + "shortName": "Ministral 8B", + "author": "mistral", + "description": "Ministral 8B is a state-of-the-art language model optimized for on-device and edge computing. Designed for efficiency in knowledge-intensive tasks, commonsense reasoning, and function-calling, it features a specialized interleaved sliding-window attention mechanism, enabling faster and more memory-efficient inference. Ministral 8B excels in local, low-latency applications such as offline translation, smart assistants, autonomous robotics, and local analytics.\n\nThe model supports up to 128k context length and can function as a performant intermediary in multi-step agentic workflows, efficiently handling tasks like input parsing, API calls, and task routing. It consistently outperforms comparable models like Mistral 7B across benchmarks, making it particularly suitable for compute-efficient, privacy-focused scenarios.", "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", + "group": "Mistral", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen2.5-vl-72b-instruct", + "permaslug": "mistral/ministral-8b-2410", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "75b071e0-0415-4543-a28f-e8d8e28d4421", - "name": "Alibaba | qwen/qwen2.5-vl-72b-instruct:free", + "id": "070490ef-6982-4a3e-86c5-92f092506f22", + "name": "Mistral | mistral/ministral-8b-2410", "contextLength": 131072, "model": { - "slug": "qwen/qwen2.5-vl-72b-instruct", - "hfSlug": "Qwen/Qwen2.5-VL-72B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-01T11:45:11.997326+00:00", + "slug": "mistral/ministral-8b", + "hfSlug": "mistralai/Ministral-8B-Instruct-2410", + "updatedAt": "2025-04-05T20:00:23.966407+00:00", + "createdAt": "2025-03-31T14:07:01.87388+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen2.5 VL 72B Instruct", - "shortName": "Qwen2.5 VL 72B Instruct", - "author": "qwen", - "description": "Qwen2.5-VL is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.", + "name": "Mistral: Ministral 8B", + "shortName": "Ministral 8B", + "author": "mistral", + "description": "Ministral 8B is a state-of-the-art language model optimized for on-device and edge computing. Designed for efficiency in knowledge-intensive tasks, commonsense reasoning, and function-calling, it features a specialized interleaved sliding-window attention mechanism, enabling faster and more memory-efficient inference. Ministral 8B excels in local, low-latency applications such as offline translation, smart assistants, autonomous robotics, and local analytics.\n\nThe model supports up to 128k context length and can function as a performant intermediary in multi-step agentic workflows, efficiently handling tasks like input parsing, API calls, and task routing. It consistently outperforms comparable models like Mistral 7B across benchmarks, making it particularly suitable for compute-efficient, privacy-focused scenarios.", "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", + "group": "Mistral", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen2.5-vl-72b-instruct", + "permaslug": "mistral/ministral-8b-2410", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "qwen/qwen2.5-vl-72b-instruct:free", - "modelVariantPermaslug": "qwen/qwen2.5-vl-72b-instruct:free", - "providerName": "Alibaba", + "modelVariantSlug": "mistral/ministral-8b", + "modelVariantPermaslug": "mistral/ministral-8b-2410", + "providerName": "Mistral", "providerInfo": { - "name": "Alibaba", - "displayName": "Alibaba", - "slug": "alibaba", + "name": "Mistral", + "displayName": "Mistral", + "slug": "mistral", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", - "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": true, + "retentionDays": 30 } }, - "headquarters": "CN", + "headquarters": "FR", "hasChatCompletions": true, - "hasCompletions": true, + "hasCompletions": false, "isAbortable": false, "moderationRequired": false, - "group": "Alibaba", + "group": "Mistral", "editors": [], "owners": [], "isMultipartSupported": true, @@ -34123,41 +36241,50 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256" + "url": "/images/icons/Mistral.png" } }, - "providerDisplayName": "Alibaba", - "providerModelId": "qwen2.5-vl-72b-instruct", - "providerGroup": "Alibaba", + "providerDisplayName": "Mistral", + "providerModelId": "ministral-8b-2410", + "providerGroup": "Mistral", "quantization": null, - "variant": "free", - "isFree": true, + "variant": "standard", + "isFree": false, "canAbort": false, - "maxPromptTokens": 129024, - "maxCompletionTokens": 2048, - "maxPromptImages": null, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "seed", + "stop", + "frequency_penalty", + "presence_penalty", "response_format", - "presence_penalty" + "structured_outputs", + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", - "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "training": false + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.0000001", + "completion": "0.0000001", "image": "0", "request": "0", "webSearch": "0", @@ -34168,7 +36295,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -34176,121 +36303,32 @@ export default { "hasCompletions": false, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 62, - "newest": 125, - "throughputHighToLow": 297, - "latencyLowToHigh": 233, - "pricingLowToHigh": 48, - "pricingHighToLow": 154 + "topWeekly": 109, + "newest": 66, + "throughputHighToLow": 60, + "latencyLowToHigh": 19, + "pricingLowToHigh": 114, + "pricingHighToLow": 202 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 32000, - "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen2.5-VL-72B-Instruct", - "pricing": { - "prompt": "0.00000025", - "completion": "0.00000075", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" - ], - "inputCost": 0.25, - "outputCost": 0.75 - }, - { - "name": "Parasail", - "slug": "parasail", - "quantization": "fp8", - "context": 128000, - "maxCompletionTokens": 128000, - "providerModelId": "parasail-qwen25-vl-72b-instruct", - "pricing": { - "prompt": "0.0000007", - "completion": "0.0000007", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "presence_penalty", - "frequency_penalty", - "repetition_penalty", - "top_k" - ], - "inputCost": 0.7, - "outputCost": 0.7 - }, - { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": null, - "context": 96000, - "maxCompletionTokens": null, - "providerModelId": "qwen/qwen2.5-vl-72b-instruct", - "pricing": { - "prompt": "0.0000008", - "completion": "0.0000008", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" - ], - "inputCost": 0.8, - "outputCost": 0.8 - }, - { - "name": "Together", - "slug": "together", + "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", + "slug": "mistral", "quantization": null, - "context": 32768, + "context": 131072, "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen2.5-VL-72B-Instruct", + "providerModelId": "ministral-8b-2410", "pricing": { - "prompt": "0.00000195", - "completion": "0.000008", + "prompt": "0.0000001", + "completion": "0.0000001", "image": "0", "request": "0", "webSearch": "0", @@ -34298,20 +36336,22 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "response_format", + "structured_outputs", + "seed" ], - "inputCost": 1.95, - "outputCost": 8 + "inputCost": 0.1, + "outputCost": 0.1, + "throughput": 115.326, + "latency": 276 } ] }, @@ -34471,16 +36511,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 109, - "newest": 225, - "throughputHighToLow": 268, - "latencyLowToHigh": 305, + "topWeekly": 110, + "newest": 224, + "throughputHighToLow": 270, + "latencyLowToHigh": 292, "pricingLowToHigh": 66, "pricingHighToLow": 78 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", "slug": "hyperbolic", "quantization": "fp8", "context": 32768, @@ -34515,6 +36557,7 @@ export default { }, { "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", "slug": "hyperbolic", "quantization": "bf16", "context": 32768, @@ -34545,7 +36588,9 @@ export default { "repetition_penalty" ], "inputCost": 4, - "outputCost": 4 + "outputCost": 4, + "throughput": 12.006, + "latency": 1129.5 } ] }, @@ -34739,16 +36784,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 110, + "topWeekly": 111, "newest": 223, - "throughputHighToLow": 119, - "latencyLowToHigh": 89, + "throughputHighToLow": 152, + "latencyLowToHigh": 106, "pricingLowToHigh": 264, "pricingHighToLow": 58 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", "slug": "openAi", "quantization": "unknown", "context": 128000, @@ -34782,10 +36829,13 @@ export default { "structured_outputs" ], "inputCost": 2.5, - "outputCost": 10 + "outputCost": 10, + "throughput": 54.8765, + "latency": 642 }, { "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", "slug": "azure", "quantization": "unknown", "context": 128000, @@ -34818,96 +36868,98 @@ export default { "structured_outputs" ], "inputCost": 2.5, - "outputCost": 10 + "outputCost": 10, + "throughput": 130.12, + "latency": 1237.5 } ] }, { - "slug": "mistral/ministral-8b", - "hfSlug": "mistralai/Ministral-8B-Instruct-2410", - "updatedAt": "2025-04-05T20:00:23.966407+00:00", - "createdAt": "2025-03-31T14:07:01.87388+00:00", + "slug": "qwen/qwen-vl-plus", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-05T04:54:15.216448+00:00", "hfUpdatedAt": null, - "name": "Mistral: Ministral 8B", - "shortName": "Ministral 8B", - "author": "mistral", - "description": "Ministral 8B is a state-of-the-art language model optimized for on-device and edge computing. Designed for efficiency in knowledge-intensive tasks, commonsense reasoning, and function-calling, it features a specialized interleaved sliding-window attention mechanism, enabling faster and more memory-efficient inference. Ministral 8B excels in local, low-latency applications such as offline translation, smart assistants, autonomous robotics, and local analytics.\n\nThe model supports up to 128k context length and can function as a performant intermediary in multi-step agentic workflows, efficiently handling tasks like input parsing, API calls, and task routing. It consistently outperforms comparable models like Mistral 7B across benchmarks, making it particularly suitable for compute-efficient, privacy-focused scenarios.", + "name": "Qwen: Qwen VL Plus", + "shortName": "Qwen VL Plus", + "author": "qwen", + "description": "Qwen's Enhanced Large Visual Language Model. Significantly upgraded for detailed recognition capabilities and text recognition abilities, supporting ultra-high pixel resolutions up to millions of pixels and extreme aspect ratios for image input. It delivers significant performance across a broad range of visual tasks.\n", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 7500, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "Qwen", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistral/ministral-8b-2410", + "permaslug": "qwen/qwen-vl-plus", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "070490ef-6982-4a3e-86c5-92f092506f22", - "name": "Mistral | mistral/ministral-8b-2410", - "contextLength": 131072, + "id": "df9c15ac-870b-40e5-aa43-e4b3b44951f7", + "name": "Alibaba | qwen/qwen-vl-plus", + "contextLength": 7500, "model": { - "slug": "mistral/ministral-8b", - "hfSlug": "mistralai/Ministral-8B-Instruct-2410", - "updatedAt": "2025-04-05T20:00:23.966407+00:00", - "createdAt": "2025-03-31T14:07:01.87388+00:00", + "slug": "qwen/qwen-vl-plus", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-05T04:54:15.216448+00:00", "hfUpdatedAt": null, - "name": "Mistral: Ministral 8B", - "shortName": "Ministral 8B", - "author": "mistral", - "description": "Ministral 8B is a state-of-the-art language model optimized for on-device and edge computing. Designed for efficiency in knowledge-intensive tasks, commonsense reasoning, and function-calling, it features a specialized interleaved sliding-window attention mechanism, enabling faster and more memory-efficient inference. Ministral 8B excels in local, low-latency applications such as offline translation, smart assistants, autonomous robotics, and local analytics.\n\nThe model supports up to 128k context length and can function as a performant intermediary in multi-step agentic workflows, efficiently handling tasks like input parsing, API calls, and task routing. It consistently outperforms comparable models like Mistral 7B across benchmarks, making it particularly suitable for compute-efficient, privacy-focused scenarios.", + "name": "Qwen: Qwen VL Plus", + "shortName": "Qwen VL Plus", + "author": "qwen", + "description": "Qwen's Enhanced Large Visual Language Model. Significantly upgraded for detailed recognition capabilities and text recognition abilities, supporting ultra-high pixel resolutions up to millions of pixels and extreme aspect ratios for image input. It delivers significant performance across a broad range of visual tasks.\n", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 7500, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "Qwen", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistral/ministral-8b-2410", + "permaslug": "qwen/qwen-vl-plus", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "mistral/ministral-8b", - "modelVariantPermaslug": "mistral/ministral-8b-2410", - "providerName": "Mistral", + "modelVariantSlug": "qwen/qwen-vl-plus", + "modelVariantPermaslug": "qwen/qwen-vl-plus", + "providerName": "Alibaba", "providerInfo": { - "name": "Mistral", - "displayName": "Mistral", - "slug": "mistral", + "name": "Alibaba", + "displayName": "Alibaba", + "slug": "alibaba", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", + "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false } }, - "headquarters": "FR", + "headquarters": "CN", "hasChatCompletions": true, - "hasCompletions": false, + "hasCompletions": true, "isAbortable": false, "moderationRequired": false, - "group": "Mistral", + "group": "Alibaba", "editors": [], "owners": [], "isMultipartSupported": true, @@ -34915,51 +36967,42 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Mistral.png" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256" } }, - "providerDisplayName": "Mistral", - "providerModelId": "ministral-8b-2410", - "providerGroup": "Mistral", + "providerDisplayName": "Alibaba", + "providerModelId": "qwen-vl-plus", + "providerGroup": "Alibaba", "quantization": null, "variant": "standard", "isFree": false, "canAbort": false, - "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxPromptTokens": 6000, + "maxCompletionTokens": 1500, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", + "seed", "response_format", - "structured_outputs", - "seed" + "presence_penalty" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", + "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", - "image": "0", + "prompt": "0.00000021", + "completion": "0.00000063", + "image": "0.0002688", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -34969,7 +37012,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -34977,216 +37020,196 @@ export default { "hasCompletions": false, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 111, - "newest": 66, - "throughputHighToLow": 60, - "latencyLowToHigh": 24, - "pricingLowToHigh": 114, - "pricingHighToLow": 202 + "topWeekly": 112, + "newest": 119, + "throughputHighToLow": 95, + "latencyLowToHigh": 155, + "pricingLowToHigh": 159, + "pricingHighToLow": 159 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { - "name": "Mistral", - "slug": "mistral", + "name": "Alibaba", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256", + "slug": "alibaba", "quantization": null, - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "ministral-8b-2410", + "context": 7500, + "maxCompletionTokens": 1500, + "providerModelId": "qwen-vl-plus", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", - "image": "0", + "prompt": "0.00000021", + "completion": "0.00000063", + "image": "0.0002688", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", + "seed", "response_format", - "structured_outputs", - "seed" + "presence_penalty" ], - "inputCost": 0.1, - "outputCost": 0.1 + "inputCost": 0.21, + "outputCost": 0.63, + "throughput": 81.893, + "latency": 966 } ] }, { - "slug": "qwen/qwen3-30b-a3b", - "hfSlug": "Qwen/Qwen3-30B-A3B", - "updatedAt": "2025-05-12T00:33:13.151167+00:00", - "createdAt": "2025-04-28T22:16:44.177326+00:00", + "slug": "openai/o3", + "hfSlug": "", + "updatedAt": "2025-05-02T21:34:35.534922+00:00", + "createdAt": "2025-04-16T17:10:57.049467+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 30B A3B", - "shortName": "Qwen3 30B A3B", - "author": "qwen", - "description": "Qwen3, the latest generation in the Qwen large language model series, features both dense and mixture-of-experts (MoE) architectures to excel in reasoning, multilingual support, and advanced agent tasks. Its unique ability to switch seamlessly between a thinking mode for complex reasoning and a non-thinking mode for efficient dialogue ensures versatile, high-quality performance.\n\nSignificantly outperforming prior models like QwQ and Qwen2.5, Qwen3 delivers superior mathematics, coding, commonsense reasoning, creative writing, and interactive dialogue capabilities. The Qwen3-30B-A3B variant includes 30.5 billion parameters (3.3 billion activated), 48 layers, 128 experts (8 activated per task), and supports up to 131K token contexts with YaRN, setting a new standard among open-source models.", + "name": "OpenAI: o3", + "shortName": "o3", + "author": "openai", + "description": "o3 is a well-rounded and powerful model across domains. It sets a new standard for math, science, coding, and visual reasoning tasks. It also excels at technical writing and instruction-following. Use it to think through multi-step problems that involve analysis across text, code, and images. Note that BYOK is required for this model. Set up here: https://openrouter.ai/settings/integrations", "modelVersionGroupId": null, - "contextLength": 40960, + "contextLength": 200000, "inputModalities": [ - "text" + "image", + "text", + "file" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>" - ], + "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "qwen/qwen3-30b-a3b-04-28", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "warningMessage": "OpenAI requires bringing your own API key to use o3 over the API. Set up here: https://openrouter.ai/settings/integrations", + "permaslug": "openai/o3-2025-04-16", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "efdfff77-9574-4695-8e08-32d968a43376", - "name": "DeepInfra | qwen/qwen3-30b-a3b-04-28", - "contextLength": 40960, + "id": "42e72619-d01c-411c-a201-f991644768b7", + "name": "OpenAI | openai/o3-2025-04-16", + "contextLength": 200000, "model": { - "slug": "qwen/qwen3-30b-a3b", - "hfSlug": "Qwen/Qwen3-30B-A3B", - "updatedAt": "2025-05-12T00:33:13.151167+00:00", - "createdAt": "2025-04-28T22:16:44.177326+00:00", + "slug": "openai/o3", + "hfSlug": "", + "updatedAt": "2025-05-02T21:34:35.534922+00:00", + "createdAt": "2025-04-16T17:10:57.049467+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 30B A3B", - "shortName": "Qwen3 30B A3B", - "author": "qwen", - "description": "Qwen3, the latest generation in the Qwen large language model series, features both dense and mixture-of-experts (MoE) architectures to excel in reasoning, multilingual support, and advanced agent tasks. Its unique ability to switch seamlessly between a thinking mode for complex reasoning and a non-thinking mode for efficient dialogue ensures versatile, high-quality performance.\n\nSignificantly outperforming prior models like QwQ and Qwen2.5, Qwen3 delivers superior mathematics, coding, commonsense reasoning, creative writing, and interactive dialogue capabilities. The Qwen3-30B-A3B variant includes 30.5 billion parameters (3.3 billion activated), 48 layers, 128 experts (8 activated per task), and supports up to 131K token contexts with YaRN, setting a new standard among open-source models.", + "name": "OpenAI: o3", + "shortName": "o3", + "author": "openai", + "description": "o3 is a well-rounded and powerful model across domains. It sets a new standard for math, science, coding, and visual reasoning tasks. It also excels at technical writing and instruction-following. Use it to think through multi-step problems that involve analysis across text, code, and images. Note that BYOK is required for this model. Set up here: https://openrouter.ai/settings/integrations", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 200000, "inputModalities": [ - "text" + "image", + "text", + "file" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>" - ], + "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "qwen/qwen3-30b-a3b-04-28", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "warningMessage": "OpenAI requires bringing your own API key to use o3 over the API. Set up here: https://openrouter.ai/settings/integrations", + "permaslug": "openai/o3-2025-04-16", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "qwen/qwen3-30b-a3b", - "modelVariantPermaslug": "qwen/qwen3-30b-a3b-04-28", - "providerName": "DeepInfra", + "modelVariantSlug": "openai/o3", + "modelVariantPermaslug": "openai/o3-2025-04-16", + "providerName": "OpenAI", "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false - } + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": false, - "group": "DeepInfra", + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/DeepInfra.webp" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "DeepInfra", - "providerModelId": "Qwen/Qwen3-30B-A3B", - "providerGroup": "DeepInfra", - "quantization": "fp8", + "providerDisplayName": "OpenAI", + "providerModelId": "o3-2025-04-16", + "providerGroup": "OpenAI", + "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 40960, + "maxCompletionTokens": 100000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", + "seed", "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", "response_format", - "top_k", - "seed", - "min_p" + "structured_outputs" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, + "requiresUserIds": true, "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", - "image": "0", + "prompt": "0.00001", + "completion": "0.00004", + "image": "0.00765", "request": "0", + "inputCacheRead": "0.0000025", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -35195,7 +37218,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, @@ -35203,164 +37226,38 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, + "supportedParameters": { + "responseFormat": true, + "structuredOutputs": true + }, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 89, - "newest": 22, - "throughputHighToLow": 23, - "latencyLowToHigh": 127, - "pricingLowToHigh": 9, - "pricingHighToLow": 193 + "topWeekly": 113, + "newest": 44, + "throughputHighToLow": 205, + "latencyLowToHigh": 304, + "pricingLowToHigh": 305, + "pricingHighToLow": 13 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": 40960, - "providerModelId": "Qwen/Qwen3-30B-A3B", - "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "inputCost": 0.1, - "outputCost": 0.3 - }, - { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen3-30B-A3B", - "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" - ], - "inputCost": 0.1, - "outputCost": 0.3 - }, - { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": "fp8", - "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "qwen/qwen3-30b-a3b-fp8", - "pricing": { - "prompt": "0.0000001", - "completion": "0.00000045", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" - ], - "inputCost": 0.1, - "outputCost": 0.45 - }, - { - "name": "Parasail", - "slug": "parasail", - "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": 40960, - "providerModelId": "parasail-qwen3-30b-a3b", - "pricing": { - "prompt": "0.0000001", - "completion": "0.0000005", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "presence_penalty", - "frequency_penalty", - "repetition_penalty", - "top_k" - ], - "inputCost": 0.1, - "outputCost": 0.5 - }, - { - "name": "Fireworks", - "slug": "fireworks", + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", "quantization": null, - "context": 40000, - "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/qwen3-30b-a3b", + "context": 200000, + "maxCompletionTokens": 100000, + "providerModelId": "o3-2025-04-16", "pricing": { - "prompt": "0.00000015", - "completion": "0.0000006", - "image": "0", + "prompt": "0.00001", + "completion": "0.00004", + "image": "0.00765", "request": "0", + "inputCacheRead": "0.0000025", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -35368,136 +37265,123 @@ export default { "supportedParameters": [ "tools", "tool_choice", + "seed", "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", "response_format", - "structured_outputs", - "logit_bias", - "logprobs", - "top_logprobs" + "structured_outputs" ], - "inputCost": 0.15, - "outputCost": 0.6 + "inputCost": 10, + "outputCost": 40, + "throughput": 39.192, + "latency": 6111 } ] }, { - "slug": "microsoft/phi-3.5-mini-128k-instruct", - "hfSlug": "microsoft/Phi-3.5-mini-instruct", + "slug": "qwen/qwen2.5-vl-32b-instruct", + "hfSlug": "Qwen/Qwen2.5-VL-32B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-21T00:00:00+00:00", + "createdAt": "2025-03-24T18:10:38.542849+00:00", "hfUpdatedAt": null, - "name": "Microsoft: Phi-3.5 Mini 128K Instruct", - "shortName": "Phi-3.5 Mini 128K Instruct", - "author": "microsoft", - "description": "Phi-3.5 models are lightweight, state-of-the-art open models. These models were trained with Phi-3 datasets that include both synthetic data and the filtered, publicly available websites data, with a focus on high quality and reasoning-dense properties. Phi-3.5 Mini uses 3.8B parameters, and is a dense decoder-only transformer model using the same tokenizer as [Phi-3 Mini](/models/microsoft/phi-3-mini-128k-instruct).\n\nThe models underwent a rigorous enhancement process, incorporating both supervised fine-tuning, proximal policy optimization, and direct preference optimization to ensure precise instruction adherence and robust safety measures. When assessed against benchmarks that test common sense, language understanding, math, code, long context and logical reasoning, Phi-3.5 models showcased robust and state-of-the-art performance among models with less than 13 billion parameters.", + "name": "Qwen: Qwen2.5 VL 32B Instruct", + "shortName": "Qwen2.5 VL 32B Instruct", + "author": "qwen", + "description": "Qwen2.5-VL-32B is a multimodal vision-language model fine-tuned through reinforcement learning for enhanced mathematical reasoning, structured outputs, and visual problem-solving capabilities. It excels at visual analysis tasks, including object recognition, textual interpretation within images, and precise event localization in extended videos. Qwen2.5-VL-32B demonstrates state-of-the-art performance across multimodal benchmarks such as MMMU, MathVista, and VideoMME, while maintaining strong reasoning and clarity in text-based tasks like MMLU, mathematical problem-solving, and code generation.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 128000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "phi3", + "group": "Qwen", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "microsoft/phi-3.5-mini-128k-instruct", + "permaslug": "qwen/qwen2.5-vl-32b-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "c284390d-1916-4a33-9bd7-b49c8165c2ca", - "name": "Nebius | microsoft/phi-3.5-mini-128k-instruct", - "contextLength": 131072, + "id": "aac76f56-eedb-405f-a07c-5bc66b877c1e", + "name": "Fireworks | qwen/qwen2.5-vl-32b-instruct", + "contextLength": 128000, "model": { - "slug": "microsoft/phi-3.5-mini-128k-instruct", - "hfSlug": "microsoft/Phi-3.5-mini-instruct", + "slug": "qwen/qwen2.5-vl-32b-instruct", + "hfSlug": "Qwen/Qwen2.5-VL-32B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-21T00:00:00+00:00", + "createdAt": "2025-03-24T18:10:38.542849+00:00", "hfUpdatedAt": null, - "name": "Microsoft: Phi-3.5 Mini 128K Instruct", - "shortName": "Phi-3.5 Mini 128K Instruct", - "author": "microsoft", - "description": "Phi-3.5 models are lightweight, state-of-the-art open models. These models were trained with Phi-3 datasets that include both synthetic data and the filtered, publicly available websites data, with a focus on high quality and reasoning-dense properties. Phi-3.5 Mini uses 3.8B parameters, and is a dense decoder-only transformer model using the same tokenizer as [Phi-3 Mini](/models/microsoft/phi-3-mini-128k-instruct).\n\nThe models underwent a rigorous enhancement process, incorporating both supervised fine-tuning, proximal policy optimization, and direct preference optimization to ensure precise instruction adherence and robust safety measures. When assessed against benchmarks that test common sense, language understanding, math, code, long context and logical reasoning, Phi-3.5 models showcased robust and state-of-the-art performance among models with less than 13 billion parameters.", + "name": "Qwen: Qwen2.5 VL 32B Instruct", + "shortName": "Qwen2.5 VL 32B Instruct", + "author": "qwen", + "description": "Qwen2.5-VL-32B is a multimodal vision-language model fine-tuned through reinforcement learning for enhanced mathematical reasoning, structured outputs, and visual problem-solving capabilities. It excels at visual analysis tasks, including object recognition, textual interpretation within images, and precise event localization in extended videos. Qwen2.5-VL-32B demonstrates state-of-the-art performance across multimodal benchmarks such as MMMU, MathVista, and VideoMME, while maintaining strong reasoning and clarity in text-based tasks like MMLU, mathematical problem-solving, and code generation.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 32768, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "phi3", + "group": "Qwen", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "microsoft/phi-3.5-mini-128k-instruct", + "permaslug": "qwen/qwen2.5-vl-32b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "microsoft/phi-3.5-mini-128k-instruct", - "modelVariantPermaslug": "microsoft/phi-3.5-mini-128k-instruct", - "providerName": "Nebius", + "modelVariantSlug": "qwen/qwen2.5-vl-32b-instruct", + "modelVariantPermaslug": "qwen/qwen2.5-vl-32b-instruct", + "providerName": "Fireworks", "providerInfo": { - "name": "Nebius", - "displayName": "Nebius AI Studio", - "slug": "nebius", + "name": "Fireworks", + "displayName": "Fireworks", + "slug": "fireworks", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "termsOfServiceUrl": "https://fireworks.ai/terms-of-service", + "privacyPolicyUrl": "https://fireworks.ai/privacy-policy", + "dataPolicyUrl": "https://docs.fireworks.ai/guides/security_compliance/data_handling", "paidModels": { "training": false, "retainsPrompts": false } }, - "headquarters": "DE", + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, + "isAbortable": true, "moderationRequired": false, - "group": "Nebius", + "group": "Fireworks", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.fireworks.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", - "invertRequired": true + "url": "/images/icons/Fireworks.png" } }, - "providerDisplayName": "Nebius AI Studio", - "providerModelId": "microsoft/Phi-3.5-mini-instruct", - "providerGroup": "Nebius", + "providerDisplayName": "Fireworks", + "providerModelId": "accounts/fireworks/models/qwen2p5-vl-32b-instruct", + "providerGroup": "Fireworks", "quantization": null, "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, "maxPromptImages": null, @@ -35509,8 +37393,10 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", "top_k", + "repetition_penalty", + "response_format", + "structured_outputs", "logit_bias", "logprobs", "top_logprobs" @@ -35518,8 +37404,9 @@ export default { "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "termsOfServiceUrl": "https://fireworks.ai/terms-of-service", + "privacyPolicyUrl": "https://fireworks.ai/privacy-policy", + "dataPolicyUrl": "https://docs.fireworks.ai/guides/security_compliance/data_handling", "paidModels": { "training": false, "retainsPrompts": false @@ -35528,8 +37415,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000003", - "completion": "0.00000009", + "prompt": "0.0000009", + "completion": "0.0000009", "image": "0", "request": "0", "webSearch": "0", @@ -35554,24 +37441,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 113, - "newest": 217, - "throughputHighToLow": 103, - "latencyLowToHigh": 91, - "pricingLowToHigh": 86, - "pricingHighToLow": 232 + "topWeekly": 114, + "newest": 73, + "throughputHighToLow": 129, + "latencyLowToHigh": 161, + "pricingLowToHigh": 32, + "pricingHighToLow": 103 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", + "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", + "slug": "fireworks", "quantization": null, - "context": 131072, + "context": 128000, "maxCompletionTokens": null, - "providerModelId": "microsoft/Phi-3.5-mini-instruct", + "providerModelId": "accounts/fireworks/models/qwen2p5-vl-32b-instruct", "pricing": { - "prompt": "0.00000003", - "completion": "0.00000009", + "prompt": "0.0000009", + "completion": "0.0000009", "image": "0", "request": "0", "webSearch": "0", @@ -35585,25 +37474,30 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", "top_k", + "repetition_penalty", + "response_format", + "structured_outputs", "logit_bias", "logprobs", "top_logprobs" ], - "inputCost": 0.03, - "outputCost": 0.09 + "inputCost": 0.9, + "outputCost": 0.9, + "throughput": 61.0595, + "latency": 937 }, { - "name": "Azure", - "slug": "azure", - "quantization": "unknown", + "name": "InoCloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inocloud.com/&size=256", + "slug": "inoCloud", + "quantization": "bf16", "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "phi-35-mini-128k-instruct", + "maxCompletionTokens": 128000, + "providerModelId": "qwen/qwen2.5-vl-32b-instruct", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", + "prompt": "0.0000011", + "completion": "0.0000011", "image": "0", "request": "0", "webSearch": "0", @@ -35611,117 +37505,98 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", - "top_p" + "top_p", + "stop", + "frequency_penalty", + "presence_penalty" ], - "inputCost": 0.1, - "outputCost": 0.1 + "inputCost": 1.1, + "outputCost": 1.1, + "throughput": 26.49, + "latency": 3097 } ] }, { - "slug": "deepseek/deepseek-r1-distill-qwen-32b", - "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-01-29T23:53:50.865297+00:00", + "slug": "qwen/qwen-2.5-vl-7b-instruct", + "hfSlug": "Qwen/Qwen2.5-VL-7B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-08-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: R1 Distill Qwen 32B", - "shortName": "R1 Distill Qwen 32B", - "author": "deepseek", - "description": "DeepSeek R1 Distill Qwen 32B is a distilled large language model based on [Qwen 2.5 32B](https://huggingface.co/Qwen/Qwen2.5-32B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 72.6\n- MATH-500 pass@1: 94.3\n- CodeForces Rating: 1691\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", - "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", - "contextLength": 131072, + "name": "Qwen: Qwen2.5-VL 7B Instruct", + "shortName": "Qwen2.5-VL 7B Instruct", + "author": "qwen", + "description": "Qwen2.5 VL 7B is a multimodal LLM from the Qwen Team with the following key enhancements:\n\n- SoTA understanding of images of various resolution & ratio: Qwen2.5-VL achieves state-of-the-art performance on visual understanding benchmarks, including MathVista, DocVQA, RealWorldQA, MTVQA, etc.\n\n- Understanding videos of 20min+: Qwen2.5-VL can understand videos over 20 minutes for high-quality video-based question answering, dialog, content creation, etc.\n\n- Agent that can operate your mobiles, robots, etc.: with the abilities of complex reasoning and decision making, Qwen2.5-VL can be integrated with devices like mobile phones, robots, etc., for automatic operation based on visual environment and text instructions.\n\n- Multilingual Support: to serve global users, besides English and Chinese, Qwen2.5-VL now supports the understanding of texts in different languages inside images, including most European languages, Japanese, Korean, Arabic, Vietnamese, etc.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2-vl/) and [GitHub repo](https://github.com/QwenLM/Qwen2-VL).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "modelVersionGroupId": null, + "contextLength": 32768, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, "group": "Qwen", - "instructType": "deepseek-r1", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "deepseek/deepseek-r1-distill-qwen-32b", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "permaslug": "qwen/qwen-2-vl-7b-instruct", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "ce318fe0-48e0-444d-af14-08708e9e1f76", - "name": "DeepInfra | deepseek/deepseek-r1-distill-qwen-32b", - "contextLength": 131072, + "id": "fae5bc3c-f799-4657-8d05-3cf6f489ed0c", + "name": "Hyperbolic | qwen/qwen-2-vl-7b-instruct", + "contextLength": 32768, "model": { - "slug": "deepseek/deepseek-r1-distill-qwen-32b", - "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-01-29T23:53:50.865297+00:00", + "slug": "qwen/qwen-2.5-vl-7b-instruct", + "hfSlug": "Qwen/Qwen2.5-VL-7B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-08-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: R1 Distill Qwen 32B", - "shortName": "R1 Distill Qwen 32B", - "author": "deepseek", - "description": "DeepSeek R1 Distill Qwen 32B is a distilled large language model based on [Qwen 2.5 32B](https://huggingface.co/Qwen/Qwen2.5-32B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 72.6\n- MATH-500 pass@1: 94.3\n- CodeForces Rating: 1691\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", - "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", - "contextLength": 128000, + "name": "Qwen: Qwen2.5-VL 7B Instruct", + "shortName": "Qwen2.5-VL 7B Instruct", + "author": "qwen", + "description": "Qwen2.5 VL 7B is a multimodal LLM from the Qwen Team with the following key enhancements:\n\n- SoTA understanding of images of various resolution & ratio: Qwen2.5-VL achieves state-of-the-art performance on visual understanding benchmarks, including MathVista, DocVQA, RealWorldQA, MTVQA, etc.\n\n- Understanding videos of 20min+: Qwen2.5-VL can understand videos over 20 minutes for high-quality video-based question answering, dialog, content creation, etc.\n\n- Agent that can operate your mobiles, robots, etc.: with the abilities of complex reasoning and decision making, Qwen2.5-VL can be integrated with devices like mobile phones, robots, etc., for automatic operation based on visual environment and text instructions.\n\n- Multilingual Support: to serve global users, besides English and Chinese, Qwen2.5-VL now supports the understanding of texts in different languages inside images, including most European languages, Japanese, Korean, Arabic, Vietnamese, etc.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2-vl/) and [GitHub repo](https://github.com/QwenLM/Qwen2-VL).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "modelVersionGroupId": null, + "contextLength": 32768, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, "group": "Qwen", - "instructType": "deepseek-r1", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "deepseek/deepseek-r1-distill-qwen-32b", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "permaslug": "qwen/qwen-2-vl-7b-instruct", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "deepseek/deepseek-r1-distill-qwen-32b", - "modelVariantPermaslug": "deepseek/deepseek-r1-distill-qwen-32b", - "providerName": "DeepInfra", + "modelVariantSlug": "qwen/qwen-2.5-vl-7b-instruct", + "modelVariantPermaslug": "qwen/qwen-2-vl-7b-instruct", + "providerName": "Hyperbolic", "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", + "name": "Hyperbolic", + "displayName": "Hyperbolic", + "slug": "hyperbolic", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "privacyPolicyUrl": "https://hyperbolic.xyz/privacy", + "termsOfServiceUrl": "https://hyperbolic.xyz/terms", "paidModels": { - "training": false, - "retainsPrompts": false + "training": false } }, "headquarters": "US", @@ -35729,7 +37604,7 @@ export default { "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "DeepInfra", + "group": "Hyperbolic", "editors": [], "owners": [], "isMultipartSupported": true, @@ -35737,52 +37612,49 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/DeepInfra.webp" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256" } }, - "providerDisplayName": "DeepInfra", - "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", - "providerGroup": "DeepInfra", - "quantization": "fp8", + "providerDisplayName": "Hyperbolic", + "providerModelId": "Qwen/Qwen2.5-VL-7B-Instruct", + "providerGroup": "Hyperbolic", + "quantization": "bf16", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 16384, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", + "logprobs", + "top_logprobs", "seed", - "min_p" + "logit_bias", + "top_k", + "min_p", + "repetition_penalty" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "privacyPolicyUrl": "https://hyperbolic.xyz/privacy", + "termsOfServiceUrl": "https://hyperbolic.xyz/terms", "paidModels": { - "training": false, - "retainsPrompts": false + "training": false }, - "training": false, - "retainsPrompts": false + "training": false }, "pricing": { - "prompt": "0.00000012", - "completion": "0.00000018", - "image": "0", + "prompt": "0.0000002", + "completion": "0.0000002", + "image": "0.0001445", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -35793,37 +37665,40 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 114, - "newest": 133, - "throughputHighToLow": 49, - "latencyLowToHigh": 74, - "pricingLowToHigh": 50, - "pricingHighToLow": 192 + "topWeekly": 115, + "newest": 214, + "throughputHighToLow": 80, + "latencyLowToHigh": 153, + "pricingLowToHigh": 65, + "pricingHighToLow": 171 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 16384, - "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", + "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", + "slug": "hyperbolic", + "quantization": "bf16", + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "Qwen/Qwen2.5-VL-7B-Instruct", "pricing": { - "prompt": "0.00000012", - "completion": "0.00000018", - "image": "0", + "prompt": "0.0000002", + "completion": "0.0000002", + "image": "0.0001445", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -35833,30 +37708,33 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", + "logprobs", + "top_logprobs", "seed", - "min_p" + "logit_bias", + "top_k", + "min_p", + "repetition_penalty" ], - "inputCost": 0.12, - "outputCost": 0.18 + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 49.225, + "latency": 2066 }, { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": null, - "context": 64000, - "maxCompletionTokens": 64000, - "providerModelId": "deepseek/deepseek-r1-distill-qwen-32b", + "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", + "slug": "inferenceNet", + "quantization": "bf16", + "context": 128000, + "maxCompletionTokens": 8192, + "providerModelId": "qwen/qwen2.5-7b-instruct/bf-16", "pricing": { - "prompt": "0.0000003", - "completion": "0.0000003", + "prompt": "0.0000002", + "completion": "0.0000002", "image": "0", "request": "0", "webSearch": "0", @@ -35867,224 +37745,31 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", "seed", "top_k", "min_p", - "repetition_penalty", - "logit_bias" + "logit_bias", + "top_logprobs" ], - "inputCost": 0.3, - "outputCost": 0.3 + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 58.101, + "latency": 3841 }, { - "name": "Cloudflare", - "slug": "cloudflare", + "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", + "slug": "klusterAi", "quantization": null, - "context": 80000, - "maxCompletionTokens": null, - "providerModelId": "@cf/deepseek-ai/deepseek-r1-distill-qwen-32b", - "pricing": { - "prompt": "0.0000005", - "completion": "0.00000488", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "top_k", - "seed", - "repetition_penalty", - "frequency_penalty", - "presence_penalty" - ], - "inputCost": 0.5, - "outputCost": 4.88 - } - ] - }, - { - "slug": "thedrummer/skyfall-36b-v2", - "hfSlug": "TheDrummer/Skyfall-36B-v2", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-10T19:56:06.00791+00:00", - "hfUpdatedAt": null, - "name": "TheDrummer: Skyfall 36B V2", - "shortName": "Skyfall 36B V2", - "author": "thedrummer", - "description": "Skyfall 36B v2 is an enhanced iteration of Mistral Small 2501, specifically fine-tuned for improved creativity, nuanced writing, role-playing, and coherent storytelling.", - "modelVersionGroupId": null, - "contextLength": 32768, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "thedrummer/skyfall-36b-v2", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "1eb01ded-ae11-49e6-8aa6-3067584070bd", - "name": "Parasail | thedrummer/skyfall-36b-v2", - "contextLength": 32768, - "model": { - "slug": "thedrummer/skyfall-36b-v2", - "hfSlug": "TheDrummer/Skyfall-36B-v2", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-10T19:56:06.00791+00:00", - "hfUpdatedAt": null, - "name": "TheDrummer: Skyfall 36B V2", - "shortName": "Skyfall 36B V2", - "author": "thedrummer", - "description": "Skyfall 36B v2 is an enhanced iteration of Mistral Small 2501, specifically fine-tuned for improved creativity, nuanced writing, role-playing, and coherent storytelling.", - "modelVersionGroupId": null, - "contextLength": 32768, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "thedrummer/skyfall-36b-v2", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "thedrummer/skyfall-36b-v2", - "modelVariantPermaslug": "thedrummer/skyfall-36b-v2", - "providerName": "Parasail", - "providerInfo": { - "name": "Parasail", - "displayName": "Parasail", - "slug": "parasail", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://www.parasail.io/legal/terms", - "privacyPolicyUrl": "https://www.parasail.io/legal/privacy-policy", - "paidModels": { - "training": false, - "retainsPrompts": true - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "Parasail", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256" - } - }, - "providerDisplayName": "Parasail", - "providerModelId": "parasail-skyfall-36b-v2-fp8", - "providerGroup": "Parasail", - "quantization": "fp8", - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 32768, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "presence_penalty", - "frequency_penalty", - "repetition_penalty", - "top_k" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://www.parasail.io/legal/terms", - "privacyPolicyUrl": "https://www.parasail.io/legal/privacy-policy", - "paidModels": { - "training": false, - "retainsPrompts": true - }, - "training": false, - "retainsPrompts": true - }, - "pricing": { - "prompt": "0.0000005", - "completion": "0.0000008", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 115, - "newest": 96, - "throughputHighToLow": 152, - "latencyLowToHigh": 156, - "pricingLowToHigh": 184, - "pricingHighToLow": 136 - }, - "providers": [ - { - "name": "Parasail", - "slug": "parasail", - "quantization": "fp8", "context": 32768, "maxCompletionTokens": 32768, - "providerModelId": "parasail-skyfall-36b-v2-fp8", + "providerModelId": "Qwen/Qwen2.5-VL-7B-Instruct", "pricing": { - "prompt": "0.0000005", - "completion": "0.0000008", + "prompt": "0.0000003", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -36095,13 +37780,21 @@ export default { "max_tokens", "temperature", "top_p", - "presence_penalty", + "stop", "frequency_penalty", + "presence_penalty", + "top_k", "repetition_penalty", - "top_k" + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", + "seed" ], - "inputCost": 0.5, - "outputCost": 0.8 + "inputCost": 0.3, + "outputCost": 0.3, + "throughput": 39.5905, + "latency": 2843.5 } ] }, @@ -36268,14 +37961,16 @@ export default { "sorting": { "topWeekly": 116, "newest": 167, - "throughputHighToLow": 142, - "latencyLowToHigh": 83, + "throughputHighToLow": 131, + "latencyLowToHigh": 97, "pricingLowToHigh": 243, "pricingHighToLow": 72 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", "slug": "mistral", "quantization": null, "context": 131072, @@ -36304,22 +37999,24 @@ export default { "seed" ], "inputCost": 2, - "outputCost": 6 + "outputCost": 6, + "throughput": 59.5, + "latency": 630 } ] }, { - "slug": "qwen/qwen3-32b", - "hfSlug": "Qwen/Qwen3-32B", - "updatedAt": "2025-05-11T22:45:44.273213+00:00", - "createdAt": "2025-04-28T21:32:25.189881+00:00", + "slug": "deepseek/deepseek-r1-distill-qwen-32b", + "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-01-29T23:53:50.865297+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 32B (free)", - "shortName": "Qwen3 32B (free)", - "author": "qwen", - "description": "Qwen3-32B is a dense 32.8B parameter causal language model from the Qwen3 series, optimized for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, coding, and logical inference, and a \"non-thinking\" mode for faster, general-purpose conversation. The model demonstrates strong performance in instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling. ", - "modelVersionGroupId": null, - "contextLength": 40960, + "name": "DeepSeek: R1 Distill Qwen 32B", + "shortName": "R1 Distill Qwen 32B", + "author": "deepseek", + "description": "DeepSeek R1 Distill Qwen 32B is a distilled large language model based on [Qwen 2.5 32B](https://huggingface.co/Qwen/Qwen2.5-32B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 72.6\n- MATH-500 pass@1: 94.3\n- CodeForces Rating: 1691\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", + "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", + "contextLength": 131072, "inputModalities": [ "text" ], @@ -36327,17 +38024,17 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Qwen", + "instructType": "deepseek-r1", "defaultSystem": null, "defaultStops": [ - "<|im_start|>", - "<|im_end|>" + "<｜User｜>", + "<｜end▁of▁sentence｜>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen3-32b-04-28", + "permaslug": "deepseek/deepseek-r1-distill-qwen-32b", "reasoningConfig": { "startToken": "", "endToken": "" @@ -36349,21 +38046,21 @@ export default { } }, "endpoint": { - "id": "9021227d-9e2a-4718-83d7-3423c7031b49", - "name": "Chutes | qwen/qwen3-32b-04-28:free", - "contextLength": 40960, + "id": "ce318fe0-48e0-444d-af14-08708e9e1f76", + "name": "DeepInfra | deepseek/deepseek-r1-distill-qwen-32b", + "contextLength": 131072, "model": { - "slug": "qwen/qwen3-32b", - "hfSlug": "Qwen/Qwen3-32B", - "updatedAt": "2025-05-11T22:45:44.273213+00:00", - "createdAt": "2025-04-28T21:32:25.189881+00:00", + "slug": "deepseek/deepseek-r1-distill-qwen-32b", + "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-01-29T23:53:50.865297+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 32B", - "shortName": "Qwen3 32B", - "author": "qwen", - "description": "Qwen3-32B is a dense 32.8B parameter causal language model from the Qwen3 series, optimized for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, coding, and logical inference, and a \"non-thinking\" mode for faster, general-purpose conversation. The model demonstrates strong performance in instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling. ", - "modelVersionGroupId": null, - "contextLength": 131072, + "name": "DeepSeek: R1 Distill Qwen 32B", + "shortName": "R1 Distill Qwen 32B", + "author": "deepseek", + "description": "DeepSeek R1 Distill Qwen 32B is a distilled large language model based on [Qwen 2.5 32B](https://huggingface.co/Qwen/Qwen2.5-32B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 72.6\n- MATH-500 pass@1: 94.3\n- CodeForces Rating: 1691\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", + "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", + "contextLength": 128000, "inputModalities": [ "text" ], @@ -36371,17 +38068,17 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Qwen", + "instructType": "deepseek-r1", "defaultSystem": null, "defaultStops": [ - "<|im_start|>", - "<|im_end|>" + "<｜User｜>", + "<｜end▁of▁sentence｜>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen3-32b-04-28", + "permaslug": "deepseek/deepseek-r1-distill-qwen-32b", "reasoningConfig": { "startToken": "", "endToken": "" @@ -36393,26 +38090,29 @@ export default { } } }, - "modelVariantSlug": "qwen/qwen3-32b:free", - "modelVariantPermaslug": "qwen/qwen3-32b-04-28:free", - "providerName": "Chutes", + "modelVariantSlug": "deepseek/deepseek-r1-distill-qwen-32b", + "modelVariantPermaslug": "deepseek/deepseek-r1-distill-qwen-32b", + "providerName": "DeepInfra", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "DeepInfra", + "displayName": "DeepInfra", + "slug": "deepinfra", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false } }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Chutes", + "group": "DeepInfra", "editors": [], "owners": [], "isMultipartSupported": true, @@ -36420,18 +38120,18 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "/images/icons/DeepInfra.webp" } }, - "providerDisplayName": "Chutes", - "providerModelId": "Qwen/Qwen3-32B", - "providerGroup": "Chutes", - "quantization": null, - "variant": "free", - "isFree": true, + "providerDisplayName": "DeepInfra", + "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", + "providerGroup": "DeepInfra", + "quantization": "fp8", + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 16384, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -36443,28 +38143,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "min_p", "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" + "response_format", + "top_k", + "seed", + "min_p" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.00000012", + "completion": "0.00000018", "image": "0", "request": "0", "webSearch": "0", @@ -36483,30 +38183,31 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 40, - "newest": 28, - "throughputHighToLow": 141, - "latencyLowToHigh": 173, - "pricingLowToHigh": 12, - "pricingHighToLow": 194 + "topWeekly": 117, + "newest": 133, + "throughputHighToLow": 53, + "latencyLowToHigh": 118, + "pricingLowToHigh": 50, + "pricingHighToLow": 192 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen3-32B", + "context": 131072, + "maxCompletionTokens": 16384, + "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", + "prompt": "0.00000012", + "completion": "0.00000018", "image": "0", "request": "0", "webSearch": "0", @@ -36528,53 +38229,22 @@ export default { "seed", "min_p" ], - "inputCost": 0.1, - "outputCost": 0.3 - }, - { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen3-32B", - "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" - ], - "inputCost": 0.1, - "outputCost": 0.3 + "inputCost": 0.12, + "outputCost": 0.18, + "throughput": 45.5815, + "latency": 668 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", - "quantization": "fp8", - "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "qwen/qwen3-32b-fp8", + "quantization": null, + "context": 64000, + "maxCompletionTokens": 64000, + "providerModelId": "deepseek/deepseek-r1-distill-qwen-32b", "pricing": { - "prompt": "0.0000001", - "completion": "0.00000045", + "prompt": "0.0000003", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -36596,19 +38266,22 @@ export default { "repetition_penalty", "logit_bias" ], - "inputCost": 0.1, - "outputCost": 0.45 + "inputCost": 0.3, + "outputCost": 0.3, + "throughput": 19.1495, + "latency": 2299.5 }, { - "name": "Parasail", - "slug": "parasail", - "quantization": "fp8", - "context": 40960, - "maxCompletionTokens": 40960, - "providerModelId": "parasail-qwen3-32b", + "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", + "slug": "cloudflare", + "quantization": null, + "context": 80000, + "maxCompletionTokens": null, + "providerModelId": "@cf/deepseek-ai/deepseek-r1-distill-qwen-32b", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000005", + "prompt": "0.0000005", + "completion": "0.00000488", "image": "0", "request": "0", "webSearch": "0", @@ -36621,149 +38294,96 @@ export default { "top_p", "reasoning", "include_reasoning", - "presence_penalty", - "frequency_penalty", + "top_k", + "seed", "repetition_penalty", - "top_k" + "frequency_penalty", + "presence_penalty" ], - "inputCost": 0.1, - "outputCost": 0.5 - }, - { - "name": "GMICloud", - "slug": "gmiCloud", - "quantization": "fp8", - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen3-32B-FP8", - "pricing": { - "prompt": "0.0000003", - "completion": "0.0000006", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "seed" - ], - "inputCost": 0.3, - "outputCost": 0.6 - }, - { - "name": "SambaNova", - "slug": "sambaNova", - "quantization": null, - "context": 8192, - "maxCompletionTokens": 4096, - "providerModelId": "Qwen3-32B", - "pricing": { - "prompt": "0.0000004", - "completion": "0.0000008", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "top_k", - "stop" - ], - "inputCost": 0.4, - "outputCost": 0.8 + "inputCost": 0.5, + "outputCost": 4.88, + "throughput": 32.7715, + "latency": 867 } ] }, { - "slug": "qwen/qwen-2.5-vl-7b-instruct", - "hfSlug": "Qwen/Qwen2.5-VL-7B-Instruct", + "slug": "thedrummer/skyfall-36b-v2", + "hfSlug": "TheDrummer/Skyfall-36B-v2", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-28T00:00:00+00:00", + "createdAt": "2025-03-10T19:56:06.00791+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen2.5-VL 7B Instruct", - "shortName": "Qwen2.5-VL 7B Instruct", - "author": "qwen", - "description": "Qwen2.5 VL 7B is a multimodal LLM from the Qwen Team with the following key enhancements:\n\n- SoTA understanding of images of various resolution & ratio: Qwen2.5-VL achieves state-of-the-art performance on visual understanding benchmarks, including MathVista, DocVQA, RealWorldQA, MTVQA, etc.\n\n- Understanding videos of 20min+: Qwen2.5-VL can understand videos over 20 minutes for high-quality video-based question answering, dialog, content creation, etc.\n\n- Agent that can operate your mobiles, robots, etc.: with the abilities of complex reasoning and decision making, Qwen2.5-VL can be integrated with devices like mobile phones, robots, etc., for automatic operation based on visual environment and text instructions.\n\n- Multilingual Support: to serve global users, besides English and Chinese, Qwen2.5-VL now supports the understanding of texts in different languages inside images, including most European languages, Japanese, Korean, Arabic, Vietnamese, etc.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2-vl/) and [GitHub repo](https://github.com/QwenLM/Qwen2-VL).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "name": "TheDrummer: Skyfall 36B V2", + "shortName": "Skyfall 36B V2", + "author": "thedrummer", + "description": "Skyfall 36B v2 is an enhanced iteration of Mistral Small 2501, specifically fine-tuned for improved creativity, nuanced writing, role-playing, and coherent storytelling.", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-2-vl-7b-instruct", + "permaslug": "thedrummer/skyfall-36b-v2", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "fae5bc3c-f799-4657-8d05-3cf6f489ed0c", - "name": "Hyperbolic | qwen/qwen-2-vl-7b-instruct", + "id": "1eb01ded-ae11-49e6-8aa6-3067584070bd", + "name": "Parasail | thedrummer/skyfall-36b-v2", "contextLength": 32768, "model": { - "slug": "qwen/qwen-2.5-vl-7b-instruct", - "hfSlug": "Qwen/Qwen2.5-VL-7B-Instruct", + "slug": "thedrummer/skyfall-36b-v2", + "hfSlug": "TheDrummer/Skyfall-36B-v2", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-28T00:00:00+00:00", + "createdAt": "2025-03-10T19:56:06.00791+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen2.5-VL 7B Instruct", - "shortName": "Qwen2.5-VL 7B Instruct", - "author": "qwen", - "description": "Qwen2.5 VL 7B is a multimodal LLM from the Qwen Team with the following key enhancements:\n\n- SoTA understanding of images of various resolution & ratio: Qwen2.5-VL achieves state-of-the-art performance on visual understanding benchmarks, including MathVista, DocVQA, RealWorldQA, MTVQA, etc.\n\n- Understanding videos of 20min+: Qwen2.5-VL can understand videos over 20 minutes for high-quality video-based question answering, dialog, content creation, etc.\n\n- Agent that can operate your mobiles, robots, etc.: with the abilities of complex reasoning and decision making, Qwen2.5-VL can be integrated with devices like mobile phones, robots, etc., for automatic operation based on visual environment and text instructions.\n\n- Multilingual Support: to serve global users, besides English and Chinese, Qwen2.5-VL now supports the understanding of texts in different languages inside images, including most European languages, Japanese, Korean, Arabic, Vietnamese, etc.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2-vl/) and [GitHub repo](https://github.com/QwenLM/Qwen2-VL).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "name": "TheDrummer: Skyfall 36B V2", + "shortName": "Skyfall 36B V2", + "author": "thedrummer", + "description": "Skyfall 36B v2 is an enhanced iteration of Mistral Small 2501, specifically fine-tuned for improved creativity, nuanced writing, role-playing, and coherent storytelling.", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-2-vl-7b-instruct", + "permaslug": "thedrummer/skyfall-36b-v2", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "qwen/qwen-2.5-vl-7b-instruct", - "modelVariantPermaslug": "qwen/qwen-2-vl-7b-instruct", - "providerName": "Hyperbolic", + "modelVariantSlug": "thedrummer/skyfall-36b-v2", + "modelVariantPermaslug": "thedrummer/skyfall-36b-v2", + "providerName": "Parasail", "providerInfo": { - "name": "Hyperbolic", - "displayName": "Hyperbolic", - "slug": "hyperbolic", + "name": "Parasail", + "displayName": "Parasail", + "slug": "parasail", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://hyperbolic.xyz/privacy", - "termsOfServiceUrl": "https://hyperbolic.xyz/terms", + "termsOfServiceUrl": "https://www.parasail.io/legal/terms", + "privacyPolicyUrl": "https://www.parasail.io/legal/privacy-policy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": true } }, "headquarters": "US", @@ -36771,7 +38391,7 @@ export default { "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Hyperbolic", + "group": "Parasail", "editors": [], "owners": [], "isMultipartSupported": true, @@ -36779,49 +38399,45 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256" } }, - "providerDisplayName": "Hyperbolic", - "providerModelId": "Qwen/Qwen2.5-VL-7B-Instruct", - "providerGroup": "Hyperbolic", - "quantization": "bf16", + "providerDisplayName": "Parasail", + "providerModelId": "parasail-skyfall-36b-v2-fp8", + "providerGroup": "Parasail", + "quantization": "fp8", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 32768, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", "presence_penalty", - "logprobs", - "top_logprobs", - "seed", - "logit_bias", - "top_k", - "min_p", - "repetition_penalty" + "frequency_penalty", + "repetition_penalty", + "top_k" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://hyperbolic.xyz/privacy", - "termsOfServiceUrl": "https://hyperbolic.xyz/terms", + "termsOfServiceUrl": "https://www.parasail.io/legal/terms", + "privacyPolicyUrl": "https://www.parasail.io/legal/privacy-policy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": true }, - "training": false + "training": false, + "retainsPrompts": true }, "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", - "image": "0.0001445", + "prompt": "0.0000005", + "completion": "0.0000008", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -36839,64 +38455,31 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 118, - "newest": 214, - "throughputHighToLow": 74, - "latencyLowToHigh": 126, - "pricingLowToHigh": 65, - "pricingHighToLow": 168 + "newest": 96, + "throughputHighToLow": 118, + "latencyLowToHigh": 124, + "pricingLowToHigh": 184, + "pricingHighToLow": 136 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { - "name": "Hyperbolic", - "slug": "hyperbolic", - "quantization": "bf16", + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", + "quantization": "fp8", "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen2.5-VL-7B-Instruct", - "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", - "image": "0.0001445", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "logprobs", - "top_logprobs", - "seed", - "logit_bias", - "top_k", - "min_p", - "repetition_penalty" - ], - "inputCost": 0.2, - "outputCost": 0.2 - }, - { - "name": "inference.net", - "slug": "inferenceNet", - "quantization": "bf16", - "context": 128000, - "maxCompletionTokens": 8192, - "providerModelId": "qwen/qwen2.5-7b-instruct/bf-16", + "maxCompletionTokens": 32768, + "providerModelId": "parasail-skyfall-36b-v2-fp8", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", + "prompt": "0.0000005", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -36907,147 +38490,139 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "min_p", - "logit_bias", - "top_logprobs" - ], - "inputCost": 0.2, - "outputCost": 0.2 - }, - { - "name": "kluster.ai", - "slug": "klusterAi", - "quantization": null, - "context": 32768, - "maxCompletionTokens": 32768, - "providerModelId": "Qwen/Qwen2.5-VL-7B-Instruct", - "pricing": { - "prompt": "0.0000003", - "completion": "0.0000003", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature" + "frequency_penalty", + "repetition_penalty", + "top_k" ], - "inputCost": 0.3, - "outputCost": 0.3 + "inputCost": 0.5, + "outputCost": 0.8, + "throughput": 66.5415, + "latency": 728 } ] }, { - "slug": "qwen/qwen2.5-vl-32b-instruct", - "hfSlug": "Qwen/Qwen2.5-VL-32B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-24T18:10:38.542849+00:00", + "slug": "qwen/qwen3-32b", + "hfSlug": "Qwen/Qwen3-32B", + "updatedAt": "2025-05-11T22:45:44.273213+00:00", + "createdAt": "2025-04-28T21:32:25.189881+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen2.5 VL 32B Instruct", - "shortName": "Qwen2.5 VL 32B Instruct", + "name": "Qwen: Qwen3 32B (free)", + "shortName": "Qwen3 32B (free)", "author": "qwen", - "description": "Qwen2.5-VL-32B is a multimodal vision-language model fine-tuned through reinforcement learning for enhanced mathematical reasoning, structured outputs, and visual problem-solving capabilities. It excels at visual analysis tasks, including object recognition, textual interpretation within images, and precise event localization in extended videos. Qwen2.5-VL-32B demonstrates state-of-the-art performance across multimodal benchmarks such as MMMU, MathVista, and VideoMME, while maintaining strong reasoning and clarity in text-based tasks like MMLU, mathematical problem-solving, and code generation.", + "description": "Qwen3-32B is a dense 32.8B parameter causal language model from the Qwen3 series, optimized for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, coding, and logical inference, and a \"non-thinking\" mode for faster, general-purpose conversation. The model demonstrates strong performance in instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling. ", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 40960, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": null, + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen2.5-vl-32b-instruct", - "reasoningConfig": null, - "features": {}, + "permaslug": "qwen/qwen3-32b-04-28", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "aac76f56-eedb-405f-a07c-5bc66b877c1e", - "name": "Fireworks | qwen/qwen2.5-vl-32b-instruct", - "contextLength": 128000, + "id": "9021227d-9e2a-4718-83d7-3423c7031b49", + "name": "Chutes | qwen/qwen3-32b-04-28:free", + "contextLength": 40960, "model": { - "slug": "qwen/qwen2.5-vl-32b-instruct", - "hfSlug": "Qwen/Qwen2.5-VL-32B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-24T18:10:38.542849+00:00", + "slug": "qwen/qwen3-32b", + "hfSlug": "Qwen/Qwen3-32B", + "updatedAt": "2025-05-11T22:45:44.273213+00:00", + "createdAt": "2025-04-28T21:32:25.189881+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen2.5 VL 32B Instruct", - "shortName": "Qwen2.5 VL 32B Instruct", + "name": "Qwen: Qwen3 32B", + "shortName": "Qwen3 32B", "author": "qwen", - "description": "Qwen2.5-VL-32B is a multimodal vision-language model fine-tuned through reinforcement learning for enhanced mathematical reasoning, structured outputs, and visual problem-solving capabilities. It excels at visual analysis tasks, including object recognition, textual interpretation within images, and precise event localization in extended videos. Qwen2.5-VL-32B demonstrates state-of-the-art performance across multimodal benchmarks such as MMMU, MathVista, and VideoMME, while maintaining strong reasoning and clarity in text-based tasks like MMLU, mathematical problem-solving, and code generation.", + "description": "Qwen3-32B is a dense 32.8B parameter causal language model from the Qwen3 series, optimized for both complex reasoning and efficient dialogue. It supports seamless switching between a \"thinking\" mode for tasks like math, coding, and logical inference, and a \"non-thinking\" mode for faster, general-purpose conversation. The model demonstrates strong performance in instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling. ", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": null, + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen2.5-vl-32b-instruct", - "reasoningConfig": null, - "features": {} + "permaslug": "qwen/qwen3-32b-04-28", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "qwen/qwen2.5-vl-32b-instruct", - "modelVariantPermaslug": "qwen/qwen2.5-vl-32b-instruct", - "providerName": "Fireworks", + "modelVariantSlug": "qwen/qwen3-32b:free", + "modelVariantPermaslug": "qwen/qwen3-32b-04-28:free", + "providerName": "Chutes", "providerInfo": { - "name": "Fireworks", - "displayName": "Fireworks", - "slug": "fireworks", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://fireworks.ai/terms-of-service", - "privacyPolicyUrl": "https://fireworks.ai/privacy-policy", - "dataPolicyUrl": "https://docs.fireworks.ai/guides/security_compliance/data_handling", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Fireworks", + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.fireworks.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Fireworks.png" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "Fireworks", - "providerModelId": "accounts/fireworks/models/qwen2p5-vl-32b-instruct", - "providerGroup": "Fireworks", + "providerDisplayName": "Chutes", + "providerModelId": "Qwen/Qwen3-32B", + "providerGroup": "Chutes", "quantization": null, - "variant": "standard", - "isFree": false, + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, @@ -37057,33 +38632,33 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", + "seed", "top_k", + "min_p", "repetition_penalty", - "response_format", - "structured_outputs", - "logit_bias", "logprobs", + "logit_bias", "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://fireworks.ai/terms-of-service", - "privacyPolicyUrl": "https://fireworks.ai/privacy-policy", - "dataPolicyUrl": "https://docs.fireworks.ai/guides/security_compliance/data_handling", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true }, - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.0000009", - "completion": "0.0000009", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -37095,7 +38670,7 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, @@ -37108,24 +38683,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 119, - "newest": 73, - "throughputHighToLow": 131, - "latencyLowToHigh": 118, - "pricingLowToHigh": 32, - "pricingHighToLow": 102 + "topWeekly": 43, + "newest": 28, + "throughputHighToLow": 133, + "latencyLowToHigh": 142, + "pricingLowToHigh": 12, + "pricingHighToLow": 194 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { - "name": "Fireworks", - "slug": "fireworks", - "quantization": null, - "context": 128000, + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "fp8", + "context": 40960, "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/qwen2p5-vl-32b-instruct", + "providerModelId": "Qwen/Qwen3-32B", "pricing": { - "prompt": "0.0000009", - "completion": "0.0000009", + "prompt": "0.0000001", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -37136,67 +38713,273 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "top_k", "repetition_penalty", "response_format", - "structured_outputs", + "top_k", + "seed", + "min_p" + ], + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 45.105, + "latency": 873 + }, + { + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 40960, + "maxCompletionTokens": null, + "providerModelId": "Qwen/Qwen3-32B", + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", "logit_bias", "logprobs", "top_logprobs" ], - "inputCost": 0.9, - "outputCost": 0.9 - } - ] - }, - { - "slug": "openai/gpt-3.5-turbo", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-05-28T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "OpenAI: GPT-3.5 Turbo", - "shortName": "GPT-3.5 Turbo", - "author": "openai", - "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.", - "modelVersionGroupId": null, - "contextLength": 16385, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "GPT", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "openai/gpt-3.5-turbo", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "3a632f37-731d-4200-9e38-413a5f5dd39d", - "name": "OpenAI | openai/gpt-3.5-turbo", - "contextLength": 16385, - "model": { - "slug": "openai/gpt-3.5-turbo", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-05-28T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "OpenAI: GPT-3.5 Turbo", - "shortName": "GPT-3.5 Turbo", - "author": "openai", - "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.", + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 31.8215, + "latency": 802 + }, + { + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": "fp8", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "qwen/qwen3-32b-fp8", + "pricing": { + "prompt": "0.0000001", + "completion": "0.00000045", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" + ], + "inputCost": 0.1, + "outputCost": 0.45, + "throughput": 30.374, + "latency": 1003.5 + }, + { + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", + "quantization": "fp8", + "context": 40960, + "maxCompletionTokens": 40960, + "providerModelId": "parasail-qwen3-32b", + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000005", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "presence_penalty", + "frequency_penalty", + "repetition_penalty", + "top_k" + ], + "inputCost": 0.1, + "outputCost": 0.5, + "throughput": 49.6365, + "latency": 813 + }, + { + "name": "GMICloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://gmicloud.ai/&size=256", + "slug": "gmiCloud", + "quantization": "fp8", + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "Qwen/Qwen3-32B-FP8", + "pricing": { + "prompt": "0.0000003", + "completion": "0.0000006", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "seed" + ], + "inputCost": 0.3, + "outputCost": 0.6, + "throughput": 57.572, + "latency": 734 + }, + { + "name": "Cerebras", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cerebras.ai/&size=256", + "slug": "cerebras", + "quantization": null, + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "qwen-3-32b", + "pricing": { + "prompt": "0.0000004", + "completion": "0.0000008", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "seed", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.4, + "outputCost": 0.8, + "throughput": 2145.3435, + "latency": 420 + }, + { + "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", + "slug": "sambaNova", + "quantization": null, + "context": 8192, + "maxCompletionTokens": 4096, + "providerModelId": "Qwen3-32B", + "pricing": { + "prompt": "0.0000004", + "completion": "0.0000008", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "top_k", + "stop" + ], + "inputCost": 0.4, + "outputCost": 0.8, + "throughput": 320.4805, + "latency": 1143 + } + ] + }, + { + "slug": "microsoft/phi-3.5-mini-128k-instruct", + "hfSlug": "microsoft/Phi-3.5-mini-instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-08-21T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Microsoft: Phi-3.5 Mini 128K Instruct", + "shortName": "Phi-3.5 Mini 128K Instruct", + "author": "microsoft", + "description": "Phi-3.5 models are lightweight, state-of-the-art open models. These models were trained with Phi-3 datasets that include both synthetic data and the filtered, publicly available websites data, with a focus on high quality and reasoning-dense properties. Phi-3.5 Mini uses 3.8B parameters, and is a dense decoder-only transformer model using the same tokenizer as [Phi-3 Mini](/models/microsoft/phi-3-mini-128k-instruct).\n\nThe models underwent a rigorous enhancement process, incorporating both supervised fine-tuning, proximal policy optimization, and direct preference optimization to ensure precise instruction adherence and robust safety measures. When assessed against benchmarks that test common sense, language understanding, math, code, long context and logical reasoning, Phi-3.5 models showcased robust and state-of-the-art performance among models with less than 13 billion parameters.", + "modelVersionGroupId": null, + "contextLength": 131072, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Other", + "instructType": "phi3", + "defaultSystem": null, + "defaultStops": [ + "<|end|>", + "<|endoftext|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "microsoft/phi-3.5-mini-128k-instruct", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "c284390d-1916-4a33-9bd7-b49c8165c2ca", + "name": "Nebius | microsoft/phi-3.5-mini-128k-instruct", + "contextLength": 131072, + "model": { + "slug": "microsoft/phi-3.5-mini-128k-instruct", + "hfSlug": "microsoft/Phi-3.5-mini-instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-08-21T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Microsoft: Phi-3.5 Mini 128K Instruct", + "shortName": "Phi-3.5 Mini 128K Instruct", + "author": "microsoft", + "description": "Phi-3.5 models are lightweight, state-of-the-art open models. These models were trained with Phi-3 datasets that include both synthetic data and the filtered, publicly available websites data, with a focus on high quality and reasoning-dense properties. Phi-3.5 Mini uses 3.8B parameters, and is a dense decoder-only transformer model using the same tokenizer as [Phi-3 Mini](/models/microsoft/phi-3-mini-128k-instruct).\n\nThe models underwent a rigorous enhancement process, incorporating both supervised fine-tuning, proximal policy optimization, and direct preference optimization to ensure precise instruction adherence and robust safety measures. When assessed against benchmarks that test common sense, language understanding, math, code, long context and logical reasoning, Phi-3.5 models showcased robust and state-of-the-art performance among models with less than 13 billion parameters.", "modelVersionGroupId": null, - "contextLength": 16385, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -37204,67 +38987,65 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, + "group": "Other", + "instructType": "phi3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|end|>", + "<|endoftext|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-3.5-turbo", + "permaslug": "microsoft/phi-3.5-mini-128k-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/gpt-3.5-turbo", - "modelVariantPermaslug": "openai/gpt-3.5-turbo", - "providerName": "OpenAI", + "modelVariantSlug": "microsoft/phi-3.5-mini-128k-instruct", + "modelVariantPermaslug": "microsoft/phi-3.5-mini-128k-instruct", + "providerName": "Nebius", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Nebius", + "displayName": "Nebius AI Studio", + "slug": "nebius", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "retainsPrompts": false + } }, - "headquarters": "US", + "headquarters": "DE", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", + "isAbortable": false, + "moderationRequired": false, + "group": "Nebius", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "invertRequired": true } }, - "providerDisplayName": "OpenAI", - "providerModelId": "gpt-3.5-turbo", - "providerGroup": "OpenAI", - "quantization": "unknown", + "providerDisplayName": "Nebius AI Studio", + "providerModelId": "microsoft/Phi-3.5-mini-instruct", + "providerGroup": "Nebius", + "quantization": null, "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -37272,30 +39053,26 @@ export default { "frequency_penalty", "presence_penalty", "seed", + "top_k", "logit_bias", "logprobs", - "top_logprobs", - "response_format" + "top_logprobs" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, - "requiresUserIds": true, "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, "pricing": { - "prompt": "0.0000005", - "completion": "0.0000015", + "prompt": "0.00000003", + "completion": "0.00000009", "image": "0", "request": "0", "webSearch": "0", @@ -37306,7 +39083,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -37314,29 +39091,32 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 120, - "newest": 317, - "throughputHighToLow": 54, - "latencyLowToHigh": 53, - "pricingLowToHigh": 189, - "pricingHighToLow": 130 + "newest": 217, + "throughputHighToLow": 16, + "latencyLowToHigh": 24, + "pricingLowToHigh": 86, + "pricingHighToLow": 232 }, + "authorIcon": "https://openrouter.ai/images/icons/Microsoft.svg", "providers": [ { - "name": "OpenAI", - "slug": "openAi", - "quantization": "unknown", - "context": 16385, - "maxCompletionTokens": 4096, - "providerModelId": "gpt-3.5-turbo", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": null, + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "microsoft/Phi-3.5-mini-instruct", "pricing": { - "prompt": "0.0000005", - "completion": "0.0000015", + "prompt": "0.00000003", + "completion": "0.00000009", "image": "0", "request": "0", "webSearch": "0", @@ -37344,8 +39124,6 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -37353,160 +39131,193 @@ export default { "frequency_penalty", "presence_penalty", "seed", + "top_k", "logit_bias", "logprobs", - "top_logprobs", - "response_format" + "top_logprobs" ], - "inputCost": 0.5, - "outputCost": 1.5 + "inputCost": 0.03, + "outputCost": 0.09, + "throughput": 154.1785, + "latency": 428 + }, + { + "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", + "slug": "azure", + "quantization": "unknown", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "phi-35-mini-128k-instruct", + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p" + ], + "inputCost": 0.1, + "outputCost": 0.1, + "throughput": 17.7345, + "latency": 1411.5 } ] }, { - "slug": "openai/o3", - "hfSlug": "", - "updatedAt": "2025-05-02T21:34:35.534922+00:00", - "createdAt": "2025-04-16T17:10:57.049467+00:00", + "slug": "qwen/qwq-32b", + "hfSlug": "Qwen/QwQ-32B", + "updatedAt": "2025-05-02T17:24:23.70143+00:00", + "createdAt": "2025-03-05T21:06:54.875499+00:00", "hfUpdatedAt": null, - "name": "OpenAI: o3", - "shortName": "o3", - "author": "openai", - "description": "o3 is a well-rounded and powerful model across domains. It sets a new standard for math, science, coding, and visual reasoning tasks. It also excels at technical writing and instruction-following. Use it to think through multi-step problems that involve analysis across text, code, and images. Note that BYOK is required for this model. Set up here: https://openrouter.ai/settings/integrations", + "name": "Qwen: QwQ 32B (free)", + "shortName": "QwQ 32B (free)", + "author": "qwen", + "description": "QwQ is the reasoning model of the Qwen series. Compared with conventional instruction-tuned models, QwQ, which is capable of thinking and reasoning, can achieve significantly enhanced performance in downstream tasks, especially hard problems. QwQ-32B is the medium-sized reasoning model, which is capable of achieving competitive performance against state-of-the-art reasoning models, e.g., DeepSeek-R1, o1-mini.", "modelVersionGroupId": null, - "contextLength": 200000, + "contextLength": 40000, "inputModalities": [ - "image", - "text", - "file" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Qwen", + "instructType": "qwq", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>" + ], "hidden": false, "router": null, - "warningMessage": "OpenAI requires bringing your own API key to use o3 over the API. Set up here: https://openrouter.ai/settings/integrations", - "permaslug": "openai/o3-2025-04-16", - "reasoningConfig": null, - "features": {}, + "warningMessage": null, + "permaslug": "qwen/qwq-32b", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "42e72619-d01c-411c-a201-f991644768b7", - "name": "OpenAI | openai/o3-2025-04-16", - "contextLength": 200000, + "id": "bdb92948-0d84-423b-8204-ecc8df0669ef", + "name": "Nineteen | qwen/qwq-32b:free", + "contextLength": 40000, "model": { - "slug": "openai/o3", - "hfSlug": "", - "updatedAt": "2025-05-02T21:34:35.534922+00:00", - "createdAt": "2025-04-16T17:10:57.049467+00:00", + "slug": "qwen/qwq-32b", + "hfSlug": "Qwen/QwQ-32B", + "updatedAt": "2025-05-02T17:24:23.70143+00:00", + "createdAt": "2025-03-05T21:06:54.875499+00:00", "hfUpdatedAt": null, - "name": "OpenAI: o3", - "shortName": "o3", - "author": "openai", - "description": "o3 is a well-rounded and powerful model across domains. It sets a new standard for math, science, coding, and visual reasoning tasks. It also excels at technical writing and instruction-following. Use it to think through multi-step problems that involve analysis across text, code, and images. Note that BYOK is required for this model. Set up here: https://openrouter.ai/settings/integrations", + "name": "Qwen: QwQ 32B", + "shortName": "QwQ 32B", + "author": "qwen", + "description": "QwQ is the reasoning model of the Qwen series. Compared with conventional instruction-tuned models, QwQ, which is capable of thinking and reasoning, can achieve significantly enhanced performance in downstream tasks, especially hard problems. QwQ-32B is the medium-sized reasoning model, which is capable of achieving competitive performance against state-of-the-art reasoning models, e.g., DeepSeek-R1, o1-mini.", "modelVersionGroupId": null, - "contextLength": 200000, + "contextLength": 131072, "inputModalities": [ - "image", - "text", - "file" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Qwen", + "instructType": "qwq", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>" + ], "hidden": false, "router": null, - "warningMessage": "OpenAI requires bringing your own API key to use o3 over the API. Set up here: https://openrouter.ai/settings/integrations", - "permaslug": "openai/o3-2025-04-16", - "reasoningConfig": null, - "features": {} + "warningMessage": null, + "permaslug": "qwen/qwq-32b", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "openai/o3", - "modelVariantPermaslug": "openai/o3-2025-04-16", - "providerName": "OpenAI", + "modelVariantSlug": "qwen/qwq-32b:free", + "modelVariantPermaslug": "qwen/qwq-32b:free", + "providerName": "Nineteen", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Nineteen", + "displayName": "Nineteen", + "slug": "nineteen", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://nineteen.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "training": false + } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", + "moderationRequired": false, + "group": "Nineteen", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nineteen.ai/&size=256" } }, - "providerDisplayName": "OpenAI", - "providerModelId": "o3-2025-04-16", - "providerGroup": "OpenAI", - "quantization": null, - "variant": "standard", - "isFree": false, + "providerDisplayName": "Nineteen", + "providerModelId": "Qwen/QwQ-32B", + "providerGroup": "Nineteen", + "quantization": "bf16", + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 100000, + "maxCompletionTokens": 40000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", - "seed", "max_tokens", - "response_format", - "structured_outputs" + "temperature", + "top_p", + "reasoning", + "include_reasoning" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://nineteen.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, "pricing": { - "prompt": "0.00001", - "completion": "0.00004", - "image": "0.00765", + "prompt": "0", + "completion": "0", + "image": "0", "request": "0", - "inputCacheRead": "0.0000025", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -37515,7 +39326,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, @@ -37523,36 +39334,107 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": { - "responseFormat": true, - "structuredOutputs": true - }, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 121, - "newest": 44, - "throughputHighToLow": 199, - "latencyLowToHigh": 298, - "pricingLowToHigh": 305, - "pricingHighToLow": 13 + "topWeekly": 65, + "newest": 102, + "throughputHighToLow": 51, + "latencyLowToHigh": 83, + "pricingLowToHigh": 43, + "pricingHighToLow": 183 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { - "name": "OpenAI", - "slug": "openAi", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "Qwen/QwQ-32B", + "pricing": { + "prompt": "0.00000015", + "completion": "0.0000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" + ], + "inputCost": 0.15, + "outputCost": 0.2, + "throughput": 39.107, + "latency": 345 + }, + { + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "Qwen/QwQ-32B", + "pricing": { + "prompt": "0.00000015", + "completion": "0.00000045", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.15, + "outputCost": 0.45, + "throughput": 32.285, + "latency": 405 + }, + { + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", "quantization": null, - "context": 200000, - "maxCompletionTokens": 100000, - "providerModelId": "o3-2025-04-16", + "context": 32768, + "maxCompletionTokens": 16000, + "providerModelId": "qwen/qwq-32b", "pricing": { - "prompt": "0.00001", - "completion": "0.00004", - "image": "0.00765", + "prompt": "0.00000018", + "completion": "0.0000002", + "image": "0", "request": "0", - "inputCacheRead": "0.0000025", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -37560,28 +39442,330 @@ export default { "supportedParameters": [ "tools", "tool_choice", + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" + ], + "inputCost": 0.18, + "outputCost": 0.2, + "throughput": 33.979, + "latency": 715 + }, + { + "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", + "slug": "inferenceNet", + "quantization": "fp8", + "context": 16384, + "maxCompletionTokens": 16384, + "providerModelId": "qwen/qwq-32b/fp-8", + "pricing": { + "prompt": "0.0000002", + "completion": "0.0000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", "seed", + "top_k", + "min_p", + "logit_bias", + "top_logprobs" + ], + "inputCost": 0.2, + "outputCost": 0.2 + }, + { + "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", + "slug": "groq", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "qwen-qwq-32b", + "pricing": { + "prompt": "0.00000029", + "completion": "0.00000039", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", "response_format", - "structured_outputs" + "top_logprobs", + "logprobs", + "logit_bias", + "seed" ], - "inputCost": 10, - "outputCost": 40 + "inputCost": 0.29, + "outputCost": 0.39, + "throughput": 563.8855, + "latency": 696.5 + }, + { + "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", + "slug": "hyperbolic", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "Qwen/QwQ-32B", + "pricing": { + "prompt": "0.0000004", + "completion": "0.0000004", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "logprobs", + "top_logprobs", + "seed", + "logit_bias", + "top_k", + "min_p", + "repetition_penalty" + ], + "inputCost": 0.4, + "outputCost": 0.4, + "throughput": 33.561, + "latency": 1027 + }, + { + "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", + "slug": "sambaNova", + "quantization": "bf16", + "context": 16384, + "maxCompletionTokens": 4096, + "providerModelId": "QwQ-32B", + "pricing": { + "prompt": "0.0000005", + "completion": "0.000001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "top_k", + "stop" + ], + "inputCost": 0.5, + "outputCost": 1, + "throughput": 262.86, + "latency": 681 + }, + { + "name": "Nebius AI Studio (Fast)", + "icon": "", + "slug": "nebiusAiStudio (fast)", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "Qwen/QwQ-32B-fast", + "pricing": { + "prompt": "0.0000005", + "completion": "0.0000015", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.5, + "outputCost": 1.5, + "throughput": 75.252, + "latency": 285 + }, + { + "name": "CentML", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://centml.ai/&size=256", + "slug": "centMl", + "quantization": "fp16", + "context": 40960, + "maxCompletionTokens": 40960, + "providerModelId": "Qwen/QwQ-32B", + "pricing": { + "prompt": "0.00000065", + "completion": "0.00000065", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty" + ], + "inputCost": 0.65, + "outputCost": 0.65, + "throughput": 58.551, + "latency": 606 + }, + { + "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", + "slug": "fireworks", + "quantization": null, + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "accounts/fireworks/models/qwq-32b", + "pricing": { + "prompt": "0.0000009", + "completion": "0.0000009", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "response_format", + "structured_outputs", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.9, + "outputCost": 0.9, + "throughput": 133.165, + "latency": 705 + }, + { + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 32768, + "providerModelId": "Qwen/QwQ-32B", + "pricing": { + "prompt": "0.0000012", + "completion": "0.0000012", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" + ], + "inputCost": 1.2, + "outputCost": 1.2, + "throughput": 88.417, + "latency": 962 } ] }, { - "slug": "openai/o3-mini-high", - "hfSlug": "", + "slug": "mistralai/ministral-3b", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-12T15:03:31.504126+00:00", + "createdAt": "2024-10-17T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenAI: o3 Mini High", - "shortName": "o3 Mini High", - "author": "openai", - "description": "OpenAI o3-mini-high is the same model as [o3-mini](/openai/o3-mini) with reasoning_effort set to high. \n\no3-mini is a cost-efficient language model optimized for STEM reasoning tasks, particularly excelling in science, mathematics, and coding. The model features three adjustable reasoning effort levels and supports key developer capabilities including function calling, structured outputs, and streaming, though it does not include vision processing capabilities.\n\nThe model demonstrates significant improvements over its predecessor, with expert testers preferring its responses 56% of the time and noting a 39% reduction in major errors on complex questions. With medium reasoning effort settings, o3-mini matches the performance of the larger o1 model on challenging reasoning evaluations like AIME and GPQA, while maintaining lower latency and cost.", + "name": "Mistral: Ministral 3B", + "shortName": "Ministral 3B", + "author": "mistralai", + "description": "Ministral 3B is a 3B parameter model optimized for on-device and edge computing. It excels in knowledge, commonsense reasoning, and function-calling, outperforming larger models like Mistral 7B on most benchmarks. Supporting up to 128k context length, it’s ideal for orchestrating agentic workflows and specialist tasks with efficient inference.", "modelVersionGroupId": null, - "contextLength": 200000, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -37589,32 +39773,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "Mistral", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/o3-mini-high-2025-01-31", + "permaslug": "mistralai/ministral-3b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "4a8663f2-89f3-40e9-9b50-e4838fff0155", - "name": "OpenAI | openai/o3-mini-high-2025-01-31", - "contextLength": 200000, + "id": "71ce5f13-6ef2-4cbe-a650-66d2f11a4ecb", + "name": "Mistral | mistralai/ministral-3b", + "contextLength": 131072, "model": { - "slug": "openai/o3-mini-high", - "hfSlug": "", + "slug": "mistralai/ministral-3b", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-12T15:03:31.504126+00:00", + "createdAt": "2024-10-17T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenAI: o3 Mini High", - "shortName": "o3 Mini High", - "author": "openai", - "description": "OpenAI o3-mini-high is the same model as [o3-mini](/openai/o3-mini) with reasoning_effort set to high. \n\no3-mini is a cost-efficient language model optimized for STEM reasoning tasks, particularly excelling in science, mathematics, and coding. The model features three adjustable reasoning effort levels and supports key developer capabilities including function calling, structured outputs, and streaming, though it does not include vision processing capabilities.\n\nThe model demonstrates significant improvements over its predecessor, with expert testers preferring its responses 56% of the time and noting a 39% reduction in major errors on complex questions. With medium reasoning effort settings, o3-mini matches the performance of the larger o1 model on challenging reasoning evaluations like AIME and GPQA, while maintaining lower latency and cost.", + "name": "Mistral: Ministral 3B", + "shortName": "Ministral 3B", + "author": "mistralai", + "description": "Ministral 3B is a 3B parameter model optimized for on-device and edge computing. It excels in knowledge, commonsense reasoning, and function-calling, outperforming larger models like Mistral 7B on most benchmarks. Supporting up to 128k context length, it’s ideal for orchestrating agentic workflows and specialist tasks with efficient inference.", "modelVersionGroupId": null, - "contextLength": 200000, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -37622,94 +39806,93 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "Mistral", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/o3-mini-high-2025-01-31", + "permaslug": "mistralai/ministral-3b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/o3-mini-high", - "modelVariantPermaslug": "openai/o3-mini-high-2025-01-31", - "providerName": "OpenAI", + "modelVariantSlug": "mistralai/ministral-3b", + "modelVariantPermaslug": "mistralai/ministral-3b", + "providerName": "Mistral", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Mistral", + "displayName": "Mistral", + "slug": "mistral", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 - }, - "requiresUserIds": true + } }, - "headquarters": "US", + "headquarters": "FR", "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", + "hasCompletions": false, + "isAbortable": false, + "moderationRequired": false, + "group": "Mistral", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true + "url": "/images/icons/Mistral.png" } }, - "providerDisplayName": "OpenAI", - "providerModelId": "o3-mini-2025-01-31", - "providerGroup": "OpenAI", - "quantization": null, + "providerDisplayName": "Mistral", + "providerModelId": "ministral-3b-2410", + "providerGroup": "Mistral", + "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 100000, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "tools", "tool_choice", - "seed", "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "response_format", - "structured_outputs" + "structured_outputs", + "seed" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 }, - "requiresUserIds": true, "training": false, "retainsPrompts": true, "retentionDays": 30 }, "pricing": { - "prompt": "0.0000011", - "completion": "0.0000044", + "prompt": "0.00000004", + "completion": "0.00000004", "image": "0", "request": "0", - "inputCacheRead": "0.00000055", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -37723,35 +39906,37 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 122, - "newest": 116, - "throughputHighToLow": 312, - "latencyLowToHigh": 309, - "pricingLowToHigh": 231, - "pricingHighToLow": 86 + "newest": 187, + "throughputHighToLow": 7, + "latencyLowToHigh": 4, + "pricingLowToHigh": 88, + "pricingHighToLow": 231 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { - "name": "OpenAI", - "slug": "openAi", - "quantization": null, - "context": 200000, - "maxCompletionTokens": 100000, - "providerModelId": "o3-mini-2025-01-31", + "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", + "slug": "mistral", + "quantization": "unknown", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "ministral-3b-2410", "pricing": { - "prompt": "0.0000011", - "completion": "0.0000044", + "prompt": "0.00000004", + "completion": "0.00000004", "image": "0", "request": "0", - "inputCacheRead": "0.00000055", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -37759,28 +39944,35 @@ export default { "supportedParameters": [ "tools", "tool_choice", - "seed", "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "response_format", - "structured_outputs" + "structured_outputs", + "seed" ], - "inputCost": 1.1, - "outputCost": 4.4 + "inputCost": 0.04, + "outputCost": 0.04, + "throughput": 246.022, + "latency": 167 } ] }, { - "slug": "mistralai/ministral-3b", + "slug": "openai/gpt-3.5-turbo", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-10-17T00:00:00+00:00", + "createdAt": "2023-05-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Mistral: Ministral 3B", - "shortName": "Ministral 3B", - "author": "mistralai", - "description": "Ministral 3B is a 3B parameter model optimized for on-device and edge computing. It excels in knowledge, commonsense reasoning, and function-calling, outperforming larger models like Mistral 7B on most benchmarks. Supporting up to 128k context length, it’s ideal for orchestrating agentic workflows and specialist tasks with efficient inference.", + "name": "OpenAI: GPT-3.5 Turbo", + "shortName": "GPT-3.5 Turbo", + "author": "openai", + "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 16385, "inputModalities": [ "text" ], @@ -37788,32 +39980,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/ministral-3b", + "permaslug": "openai/gpt-3.5-turbo", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "71ce5f13-6ef2-4cbe-a650-66d2f11a4ecb", - "name": "Mistral | mistralai/ministral-3b", - "contextLength": 131072, + "id": "3a632f37-731d-4200-9e38-413a5f5dd39d", + "name": "OpenAI | openai/gpt-3.5-turbo", + "contextLength": 16385, "model": { - "slug": "mistralai/ministral-3b", + "slug": "openai/gpt-3.5-turbo", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-10-17T00:00:00+00:00", + "createdAt": "2023-05-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Mistral: Ministral 3B", - "shortName": "Ministral 3B", - "author": "mistralai", - "description": "Ministral 3B is a 3B parameter model optimized for on-device and edge computing. It excels in knowledge, commonsense reasoning, and function-calling, outperforming larger models like Mistral 7B on most benchmarks. Supporting up to 128k context length, it’s ideal for orchestrating agentic workflows and specialist tasks with efficient inference.", + "name": "OpenAI: GPT-3.5 Turbo", + "shortName": "GPT-3.5 Turbo", + "author": "openai", + "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 16385, "inputModalities": [ "text" ], @@ -37821,59 +40013,62 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/ministral-3b", + "permaslug": "openai/gpt-3.5-turbo", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "mistralai/ministral-3b", - "modelVariantPermaslug": "mistralai/ministral-3b", - "providerName": "Mistral", + "modelVariantSlug": "openai/gpt-3.5-turbo", + "modelVariantPermaslug": "openai/gpt-3.5-turbo", + "providerName": "OpenAI", "providerInfo": { - "name": "Mistral", - "displayName": "Mistral", - "slug": "mistral", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 - } + }, + "requiresUserIds": true }, - "headquarters": "FR", + "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, - "moderationRequired": false, - "group": "Mistral", + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Mistral.png" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Mistral", - "providerModelId": "ministral-3b-2410", - "providerGroup": "Mistral", + "providerDisplayName": "OpenAI", + "providerModelId": "gpt-3.5-turbo", + "providerGroup": "OpenAI", "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -37885,27 +40080,31 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "response_format", - "structured_outputs", - "seed" + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 }, + "requiresUserIds": true, "training": false, "retainsPrompts": true, "retentionDays": 30 }, "pricing": { - "prompt": "0.00000004", - "completion": "0.00000004", + "prompt": "0.0000005", + "completion": "0.0000015", "image": "0", "request": "0", "webSearch": "0", @@ -37921,33 +40120,34 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 123, - "newest": 186, - "throughputHighToLow": 7, - "latencyLowToHigh": 5, - "pricingLowToHigh": 88, - "pricingHighToLow": 231 + "newest": 315, + "throughputHighToLow": 48, + "latencyLowToHigh": 54, + "pricingLowToHigh": 190, + "pricingHighToLow": 129 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "Mistral", - "slug": "mistral", + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", "quantization": "unknown", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "ministral-3b-2410", + "context": 16385, + "maxCompletionTokens": 4096, + "providerModelId": "gpt-3.5-turbo", "pricing": { - "prompt": "0.00000004", - "completion": "0.00000004", + "prompt": "0.0000005", + "completion": "0.0000015", "image": "0", "request": "0", "webSearch": "0", @@ -37963,145 +40163,159 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "response_format", - "structured_outputs", - "seed" + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" ], - "inputCost": 0.04, - "outputCost": 0.04 + "inputCost": 0.5, + "outputCost": 1.5, + "throughput": 128.897, + "latency": 421 } ] }, { - "slug": "qwen/qwen-vl-plus", + "slug": "openai/o3-mini-high", "hfSlug": "", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-05T04:54:15.216448+00:00", + "createdAt": "2025-02-12T15:03:31.504126+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen VL Plus", - "shortName": "Qwen VL Plus", - "author": "qwen", - "description": "Qwen's Enhanced Large Visual Language Model. Significantly upgraded for detailed recognition capabilities and text recognition abilities, supporting ultra-high pixel resolutions up to millions of pixels and extreme aspect ratios for image input. It delivers significant performance across a broad range of visual tasks.\n", + "name": "OpenAI: o3 Mini High", + "shortName": "o3 Mini High", + "author": "openai", + "description": "OpenAI o3-mini-high is the same model as [o3-mini](/openai/o3-mini) with reasoning_effort set to high. \n\no3-mini is a cost-efficient language model optimized for STEM reasoning tasks, particularly excelling in science, mathematics, and coding. The model features three adjustable reasoning effort levels and supports key developer capabilities including function calling, structured outputs, and streaming, though it does not include vision processing capabilities.\n\nThe model demonstrates significant improvements over its predecessor, with expert testers preferring its responses 56% of the time and noting a 39% reduction in major errors on complex questions. With medium reasoning effort settings, o3-mini matches the performance of the larger o1 model on challenging reasoning evaluations like AIME and GPQA, while maintaining lower latency and cost.", "modelVersionGroupId": null, - "contextLength": 7500, + "contextLength": 200000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-vl-plus", + "permaslug": "openai/o3-mini-high-2025-01-31", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "df9c15ac-870b-40e5-aa43-e4b3b44951f7", - "name": "Alibaba | qwen/qwen-vl-plus", - "contextLength": 7500, + "id": "4a8663f2-89f3-40e9-9b50-e4838fff0155", + "name": "OpenAI | openai/o3-mini-high-2025-01-31", + "contextLength": 200000, "model": { - "slug": "qwen/qwen-vl-plus", + "slug": "openai/o3-mini-high", "hfSlug": "", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-05T04:54:15.216448+00:00", + "createdAt": "2025-02-12T15:03:31.504126+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen VL Plus", - "shortName": "Qwen VL Plus", - "author": "qwen", - "description": "Qwen's Enhanced Large Visual Language Model. Significantly upgraded for detailed recognition capabilities and text recognition abilities, supporting ultra-high pixel resolutions up to millions of pixels and extreme aspect ratios for image input. It delivers significant performance across a broad range of visual tasks.\n", + "name": "OpenAI: o3 Mini High", + "shortName": "o3 Mini High", + "author": "openai", + "description": "OpenAI o3-mini-high is the same model as [o3-mini](/openai/o3-mini) with reasoning_effort set to high. \n\no3-mini is a cost-efficient language model optimized for STEM reasoning tasks, particularly excelling in science, mathematics, and coding. The model features three adjustable reasoning effort levels and supports key developer capabilities including function calling, structured outputs, and streaming, though it does not include vision processing capabilities.\n\nThe model demonstrates significant improvements over its predecessor, with expert testers preferring its responses 56% of the time and noting a 39% reduction in major errors on complex questions. With medium reasoning effort settings, o3-mini matches the performance of the larger o1 model on challenging reasoning evaluations like AIME and GPQA, while maintaining lower latency and cost.", "modelVersionGroupId": null, - "contextLength": 7500, + "contextLength": 200000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-vl-plus", + "permaslug": "openai/o3-mini-high-2025-01-31", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "qwen/qwen-vl-plus", - "modelVariantPermaslug": "qwen/qwen-vl-plus", - "providerName": "Alibaba", + "modelVariantSlug": "openai/o3-mini-high", + "modelVariantPermaslug": "openai/o3-mini-high-2025-01-31", + "providerName": "OpenAI", "providerInfo": { - "name": "Alibaba", - "displayName": "Alibaba", - "slug": "alibaba", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", - "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { - "training": false - } + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, - "headquarters": "CN", + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, - "moderationRequired": false, - "group": "Alibaba", + "isAbortable": true, + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Alibaba", - "providerModelId": "qwen-vl-plus", - "providerGroup": "Alibaba", + "providerDisplayName": "OpenAI", + "providerModelId": "o3-mini-2025-01-31", + "providerGroup": "OpenAI", "quantization": null, "variant": "standard", "isFree": false, - "canAbort": false, - "maxPromptTokens": 6000, - "maxCompletionTokens": 1500, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": 100000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", + "tools", + "tool_choice", "seed", + "max_tokens", "response_format", - "presence_penalty" + "structured_outputs" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", - "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { - "training": false + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "training": false + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.00000021", - "completion": "0.00000063", - "image": "0.0002688", + "prompt": "0.0000011", + "completion": "0.0000044", + "image": "0", "request": "0", + "inputCacheRead": "0.00000055", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -38110,12 +40324,12 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -38124,187 +40338,394 @@ export default { }, "sorting": { "topWeekly": 124, - "newest": 119, - "throughputHighToLow": 95, - "latencyLowToHigh": 161, - "pricingLowToHigh": 159, - "pricingHighToLow": 159 + "newest": 116, + "throughputHighToLow": 312, + "latencyLowToHigh": 303, + "pricingLowToHigh": 231, + "pricingHighToLow": 86 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "Alibaba", - "slug": "alibaba", + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", "quantization": null, - "context": 7500, - "maxCompletionTokens": 1500, - "providerModelId": "qwen-vl-plus", + "context": 200000, + "maxCompletionTokens": 100000, + "providerModelId": "o3-mini-2025-01-31", "pricing": { - "prompt": "0.00000021", - "completion": "0.00000063", - "image": "0.0002688", + "prompt": "0.0000011", + "completion": "0.0000044", + "image": "0", "request": "0", + "inputCacheRead": "0.00000055", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", + "tools", + "tool_choice", "seed", + "max_tokens", "response_format", - "presence_penalty" + "structured_outputs" ], - "inputCost": 0.21, - "outputCost": 0.63 + "inputCost": 1.1, + "outputCost": 4.4, + "throughput": 95.3105, + "latency": 8220 } ] }, { - "slug": "qwen/qwq-32b", - "hfSlug": "Qwen/QwQ-32B", - "updatedAt": "2025-05-02T17:24:23.70143+00:00", - "createdAt": "2025-03-05T21:06:54.875499+00:00", + "slug": "openai/o1", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-12-17T18:26:39.576639+00:00", "hfUpdatedAt": null, - "name": "Qwen: QwQ 32B (free)", - "shortName": "QwQ 32B (free)", - "author": "qwen", - "description": "QwQ is the reasoning model of the Qwen series. Compared with conventional instruction-tuned models, QwQ, which is capable of thinking and reasoning, can achieve significantly enhanced performance in downstream tasks, especially hard problems. QwQ-32B is the medium-sized reasoning model, which is capable of achieving competitive performance against state-of-the-art reasoning models, e.g., DeepSeek-R1, o1-mini.", + "name": "OpenAI: o1", + "shortName": "o1", + "author": "openai", + "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding. The o1 model series is trained with large-scale reinforcement learning to reason using chain of thought. \n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n", "modelVersionGroupId": null, - "contextLength": 40000, + "contextLength": 200000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "qwq", + "group": "GPT", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>" - ], + "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "qwen/qwq-32b", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "warningMessage": "", + "permaslug": "openai/o1-2024-12-17", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "bdb92948-0d84-423b-8204-ecc8df0669ef", - "name": "Nineteen | qwen/qwq-32b:free", - "contextLength": 40000, + "id": "82738f61-f3cb-44a5-b5d1-e6787ae64e3b", + "name": "OpenAI | openai/o1-2024-12-17", + "contextLength": 200000, "model": { - "slug": "qwen/qwq-32b", - "hfSlug": "Qwen/QwQ-32B", - "updatedAt": "2025-05-02T17:24:23.70143+00:00", - "createdAt": "2025-03-05T21:06:54.875499+00:00", + "slug": "openai/o1", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-12-17T18:26:39.576639+00:00", "hfUpdatedAt": null, - "name": "Qwen: QwQ 32B", - "shortName": "QwQ 32B", - "author": "qwen", - "description": "QwQ is the reasoning model of the Qwen series. Compared with conventional instruction-tuned models, QwQ, which is capable of thinking and reasoning, can achieve significantly enhanced performance in downstream tasks, especially hard problems. QwQ-32B is the medium-sized reasoning model, which is capable of achieving competitive performance against state-of-the-art reasoning models, e.g., DeepSeek-R1, o1-mini.", + "name": "OpenAI: o1", + "shortName": "o1", + "author": "openai", + "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding. The o1 model series is trained with large-scale reinforcement learning to reason using chain of thought. \n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 200000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "qwq", + "group": "GPT", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>" - ], + "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "qwen/qwq-32b", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "warningMessage": "", + "permaslug": "openai/o1-2024-12-17", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "qwen/qwq-32b:free", - "modelVariantPermaslug": "qwen/qwq-32b:free", - "providerName": "Nineteen", + "modelVariantSlug": "openai/o1", + "modelVariantPermaslug": "openai/o1-2024-12-17", + "providerName": "OpenAI", "providerInfo": { - "name": "Nineteen", - "displayName": "Nineteen", - "slug": "nineteen", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://nineteen.ai/tos", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { - "training": false - } + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": false, - "group": "Nineteen", + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nineteen.ai/&size=256" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Nineteen", - "providerModelId": "Qwen/QwQ-32B", - "providerGroup": "Nineteen", - "quantization": "bf16", - "variant": "free", - "isFree": true, + "providerDisplayName": "OpenAI", + "providerModelId": "o1-2024-12-17", + "providerGroup": "OpenAI", + "quantization": "unknown", + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 40000, + "maxCompletionTokens": 100000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", + "seed", "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning" + "response_format", + "structured_outputs" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://nineteen.ai/tos", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { - "training": false + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "training": false + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0", - "completion": "0", - "image": "0", + "prompt": "0.000015", + "completion": "0.00006", + "image": "0.021675", + "request": "0", + "inputCacheRead": "0.0000075", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": true, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 125, + "newest": 151, + "throughputHighToLow": 314, + "latencyLowToHigh": 312, + "pricingLowToHigh": 306, + "pricingHighToLow": 10 + }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", + "providers": [ + { + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 100000, + "providerModelId": "o1-2024-12-17", + "pricing": { + "prompt": "0.000015", + "completion": "0.00006", + "image": "0.021675", + "request": "0", + "inputCacheRead": "0.0000075", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "seed", + "max_tokens", + "response_format", + "structured_outputs" + ], + "inputCost": 15, + "outputCost": 60, + "throughput": 25.3135, + "latency": 18105.5 + } + ] + }, + { + "slug": "qwen/qwen-2.5-72b-instruct", + "hfSlug": "Qwen/Qwen2.5-72B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-09-19T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Qwen2.5 72B Instruct (free)", + "shortName": "Qwen2.5 72B Instruct (free)", + "author": "qwen", + "description": "Qwen2.5 72B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "modelVersionGroupId": null, + "contextLength": 32768, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Qwen", + "instructType": "chatml", + "defaultSystem": null, + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "qwen/qwen-2.5-72b-instruct", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "e0e08390-a346-4bbf-a641-0ccdedface29", + "name": "Chutes | qwen/qwen-2.5-72b-instruct:free", + "contextLength": 32768, + "model": { + "slug": "qwen/qwen-2.5-72b-instruct", + "hfSlug": "Qwen/Qwen2.5-72B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-09-19T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Qwen2.5 72B Instruct", + "shortName": "Qwen2.5 72B Instruct", + "author": "qwen", + "description": "Qwen2.5 72B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "modelVersionGroupId": null, + "contextLength": 131072, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Qwen", + "instructType": "chatml", + "defaultSystem": null, + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "qwen/qwen-2.5-72b-instruct", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "qwen/qwen-2.5-72b-instruct:free", + "modelVariantPermaslug": "qwen/qwen-2.5-72b-instruct:free", + "providerName": "Chutes", + "providerInfo": { + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://chutes.ai/tos", + "paidModels": { + "training": true, + "retainsPrompts": true + } + }, + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "Chutes", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + } + }, + "providerDisplayName": "Chutes", + "providerModelId": "Qwen/Qwen2.5-72B-Instruct", + "providerGroup": "Chutes", + "quantization": "bf16", + "variant": "free", + "isFree": true, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://chutes.ai/tos", + "paidModels": { + "training": true, + "retainsPrompts": true + }, + "training": true, + "retainsPrompts": true + }, + "pricing": { + "prompt": "0", + "completion": "0", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -38315,7 +40736,7 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, @@ -38327,24 +40748,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 66, - "newest": 102, - "throughputHighToLow": 48, - "latencyLowToHigh": 98, - "pricingLowToHigh": 43, - "pricingHighToLow": 180 + "topWeekly": 31, + "newest": 204, + "throughputHighToLow": 221, + "latencyLowToHigh": 135, + "pricingLowToHigh": 64, + "pricingHighToLow": 186 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "Qwen/QwQ-32B", + "quantization": "fp8", + "context": 32768, + "maxCompletionTokens": 16384, + "providerModelId": "Qwen/Qwen2.5-72B-Instruct", "pricing": { - "prompt": "0.00000015", - "completion": "0.0000002", + "prompt": "0.00000012", + "completion": "0.00000039", "image": "0", "request": "0", "webSearch": "0", @@ -38352,11 +40775,11 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -38366,19 +40789,22 @@ export default { "seed", "min_p" ], - "inputCost": 0.15, - "outputCost": 0.2 + "inputCost": 0.12, + "outputCost": 0.39, + "throughput": 33.379, + "latency": 675 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 131072, "maxCompletionTokens": null, - "providerModelId": "Qwen/QwQ-32B", + "providerModelId": "Qwen/Qwen2.5-72B-Instruct", "pricing": { - "prompt": "0.00000015", - "completion": "0.00000045", + "prompt": "0.00000013", + "completion": "0.0000004", "image": "0", "request": "0", "webSearch": "0", @@ -38389,8 +40815,6 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -38400,19 +40824,22 @@ export default { "logprobs", "top_logprobs" ], - "inputCost": 0.15, - "outputCost": 0.45 + "inputCost": 0.13, + "outputCost": 0.4, + "throughput": 19.889, + "latency": 1954.5 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", - "quantization": null, - "context": 32768, - "maxCompletionTokens": 16000, - "providerModelId": "qwen/qwq-32b", + "quantization": "unknown", + "context": 32000, + "maxCompletionTokens": 4096, + "providerModelId": "qwen/qwen-2.5-72b-instruct", "pricing": { - "prompt": "0.00000018", - "completion": "0.0000002", + "prompt": "0.00000038", + "completion": "0.0000004", "image": "0", "request": "0", "webSearch": "0", @@ -38425,8 +40852,6 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -38436,84 +40861,19 @@ export default { "repetition_penalty", "logit_bias" ], - "inputCost": 0.18, - "outputCost": 0.2 - }, - { - "name": "inference.net", - "slug": "inferenceNet", - "quantization": "fp8", - "context": 16384, - "maxCompletionTokens": 16384, - "providerModelId": "qwen/qwq-32b/fp-8", - "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "logit_bias", - "top_logprobs" - ], - "inputCost": 0.2, - "outputCost": 0.2 - }, - { - "name": "Groq", - "slug": "groq", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "qwen-qwq-32b", - "pricing": { - "prompt": "0.00000029", - "completion": "0.00000039", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "top_logprobs", - "logprobs", - "logit_bias", - "seed" - ], - "inputCost": 0.29, - "outputCost": 0.39 + "inputCost": 0.38, + "outputCost": 0.4, + "throughput": 21.3825, + "latency": 4101 }, { "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", "slug": "hyperbolic", "quantization": "bf16", "context": 131072, "maxCompletionTokens": null, - "providerModelId": "Qwen/QwQ-32B", + "providerModelId": "Qwen/Qwen2.5-72B-Instruct", "pricing": { "prompt": "0.0000004", "completion": "0.0000004", @@ -38527,8 +40887,6 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -38541,110 +40899,18 @@ export default { "repetition_penalty" ], "inputCost": 0.4, - "outputCost": 0.4 - }, - { - "name": "SambaNova", - "slug": "sambaNova", - "quantization": "bf16", - "context": 16384, - "maxCompletionTokens": 4096, - "providerModelId": "QwQ-32B", - "pricing": { - "prompt": "0.0000005", - "completion": "0.000001", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "top_k", - "stop" - ], - "inputCost": 0.5, - "outputCost": 1 - }, - { - "name": "Nebius AI Studio (Fast)", - "slug": "nebiusAiStudio (fast)", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "Qwen/QwQ-32B-fast", - "pricing": { - "prompt": "0.0000005", - "completion": "0.0000015", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" - ], - "inputCost": 0.5, - "outputCost": 1.5 - }, - { - "name": "CentML", - "slug": "centMl", - "quantization": "fp16", - "context": 40960, - "maxCompletionTokens": 40960, - "providerModelId": "Qwen/QwQ-32B", - "pricing": { - "prompt": "0.00000065", - "completion": "0.00000065", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty" - ], - "inputCost": 0.65, - "outputCost": 0.65 + "outputCost": 0.4, + "throughput": 32.508, + "latency": 2157 }, { "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", "slug": "fireworks", - "quantization": null, - "context": 131072, + "quantization": "unknown", + "context": 32768, "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/qwq-32b", + "providerModelId": "accounts/fireworks/models/qwen2p5-72b-instruct", "pricing": { "prompt": "0.0000009", "completion": "0.0000009", @@ -38655,11 +40921,11 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -38672,15 +40938,18 @@ export default { "top_logprobs" ], "inputCost": 0.9, - "outputCost": 0.9 + "outputCost": 0.9, + "throughput": 70.3035, + "latency": 354 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", - "quantization": null, + "quantization": "fp8", "context": 131072, - "maxCompletionTokens": 32768, - "providerModelId": "Qwen/QwQ-32B", + "maxCompletionTokens": 2048, + "providerModelId": "Qwen/Qwen2.5-72B-Instruct-Turbo", "pricing": { "prompt": "0.0000012", "completion": "0.0000012", @@ -38694,8 +40963,6 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -38706,152 +40973,142 @@ export default { "response_format" ], "inputCost": 1.2, - "outputCost": 1.2 + "outputCost": 1.2, + "throughput": 95.0335, + "latency": 808 } ] }, { - "slug": "openai/o1", - "hfSlug": "", + "slug": "qwen/qwen-2.5-7b-instruct", + "hfSlug": "Qwen/Qwen2.5-7B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-17T18:26:39.576639+00:00", + "createdAt": "2024-10-16T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenAI: o1", - "shortName": "o1", - "author": "openai", - "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding. The o1 model series is trained with large-scale reinforcement learning to reason using chain of thought. \n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n", + "name": "Qwen2.5 7B Instruct (free)", + "shortName": "Qwen2.5 7B Instruct (free)", + "author": "qwen", + "description": "Qwen2.5 7B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", "modelVersionGroupId": null, - "contextLength": 200000, + "contextLength": 32768, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, + "group": "Qwen", + "instructType": "chatml", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "openai/o1-2024-12-17", + "warningMessage": null, + "permaslug": "qwen/qwen-2.5-7b-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "82738f61-f3cb-44a5-b5d1-e6787ae64e3b", - "name": "OpenAI | openai/o1-2024-12-17", - "contextLength": 200000, + "id": "8a40a3b8-80d8-4f75-bb8a-c0c710b31eb5", + "name": "Nineteen | qwen/qwen-2.5-7b-instruct:free", + "contextLength": 32768, "model": { - "slug": "openai/o1", - "hfSlug": "", + "slug": "qwen/qwen-2.5-7b-instruct", + "hfSlug": "Qwen/Qwen2.5-7B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-17T18:26:39.576639+00:00", + "createdAt": "2024-10-16T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenAI: o1", - "shortName": "o1", - "author": "openai", - "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding. The o1 model series is trained with large-scale reinforcement learning to reason using chain of thought. \n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n", + "name": "Qwen2.5 7B Instruct", + "shortName": "Qwen2.5 7B Instruct", + "author": "qwen", + "description": "Qwen2.5 7B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", "modelVersionGroupId": null, - "contextLength": 200000, + "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, + "group": "Qwen", + "instructType": "chatml", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "openai/o1-2024-12-17", + "warningMessage": null, + "permaslug": "qwen/qwen-2.5-7b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/o1", - "modelVariantPermaslug": "openai/o1-2024-12-17", - "providerName": "OpenAI", + "modelVariantSlug": "qwen/qwen-2.5-7b-instruct:free", + "modelVariantPermaslug": "qwen/qwen-2.5-7b-instruct:free", + "providerName": "Nineteen", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Nineteen", + "displayName": "Nineteen", + "slug": "nineteen", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://nineteen.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "training": false + } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", + "moderationRequired": false, + "group": "Nineteen", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nineteen.ai/&size=256" } }, - "providerDisplayName": "OpenAI", - "providerModelId": "o1-2024-12-17", - "providerGroup": "OpenAI", - "quantization": "unknown", - "variant": "standard", - "isFree": false, + "providerDisplayName": "Nineteen", + "providerModelId": "Qwen/Qwen2.5-7B-Instruct", + "providerGroup": "Nineteen", + "quantization": "bf16", + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 100000, + "maxCompletionTokens": 32768, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", - "seed", "max_tokens", - "response_format", - "structured_outputs" + "temperature", + "top_p" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://nineteen.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, "pricing": { - "prompt": "0.000015", - "completion": "0.00006", - "image": "0.021675", + "prompt": "0", + "completion": "0", + "image": "0", "request": "0", - "inputCacheRead": "0.0000075", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -38860,7 +41117,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -38868,228 +41125,133 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 126, - "newest": 151, - "throughputHighToLow": 314, - "latencyLowToHigh": 312, - "pricingLowToHigh": 306, - "pricingHighToLow": 10 + "topWeekly": 22, + "newest": 188, + "throughputHighToLow": 5, + "latencyLowToHigh": 89, + "pricingLowToHigh": 60, + "pricingHighToLow": 227 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { - "name": "OpenAI", - "slug": "openAi", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 100000, - "providerModelId": "o1-2024-12-17", + "name": "NextBit", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nextbit256.com/&size=256", + "slug": "nextBit", + "quantization": "bf16", + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "qwen:2-5-7b", "pricing": { - "prompt": "0.000015", - "completion": "0.00006", - "image": "0.021675", + "prompt": "0.00000004", + "completion": "0.0000001", + "image": "0", "request": "0", - "inputCacheRead": "0.0000075", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", - "seed", "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "response_format", "structured_outputs" ], - "inputCost": 15, - "outputCost": 60 - } - ] - }, - { - "slug": "x-ai/grok-2-1212", - "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-15T03:20:14.161318+00:00", - "hfUpdatedAt": null, - "name": "xAI: Grok 2 1212", - "shortName": "Grok 2 1212", - "author": "x-ai", - "description": "Grok 2 1212 introduces significant enhancements to accuracy, instruction adherence, and multilingual support, making it a powerful and flexible choice for developers seeking a highly steerable, intelligent model.", - "modelVersionGroupId": null, - "contextLength": 131072, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Grok", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "x-ai/grok-2-1212", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "9d94b368-75f7-44b1-a056-f8a1240da545", - "name": "xAI | x-ai/grok-2-1212", - "contextLength": 131072, - "model": { - "slug": "x-ai/grok-2-1212", - "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-15T03:20:14.161318+00:00", - "hfUpdatedAt": null, - "name": "xAI: Grok 2 1212", - "shortName": "Grok 2 1212", - "author": "x-ai", - "description": "Grok 2 1212 introduces significant enhancements to accuracy, instruction adherence, and multilingual support, making it a powerful and flexible choice for developers seeking a highly steerable, intelligent model.", - "modelVersionGroupId": null, - "contextLength": 131072, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Grok", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "x-ai/grok-2-1212", - "reasoningConfig": null, - "features": {} + "inputCost": 0.04, + "outputCost": 0.1, + "throughput": 7.6005, + "latency": 1819 }, - "modelVariantSlug": "x-ai/grok-2-1212", - "modelVariantPermaslug": "x-ai/grok-2-1212", - "providerName": "xAI", - "providerInfo": { - "name": "xAI", - "displayName": "xAI", - "slug": "xai", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://x.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://x.ai/legal/privacy-policy", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - } + { + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 32768, + "maxCompletionTokens": 16384, + "providerModelId": "Qwen/Qwen2.5-7B-Instruct", + "pricing": { + "prompt": "0.00000005", + "completion": "0.0000001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "xAI", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": "https://status.x.ai/", - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://x.ai/&size=256" - } + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" + ], + "inputCost": 0.05, + "outputCost": 0.1, + "throughput": 52.994, + "latency": 241 }, - "providerDisplayName": "xAI", - "providerModelId": "grok-2-1212", - "providerGroup": "xAI", - "quantization": null, - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logprobs", - "top_logprobs", - "response_format" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://x.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://x.ai/legal/privacy-policy", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + { + "name": "nCompass", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://console.ncompass.tech/&size=256", + "slug": "nCompass", + "quantization": "bf16", + "context": 32768, + "maxCompletionTokens": 32768, + "providerModelId": "Qwen/Qwen2.5-7B-Instruct", + "pricing": { + "prompt": "0.0000002", + "completion": "0.0000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 }, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "pricing": { - "prompt": "0.000002", - "completion": "0.00001", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty" + ], + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 145.8585, + "latency": 114 }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 127, - "newest": 154, - "throughputHighToLow": 80, - "latencyLowToHigh": 29, - "pricingLowToHigh": 255, - "pricingHighToLow": 64 - }, - "providers": [ { - "name": "xAI", - "slug": "xAi", - "quantization": null, - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "grok-2-1212", + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": "fp8", + "context": 32768, + "maxCompletionTokens": 2048, + "providerModelId": "Qwen/Qwen2.5-7B-Instruct-Turbo", "pricing": { - "prompt": "0.000002", - "completion": "0.00001", + "prompt": "0.0000003", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -39097,21 +41259,22 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "seed", - "logprobs", - "top_logprobs", + "top_k", + "repetition_penalty", + "logit_bias", + "min_p", "response_format" ], - "inputCost": 2, - "outputCost": 10 + "inputCost": 0.3, + "outputCost": 0.3, + "throughput": 179.537, + "latency": 203 } ] }, @@ -39274,14 +41437,16 @@ export default { "sorting": { "topWeekly": 128, "newest": 39, - "throughputHighToLow": 200, - "latencyLowToHigh": 184, + "throughputHighToLow": 192, + "latencyLowToHigh": 112, "pricingLowToHigh": 19, "pricingHighToLow": 162 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": null, "context": 32000, @@ -39310,21 +41475,23 @@ export default { "logit_bias" ], "inputCost": 0.24, - "outputCost": 0.24 + "outputCost": 0.24, + "throughput": 22.1315, + "latency": 1098 } ] }, { - "slug": "qwen/qwen-2.5-72b-instruct", - "hfSlug": "Qwen/Qwen2.5-72B-Instruct", + "slug": "mistralai/mistral-7b-instruct", + "hfSlug": "mistralai/Mistral-7B-Instruct-v0.3", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-19T00:00:00+00:00", + "createdAt": "2024-05-27T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen2.5 72B Instruct (free)", - "shortName": "Qwen2.5 72B Instruct (free)", - "author": "qwen", - "description": "Qwen2.5 72B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", - "modelVersionGroupId": null, + "name": "Mistral: Mistral 7B Instruct", + "shortName": "Mistral 7B Instruct", + "author": "mistralai", + "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*", + "modelVersionGroupId": "1d07cc56-c54d-4587-b785-5093496397a4", "contextLength": 32768, "inputModalities": [ "text" @@ -39333,36 +41500,35 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", + "group": "Mistral", + "instructType": "mistral", "defaultSystem": null, "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" + "[INST]", + "" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-2.5-72b-instruct", + "permaslug": "mistralai/mistral-7b-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "e0e08390-a346-4bbf-a641-0ccdedface29", - "name": "Chutes | qwen/qwen-2.5-72b-instruct:free", + "id": "54a41d98-d92a-4ec1-a68c-a1da232e0946", + "name": "Enfer | mistralai/mistral-7b-instruct", "contextLength": 32768, "model": { - "slug": "qwen/qwen-2.5-72b-instruct", - "hfSlug": "Qwen/Qwen2.5-72B-Instruct", + "slug": "mistralai/mistral-7b-instruct", + "hfSlug": "mistralai/Mistral-7B-Instruct-v0.3", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-19T00:00:00+00:00", + "createdAt": "2024-05-27T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen2.5 72B Instruct", - "shortName": "Qwen2.5 72B Instruct", - "author": "qwen", - "description": "Qwen2.5 72B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", - "modelVersionGroupId": null, - "contextLength": 131072, + "name": "Mistral: Mistral 7B Instruct", + "shortName": "Mistral 7B Instruct", + "author": "mistralai", + "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*", + "modelVersionGroupId": "1d07cc56-c54d-4587-b785-5093496397a4", + "contextLength": 32768, "inputModalities": [ "text" ], @@ -39370,41 +41536,40 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", + "group": "Mistral", + "instructType": "mistral", "defaultSystem": null, "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" + "[INST]", + "" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-2.5-72b-instruct", + "permaslug": "mistralai/mistral-7b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "qwen/qwen-2.5-72b-instruct:free", - "modelVariantPermaslug": "qwen/qwen-2.5-72b-instruct:free", - "providerName": "Chutes", + "modelVariantSlug": "mistralai/mistral-7b-instruct", + "modelVariantPermaslug": "mistralai/mistral-7b-instruct", + "providerName": "Enfer", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "Enfer", + "displayName": "Enfer", + "slug": "enfer", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://enfer.ai/privacy-policy", + "privacyPolicyUrl": "https://enfer.ai/privacy-policy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false } }, "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Chutes", + "group": "Enfer", "editors": [], "owners": [], "isMultipartSupported": true, @@ -39412,18 +41577,18 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://enfer.ai/&size=256" } }, - "providerDisplayName": "Chutes", - "providerModelId": "Qwen/Qwen2.5-72B-Instruct", - "providerGroup": "Chutes", - "quantization": "bf16", - "variant": "free", - "isFree": true, + "providerDisplayName": "Enfer", + "providerModelId": "mistralai/mistral-7b-instruct-v0-3", + "providerGroup": "Enfer", + "quantization": null, + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 16384, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -39433,28 +41598,22 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logprobs", "logit_bias", - "top_logprobs" + "logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://enfer.ai/privacy-policy", + "privacyPolicyUrl": "https://enfer.ai/privacy-policy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false }, - "training": true, - "retainsPrompts": true + "training": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.000000028", + "completion": "0.000000054", "image": "0", "request": "0", "webSearch": "0", @@ -39473,63 +41632,32 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 32, - "newest": 204, - "throughputHighToLow": 222, - "latencyLowToHigh": 78, - "pricingLowToHigh": 64, - "pricingHighToLow": 186 + "topWeekly": 129, + "newest": 249, + "throughputHighToLow": 36, + "latencyLowToHigh": 74, + "pricingLowToHigh": 70, + "pricingHighToLow": 234 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "fp8", + "name": "Enfer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://enfer.ai/&size=256", + "slug": "enfer", + "quantization": null, "context": 32768, "maxCompletionTokens": 16384, - "providerModelId": "Qwen/Qwen2.5-72B-Instruct", + "providerModelId": "mistralai/mistral-7b-instruct-v0-3", "pricing": { - "prompt": "0.00000012", - "completion": "0.00000039", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "inputCost": 0.12, - "outputCost": 0.39 - }, - { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen2.5-72B-Instruct", - "pricing": { - "prompt": "0.00000013", - "completion": "0.0000004", + "prompt": "0.000000028", + "completion": "0.000000054", "image": "0", "request": "0", "webSearch": "0", @@ -39543,25 +41671,25 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", "logit_bias", - "logprobs", - "top_logprobs" + "logprobs" ], - "inputCost": 0.13, - "outputCost": 0.4 + "inputCost": 0.03, + "outputCost": 0.05, + "throughput": 113.0205, + "latency": 4302 }, { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": "unknown", - "context": 32000, - "maxCompletionTokens": 4096, - "providerModelId": "qwen/qwen-2.5-72b-instruct", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 32768, + "maxCompletionTokens": 16384, + "providerModelId": "mistralai/Mistral-7B-Instruct-v0.3", "pricing": { - "prompt": "0.00000038", - "completion": "0.0000004", + "prompt": "0.000000029", + "completion": "0.000000055", "image": "0", "request": "0", "webSearch": "0", @@ -39577,25 +41705,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "min_p", "repetition_penalty", - "logit_bias" + "response_format", + "top_k", + "seed", + "min_p" ], - "inputCost": 0.38, - "outputCost": 0.4 + "inputCost": 0.03, + "outputCost": 0.06, + "throughput": 105.912, + "latency": 593.5 }, { - "name": "Hyperbolic", - "slug": "hyperbolic", + "name": "NextBit", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nextbit256.com/&size=256", + "slug": "nextBit", "quantization": "bf16", - "context": 131072, + "context": 32768, "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen2.5-72B-Instruct", + "providerModelId": "mistral:7b", "pricing": { - "prompt": "0.0000004", - "completion": "0.0000004", + "prompt": "0.000000029", + "completion": "0.000000058", "image": "0", "request": "0", "webSearch": "0", @@ -39609,27 +41740,25 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "logprobs", - "top_logprobs", - "seed", - "logit_bias", - "top_k", - "min_p", - "repetition_penalty" + "response_format", + "structured_outputs" ], - "inputCost": 0.4, - "outputCost": 0.4 + "inputCost": 0.03, + "outputCost": 0.06, + "throughput": 78.8245, + "latency": 1221 }, { - "name": "Fireworks", - "slug": "fireworks", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", "quantization": "unknown", "context": 32768, "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/qwen2p5-72b-instruct", + "providerModelId": "mistralai/mistral-7b-instruct", "pricing": { - "prompt": "0.0000009", - "completion": "0.0000009", + "prompt": "0.000000029", + "completion": "0.000000059", "image": "0", "request": "0", "webSearch": "0", @@ -39637,35 +41766,34 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", + "seed", "top_k", + "min_p", "repetition_penalty", - "response_format", - "structured_outputs", - "logit_bias", - "logprobs", - "top_logprobs" + "logit_bias" ], - "inputCost": 0.9, - "outputCost": 0.9 + "inputCost": 0.03, + "outputCost": 0.06, + "throughput": 142.876, + "latency": 768 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 2048, - "providerModelId": "Qwen/Qwen2.5-72B-Instruct-Turbo", + "quantization": "unknown", + "context": 32768, + "maxCompletionTokens": 4096, + "providerModelId": "mistralai/Mistral-7B-Instruct-v0.3", "pricing": { - "prompt": "0.0000012", - "completion": "0.0000012", + "prompt": "0.0000002", + "completion": "0.0000002", "image": "0", "request": "0", "webSearch": "0", @@ -39685,8 +41813,10 @@ export default { "min_p", "response_format" ], - "inputCost": 1.2, - "outputCost": 1.2 + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 211.92, + "latency": 283 } ] }, @@ -39865,14 +41995,16 @@ export default { "sorting": { "topWeekly": 130, "newest": 100, - "throughputHighToLow": 172, - "latencyLowToHigh": 314, + "throughputHighToLow": 167, + "latencyLowToHigh": 313, "pricingLowToHigh": 250, "pricingHighToLow": 70 }, + "authorIcon": "https://openrouter.ai/images/icons/Perplexity.svg", "providers": [ { "name": "Perplexity", + "icon": "https://openrouter.ai/images/icons/Perplexity.svg", "slug": "perplexity", "quantization": null, "context": 128000, @@ -39898,7 +42030,9 @@ export default { "presence_penalty" ], "inputCost": 2, - "outputCost": 8 + "outputCost": 8, + "throughput": 51.864, + "latency": 33768 } ] }, @@ -40064,14 +42198,16 @@ export default { "sorting": { "topWeekly": 131, "newest": 190, - "throughputHighToLow": 157, - "latencyLowToHigh": 259, + "throughputHighToLow": 165, + "latencyLowToHigh": 69, "pricingLowToHigh": 130, "pricingHighToLow": 187 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nvidia.com/\\u0026size=256", "providers": [ { "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", "slug": "lambda", "quantization": "fp8", "context": 131072, @@ -40100,10 +42236,13 @@ export default { "response_format" ], "inputCost": 0.12, - "outputCost": 0.3 + "outputCost": 0.3, + "throughput": 51.772, + "latency": 757.5 }, { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", "context": 131072, @@ -40134,10 +42273,13 @@ export default { "min_p" ], "inputCost": 0.12, - "outputCost": 0.3 + "outputCost": 0.3, + "throughput": 26.008, + "latency": 1093 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 131072, @@ -40166,10 +42308,13 @@ export default { "top_logprobs" ], "inputCost": 0.13, - "outputCost": 0.4 + "outputCost": 0.4, + "throughput": 40.398, + "latency": 785 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": null, "context": 32768, @@ -40198,10 +42343,13 @@ export default { "response_format" ], "inputCost": 0.88, - "outputCost": 0.88 + "outputCost": 0.88, + "throughput": 68.8815, + "latency": 725.5 }, { "name": "Infermatic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://infermatic.ai/&size=256", "slug": "infermatic", "quantization": "unknown", "context": 32000, @@ -40398,16 +42546,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 82, + "topWeekly": 86, "newest": 172, - "throughputHighToLow": 164, - "latencyLowToHigh": 31, + "throughputHighToLow": 159, + "latencyLowToHigh": 66, "pricingLowToHigh": 59, "pricingHighToLow": 215 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", "context": 32768, @@ -40436,10 +42586,13 @@ export default { "min_p" ], "inputCost": 0.06, - "outputCost": 0.15 + "outputCost": 0.15, + "throughput": 53.254, + "latency": 403 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 131072, @@ -40468,10 +42621,13 @@ export default { "top_logprobs" ], "inputCost": 0.06, - "outputCost": 0.18 + "outputCost": 0.18, + "throughput": 52.075, + "latency": 464 }, { "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", "slug": "lambda", "quantization": "bf16", "context": 32768, @@ -40500,10 +42656,13 @@ export default { "response_format" ], "inputCost": 0.07, - "outputCost": 0.16 + "outputCost": 0.16, + "throughput": 44.453, + "latency": 502 }, { "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", "slug": "hyperbolic", "quantization": "fp8", "context": 32768, @@ -40534,10 +42693,13 @@ export default { "repetition_penalty" ], "inputCost": 0.2, - "outputCost": 0.2 + "outputCost": 0.2, + "throughput": 53.427, + "latency": 1119 }, { "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", "slug": "parasail", "quantization": "fp8", "context": 131072, @@ -40562,10 +42724,13 @@ export default { "top_k" ], "inputCost": 0.35, - "outputCost": 0.5 + "outputCost": 0.5, + "throughput": 56.904, + "latency": 569 }, { "name": "Mancer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", "slug": "mancer", "quantization": "fp8", "context": 32768, @@ -40595,10 +42760,13 @@ export default { "top_a" ], "inputCost": 0.38, - "outputCost": 1.5 + "outputCost": 1.5, + "throughput": 31.575, + "latency": 807 }, { "name": "Mancer (private)", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", "slug": "mancer (private)", "quantization": "fp8", "context": 32768, @@ -40628,10 +42796,13 @@ export default { "top_a" ], "inputCost": 0.5, - "outputCost": 2 + "outputCost": 2, + "throughput": 31.802, + "latency": 1431 }, { "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", "slug": "cloudflare", "quantization": null, "context": 32768, @@ -40657,10 +42828,13 @@ export default { "presence_penalty" ], "inputCost": 0.66, - "outputCost": 1 + "outputCost": 1, + "throughput": 34.9445, + "latency": 1051 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "fp16", "context": 16384, @@ -40689,10 +42863,13 @@ export default { "response_format" ], "inputCost": 0.8, - "outputCost": 0.8 + "outputCost": 0.8, + "throughput": 91.708, + "latency": 808 }, { "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", "slug": "featherless", "quantization": "fp8", "context": 16384, @@ -40888,13 +43065,15 @@ export default { "topWeekly": 133, "newest": 153, "throughputHighToLow": 158, - "latencyLowToHigh": 147, + "latencyLowToHigh": 150, "pricingLowToHigh": 254, "pricingHighToLow": 63 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://x.ai/\\u0026size=256", "providers": [ { "name": "xAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://x.ai/&size=256", "slug": "xAi", "quantization": null, "context": 32768, @@ -40922,22 +43101,24 @@ export default { "response_format" ], "inputCost": 2, - "outputCost": 10 + "outputCost": 10, + "throughput": 52.9805, + "latency": 904 } ] }, { - "slug": "mistralai/pixtral-12b", - "hfSlug": "mistralai/Pixtral-12B-2409", + "slug": "anthropic/claude-3-sonnet", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-10T00:00:00+00:00", + "createdAt": "2024-03-05T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Mistral: Pixtral 12B", - "shortName": "Pixtral 12B", - "author": "mistralai", - "description": "The first multi-modal, text+image-to-text model from Mistral AI. Its weights were launched via torrent: https://x.com/mistralai/status/1833758285167722836.", - "modelVersionGroupId": null, - "contextLength": 32768, + "name": "Anthropic: Claude 3 Sonnet", + "shortName": "Claude 3 Sonnet", + "author": "anthropic", + "description": "Claude 3 Sonnet is an ideal balance of intelligence and speed for enterprise workloads. Maximum utility at a lower price, dependable, balanced for scaled deployments.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal", + "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", + "contextLength": 200000, "inputModalities": [ "text", "image" @@ -40946,32 +43127,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "Claude", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/pixtral-12b", + "permaslug": "anthropic/claude-3-sonnet", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "b550f7af-571a-45fd-b442-b3327afaf38c", - "name": "Hyperbolic | mistralai/pixtral-12b", - "contextLength": 32768, + "id": "dc39604a-e702-4a0c-8bd8-0f2578f14a2d", + "name": "Anthropic | anthropic/claude-3-sonnet", + "contextLength": 200000, "model": { - "slug": "mistralai/pixtral-12b", - "hfSlug": "mistralai/Pixtral-12B-2409", + "slug": "anthropic/claude-3-sonnet", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-10T00:00:00+00:00", + "createdAt": "2024-03-05T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Mistral: Pixtral 12B", - "shortName": "Pixtral 12B", - "author": "mistralai", - "description": "The first multi-modal, text+image-to-text model from Mistral AI. Its weights were launched via torrent: https://x.com/mistralai/status/1833758285167722836.", - "modelVersionGroupId": null, - "contextLength": 4096, + "name": "Anthropic: Claude 3 Sonnet", + "shortName": "Claude 3 Sonnet", + "author": "anthropic", + "description": "Claude 3 Sonnet is an ideal balance of intelligence and speed for enterprise workloads. Maximum utility at a lower price, dependable, balanced for scaled deployments.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal", + "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", + "contextLength": 200000, "inputModalities": [ "text", "image" @@ -40980,89 +43161,93 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "Claude", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/pixtral-12b", + "permaslug": "anthropic/claude-3-sonnet", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "mistralai/pixtral-12b", - "modelVariantPermaslug": "mistralai/pixtral-12b", - "providerName": "Hyperbolic", + "modelVariantSlug": "anthropic/claude-3-sonnet", + "modelVariantPermaslug": "anthropic/claude-3-sonnet", + "providerName": "Anthropic", "providerInfo": { - "name": "Hyperbolic", - "displayName": "Hyperbolic", - "slug": "hyperbolic", + "name": "Anthropic", + "displayName": "Anthropic", + "slug": "anthropic", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://hyperbolic.xyz/privacy", - "termsOfServiceUrl": "https://hyperbolic.xyz/terms", + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", "paidModels": { - "training": false - } + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": false, - "group": "Hyperbolic", + "moderationRequired": true, + "group": "Anthropic", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.anthropic.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256" + "url": "/images/icons/Anthropic.svg" } }, - "providerDisplayName": "Hyperbolic", - "providerModelId": "mistralai/Pixtral-12B-2409", - "providerGroup": "Hyperbolic", - "quantization": "bf16", + "providerDisplayName": "Anthropic", + "providerModelId": "claude-3-sonnet-20240229", + "providerGroup": "Anthropic", + "quantization": "unknown", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "logprobs", - "top_logprobs", - "seed", - "logit_bias", "top_k", - "min_p", - "repetition_penalty" + "stop" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "privacyPolicyUrl": "https://hyperbolic.xyz/privacy", - "termsOfServiceUrl": "https://hyperbolic.xyz/terms", + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "training": false + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", - "image": "0.0001445", + "prompt": "0.000003", + "completion": "0.000015", + "image": "0.0048", "request": "0", + "inputCacheRead": "0.0000003", + "inputCacheWrite": "0.00000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -41071,7 +43256,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -41079,66 +43264,68 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 134, - "newest": 211, - "throughputHighToLow": 115, - "latencyLowToHigh": 215, - "pricingLowToHigh": 116, - "pricingHighToLow": 204 + "newest": 281, + "throughputHighToLow": 116, + "latencyLowToHigh": 128, + "pricingLowToHigh": 278, + "pricingHighToLow": 50 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { - "name": "Hyperbolic", - "slug": "hyperbolic", - "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "mistralai/Pixtral-12B-2409", + "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", + "slug": "anthropic", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 4096, + "providerModelId": "claude-3-sonnet-20240229", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", - "image": "0.0001445", + "prompt": "0.000003", + "completion": "0.000015", + "image": "0.0048", "request": "0", + "inputCacheRead": "0.0000003", + "inputCacheWrite": "0.00000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "logprobs", - "top_logprobs", - "seed", - "logit_bias", "top_k", - "min_p", - "repetition_penalty" + "stop" ], - "inputCost": 0.1, - "outputCost": 0.1 + "inputCost": 3, + "outputCost": 15, + "throughput": 66.9785, + "latency": 754 }, { - "name": "Mistral", - "slug": "mistral", + "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", + "slug": "vertex", "quantization": "unknown", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "pixtral-12b-2409", + "context": 200000, + "maxCompletionTokens": 4096, + "providerModelId": "claude-3-sonnet@20240229", "pricing": { - "prompt": "0.00000015", - "completion": "0.00000015", - "image": "0.0002168", + "prompt": "0.000003", + "completion": "0.000015", + "image": "0.0048", "request": "0", + "inputCacheRead": "0.0000003", + "inputCacheWrite": "0.00000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -41149,29 +43336,27 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "structured_outputs", - "seed" + "top_k", + "stop" ], - "inputCost": 0.15, - "outputCost": 0.15 + "inputCost": 3, + "outputCost": 15, + "throughput": 68.034, + "latency": 1041 } ] }, { - "slug": "anthropic/claude-3-sonnet", + "slug": "anthropic/claude-3-opus", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-03-05T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3 Sonnet", - "shortName": "Claude 3 Sonnet", + "name": "Anthropic: Claude 3 Opus", + "shortName": "Claude 3 Opus", "author": "anthropic", - "description": "Claude 3 Sonnet is an ideal balance of intelligence and speed for enterprise workloads. Maximum utility at a lower price, dependable, balanced for scaled deployments.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal", - "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", + "description": "Claude 3 Opus is Anthropic's most powerful model for highly complex tasks. It boasts top-level performance, intelligence, fluency, and understanding.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal", + "modelVersionGroupId": null, "contextLength": 200000, "inputModalities": [ "text", @@ -41188,24 +43373,24 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-sonnet", + "permaslug": "anthropic/claude-3-opus", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "dc39604a-e702-4a0c-8bd8-0f2578f14a2d", - "name": "Anthropic | anthropic/claude-3-sonnet", + "id": "ee74a4e0-2863-4f51-99a5-997c31c48ae7", + "name": "Anthropic | anthropic/claude-3-opus", "contextLength": 200000, "model": { - "slug": "anthropic/claude-3-sonnet", + "slug": "anthropic/claude-3-opus", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-03-05T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3 Sonnet", - "shortName": "Claude 3 Sonnet", + "name": "Anthropic: Claude 3 Opus", + "shortName": "Claude 3 Opus", "author": "anthropic", - "description": "Claude 3 Sonnet is an ideal balance of intelligence and speed for enterprise workloads. Maximum utility at a lower price, dependable, balanced for scaled deployments.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal", - "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", + "description": "Claude 3 Opus is Anthropic's most powerful model for highly complex tasks. It boasts top-level performance, intelligence, fluency, and understanding.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal", + "modelVersionGroupId": null, "contextLength": 200000, "inputModalities": [ "text", @@ -41222,12 +43407,12 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-sonnet", + "permaslug": "anthropic/claude-3-opus", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "anthropic/claude-3-sonnet", - "modelVariantPermaslug": "anthropic/claude-3-sonnet", + "modelVariantSlug": "anthropic/claude-3-opus", + "modelVariantPermaslug": "anthropic/claude-3-opus", "providerName": "Anthropic", "providerInfo": { "name": "Anthropic", @@ -41261,7 +43446,7 @@ export default { } }, "providerDisplayName": "Anthropic", - "providerModelId": "claude-3-sonnet-20240229", + "providerModelId": "claude-3-opus-20240229", "providerGroup": "Anthropic", "quantization": "unknown", "variant": "standard", @@ -41296,12 +43481,12 @@ export default { "retentionDays": 30 }, "pricing": { - "prompt": "0.000003", - "completion": "0.000015", - "image": "0.0048", + "prompt": "0.000015", + "completion": "0.000075", + "image": "0.024", "request": "0", - "inputCacheRead": "0.0000003", - "inputCacheWrite": "0.00000375", + "inputCacheRead": "0.0000015", + "inputCacheWrite": "0.00001875", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -41325,26 +43510,28 @@ export default { "sorting": { "topWeekly": 135, "newest": 279, - "throughputHighToLow": 121, - "latencyLowToHigh": 128, - "pricingLowToHigh": 278, - "pricingHighToLow": 50 + "throughputHighToLow": 247, + "latencyLowToHigh": 232, + "pricingLowToHigh": 309, + "pricingHighToLow": 8 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", "slug": "anthropic", "quantization": "unknown", "context": 200000, "maxCompletionTokens": 4096, - "providerModelId": "claude-3-sonnet-20240229", + "providerModelId": "claude-3-opus-20240229", "pricing": { - "prompt": "0.000003", - "completion": "0.000015", - "image": "0.0048", + "prompt": "0.000015", + "completion": "0.000075", + "image": "0.024", "request": "0", - "inputCacheRead": "0.0000003", - "inputCacheWrite": "0.00000375", + "inputCacheRead": "0.0000015", + "inputCacheWrite": "0.00001875", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -41358,23 +43545,26 @@ export default { "top_k", "stop" ], - "inputCost": 3, - "outputCost": 15 + "inputCost": 15, + "outputCost": 75, + "throughput": 30.035, + "latency": 1648.5 }, { "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", "slug": "vertex", "quantization": "unknown", "context": 200000, "maxCompletionTokens": 4096, - "providerModelId": "claude-3-sonnet@20240229", + "providerModelId": "claude-3-opus@20240229", "pricing": { - "prompt": "0.000003", - "completion": "0.000015", - "image": "0.0048", + "prompt": "0.000015", + "completion": "0.000075", + "image": "0.024", "request": "0", - "inputCacheRead": "0.0000003", - "inputCacheWrite": "0.00000375", + "inputCacheRead": "0.0000015", + "inputCacheWrite": "0.00001875", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -41388,231 +43578,10 @@ export default { "top_k", "stop" ], - "inputCost": 3, - "outputCost": 15 - } - ] - }, - { - "slug": "deepseek/deepseek-r1-distill-llama-8b", - "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Llama-8B", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-02-07T14:15:18.18663+00:00", - "hfUpdatedAt": null, - "name": "DeepSeek: R1 Distill Llama 8B", - "shortName": "R1 Distill Llama 8B", - "author": "deepseek", - "description": "DeepSeek R1 Distill Llama 8B is a distilled large language model based on [Llama-3.1-8B-Instruct](/meta-llama/llama-3.1-8b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 50.4\n- MATH-500 pass@1: 89.1\n- CodeForces Rating: 1205\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.\n\nHugging Face: \n- [Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) \n- [DeepSeek-R1-Distill-Llama-8B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B) |", - "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", - "contextLength": 32000, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Llama3", - "instructType": "deepseek-r1", - "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "deepseek/deepseek-r1-distill-llama-8b", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, - "endpoint": { - "id": "fbc63cef-f2e8-4e4b-a135-9b45485e0eaf", - "name": "Novita | deepseek/deepseek-r1-distill-llama-8b", - "contextLength": 32000, - "model": { - "slug": "deepseek/deepseek-r1-distill-llama-8b", - "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Llama-8B", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-02-07T14:15:18.18663+00:00", - "hfUpdatedAt": null, - "name": "DeepSeek: R1 Distill Llama 8B", - "shortName": "R1 Distill Llama 8B", - "author": "deepseek", - "description": "DeepSeek R1 Distill Llama 8B is a distilled large language model based on [Llama-3.1-8B-Instruct](/meta-llama/llama-3.1-8b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 50.4\n- MATH-500 pass@1: 89.1\n- CodeForces Rating: 1205\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.\n\nHugging Face: \n- [Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) \n- [DeepSeek-R1-Distill-Llama-8B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B) |", - "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", - "contextLength": 0, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Llama3", - "instructType": "deepseek-r1", - "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "deepseek/deepseek-r1-distill-llama-8b", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } - }, - "modelVariantSlug": "deepseek/deepseek-r1-distill-llama-8b", - "modelVariantPermaslug": "deepseek/deepseek-r1-distill-llama-8b", - "providerName": "Novita", - "providerInfo": { - "name": "Novita", - "displayName": "NovitaAI", - "slug": "novita", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", - "paidModels": { - "training": false - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "Novita", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": "https://status.novita.ai/", - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", - "invertRequired": true - } - }, - "providerDisplayName": "NovitaAI", - "providerModelId": "deepseek/deepseek-r1-distill-llama-8b", - "providerGroup": "Novita", - "quantization": "unknown", - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 32000, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", - "paidModels": { - "training": false - }, - "training": false - }, - "pricing": { - "prompt": "0.00000004", - "completion": "0.00000004", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": true, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 136, - "newest": 117, - "throughputHighToLow": 186, - "latencyLowToHigh": 241, - "pricingLowToHigh": 87, - "pricingHighToLow": 230 - }, - "providers": [ - { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": "unknown", - "context": 32000, - "maxCompletionTokens": 32000, - "providerModelId": "deepseek/deepseek-r1-distill-llama-8b", - "pricing": { - "prompt": "0.00000004", - "completion": "0.00000004", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" - ], - "inputCost": 0.04, - "outputCost": 0.04 + "inputCost": 15, + "outputCost": 75, + "throughput": 17.365, + "latency": 3104 } ] }, @@ -41770,16 +43739,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 137, + "topWeekly": 136, "newest": 95, - "throughputHighToLow": 261, - "latencyLowToHigh": 158, + "throughputHighToLow": 257, + "latencyLowToHigh": 165, "pricingLowToHigh": 203, - "pricingHighToLow": 114 + "pricingHighToLow": 115 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", "slug": "parasail", "quantization": "fp8", "context": 131072, @@ -41804,151 +43775,147 @@ export default { "top_k" ], "inputCost": 0.8, - "outputCost": 1 + "outputCost": 1, + "throughput": 25.403, + "latency": 1019 } ] }, { - "slug": "anthropic/claude-3-opus", - "hfSlug": null, + "slug": "meta-llama/llama-3.3-70b-instruct", + "hfSlug": "meta-llama/Llama-3.3-70B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-03-05T00:00:00+00:00", + "createdAt": "2024-12-06T17:28:57.828422+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3 Opus", - "shortName": "Claude 3 Opus", - "author": "anthropic", - "description": "Claude 3 Opus is Anthropic's most powerful model for highly complex tasks. It boasts top-level performance, intelligence, fluency, and understanding.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal", - "modelVersionGroupId": null, - "contextLength": 200000, + "name": "Meta: Llama 3.3 70B Instruct (free)", + "shortName": "Llama 3.3 70B Instruct (free)", + "author": "meta-llama", + "description": "The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.\n\nSupported languages: English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.\n\n[Model Card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/MODEL_CARD.md)", + "modelVersionGroupId": "397604e2-45fa-454e-a85d-9921f5138747", + "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", - "instructType": null, + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-opus", + "permaslug": "meta-llama/llama-3.3-70b-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "ee74a4e0-2863-4f51-99a5-997c31c48ae7", - "name": "Anthropic | anthropic/claude-3-opus", - "contextLength": 200000, + "id": "4625027b-9188-4d61-b21a-ae6808bd4e5d", + "name": "Crusoe | meta-llama/llama-3.3-70b-instruct:free", + "contextLength": 131072, "model": { - "slug": "anthropic/claude-3-opus", - "hfSlug": null, + "slug": "meta-llama/llama-3.3-70b-instruct", + "hfSlug": "meta-llama/Llama-3.3-70B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-03-05T00:00:00+00:00", + "createdAt": "2024-12-06T17:28:57.828422+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3 Opus", - "shortName": "Claude 3 Opus", - "author": "anthropic", - "description": "Claude 3 Opus is Anthropic's most powerful model for highly complex tasks. It boasts top-level performance, intelligence, fluency, and understanding.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal", - "modelVersionGroupId": null, - "contextLength": 200000, + "name": "Meta: Llama 3.3 70B Instruct", + "shortName": "Llama 3.3 70B Instruct", + "author": "meta-llama", + "description": "The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.\n\nSupported languages: English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.\n\n[Model Card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/MODEL_CARD.md)", + "modelVersionGroupId": "397604e2-45fa-454e-a85d-9921f5138747", + "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", - "instructType": null, + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-opus", + "permaslug": "meta-llama/llama-3.3-70b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "anthropic/claude-3-opus", - "modelVariantPermaslug": "anthropic/claude-3-opus", - "providerName": "Anthropic", + "modelVariantSlug": "meta-llama/llama-3.3-70b-instruct:free", + "modelVariantPermaslug": "meta-llama/llama-3.3-70b-instruct:free", + "providerName": "Crusoe", "providerInfo": { - "name": "Anthropic", - "displayName": "Anthropic", - "slug": "anthropic", + "name": "Crusoe", + "displayName": "Crusoe", + "slug": "crusoe", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "privacyPolicyUrl": "https://legal.crusoe.ai/open-router#cloud-privacy-policy", + "termsOfServiceUrl": "https://legal.crusoe.ai/open-router#managed-inference-tos-open-router", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "training": false + } }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "Anthropic", + "moderationRequired": false, + "group": "Crusoe", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.anthropic.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Anthropic.svg" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://crusoe.ai/&size=256" } }, - "providerDisplayName": "Anthropic", - "providerModelId": "claude-3-opus-20240229", - "providerGroup": "Anthropic", - "quantization": "unknown", - "variant": "standard", - "isFree": false, + "providerDisplayName": "Crusoe", + "providerModelId": "meta-llama/Llama-3.3-70B-Instruct", + "providerGroup": "Crusoe", + "quantization": "bf16", + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": 2048, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "top_k", - "stop" + "stop", + "frequency_penalty", + "presence_penalty", + "seed" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "privacyPolicyUrl": "https://legal.crusoe.ai/open-router#cloud-privacy-policy", + "termsOfServiceUrl": "https://legal.crusoe.ai/open-router#managed-inference-tos-open-router", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, "pricing": { - "prompt": "0.000015", - "completion": "0.000075", - "image": "0.024", + "prompt": "0", + "completion": "0", + "image": "0", "request": "0", - "inputCacheRead": "0.0000015", - "inputCacheWrite": "0.00001875", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -41957,7 +43924,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -41970,58 +43937,65 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 138, - "newest": 281, - "throughputHighToLow": 249, - "latencyLowToHigh": 216, - "pricingLowToHigh": 309, - "pricingHighToLow": 8 + "topWeekly": 8, + "newest": 157, + "throughputHighToLow": 235, + "latencyLowToHigh": 88, + "pricingLowToHigh": 56, + "pricingHighToLow": 211 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { - "name": "Anthropic", - "slug": "anthropic", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 4096, - "providerModelId": "claude-3-opus-20240229", + "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", + "slug": "klusterAi", + "quantization": "fp8", + "context": 131000, + "maxCompletionTokens": 131000, + "providerModelId": "klusterai/Meta-Llama-3.3-70B-Instruct-Turbo", "pricing": { - "prompt": "0.000015", - "completion": "0.000075", - "image": "0.024", + "prompt": "0.00000007", + "completion": "0.00000033", + "image": "0", "request": "0", - "inputCacheRead": "0.0000015", - "inputCacheWrite": "0.00001875", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "top_k", - "stop" + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", + "seed" ], - "inputCost": 15, - "outputCost": 75 + "inputCost": 0.07, + "outputCost": 0.33, + "throughput": 33.095, + "latency": 571.5 }, { - "name": "Google Vertex", - "slug": "vertex", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 4096, - "providerModelId": "claude-3-opus@20240229", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/Llama-3.3-70B-Instruct-Turbo", "pricing": { - "prompt": "0.000015", - "completion": "0.000075", - "image": "0.024", + "prompt": "0.00000008", + "completion": "0.00000025", + "image": "0", "request": "0", - "inputCacheRead": "0.0000015", - "inputCacheWrite": "0.00001875", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -42032,192 +44006,66 @@ export default { "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", "top_k", - "stop" + "seed", + "min_p" ], - "inputCost": 15, - "outputCost": 75 - } - ] - }, - { - "slug": "mistralai/ministral-8b", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-10-17T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Mistral: Ministral 8B", - "shortName": "Ministral 8B", - "author": "mistralai", - "description": "Ministral 8B is an 8B parameter model featuring a unique interleaved sliding-window attention pattern for faster, memory-efficient inference. Designed for edge use cases, it supports up to 128k context length and excels in knowledge and reasoning tasks. It outperforms peers in the sub-10B category, making it perfect for low-latency, privacy-first applications.", - "modelVersionGroupId": null, - "contextLength": 128000, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Mistral", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "mistralai/ministral-8b", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "c70696c7-73c0-47c3-96b4-20c44b17101b", - "name": "Mistral | mistralai/ministral-8b", - "contextLength": 128000, - "model": { - "slug": "mistralai/ministral-8b", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-10-17T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Mistral: Ministral 8B", - "shortName": "Ministral 8B", - "author": "mistralai", - "description": "Ministral 8B is an 8B parameter model featuring a unique interleaved sliding-window attention pattern for faster, memory-efficient inference. Designed for edge use cases, it supports up to 128k context length and excels in knowledge and reasoning tasks. It outperforms peers in the sub-10B category, making it perfect for low-latency, privacy-first applications.", - "modelVersionGroupId": null, - "contextLength": 128000, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Mistral", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "mistralai/ministral-8b", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "mistralai/ministral-8b", - "modelVariantPermaslug": "mistralai/ministral-8b", - "providerName": "Mistral", - "providerInfo": { - "name": "Mistral", - "displayName": "Mistral", - "slug": "mistral", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - } - }, - "headquarters": "FR", - "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, - "moderationRequired": false, - "group": "Mistral", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/Mistral.png" - } + "inputCost": 0.08, + "outputCost": 0.25, + "throughput": 34.0525, + "latency": 268 }, - "providerDisplayName": "Mistral", - "providerModelId": "ministral-8b-2410", - "providerGroup": "Mistral", - "quantization": "unknown", - "variant": "standard", - "isFree": false, - "canAbort": false, - "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "structured_outputs", - "seed" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + { + "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", + "slug": "lambda", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "llama3.3-70b-instruct-fp8", + "pricing": { + "prompt": "0.00000012", + "completion": "0.0000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 }, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": false, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" + ], + "inputCost": 0.12, + "outputCost": 0.3, + "throughput": 60.483, + "latency": 510 }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 139, - "newest": 187, - "throughputHighToLow": 56, - "latencyLowToHigh": 2, - "pricingLowToHigh": 115, - "pricingHighToLow": 203 - }, - "providers": [ { - "name": "Mistral", - "slug": "mistral", - "quantization": "unknown", - "context": 128000, + "name": "Phala", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://phala.network/&size=256", + "slug": "phala", + "quantization": null, + "context": 131072, "maxCompletionTokens": null, - "providerModelId": "ministral-8b-2410", + "providerModelId": "phala/llama-3.3-70b-instruct", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", + "prompt": "0.00000012", + "completion": "0.00000035", "image": "0", "request": "0", "webSearch": "0", @@ -42233,185 +44081,64 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "response_format", - "structured_outputs", - "seed" - ], - "inputCost": 0.1, - "outputCost": 0.1 - } - ] - }, - { - "slug": "qwen/qwen-2.5-7b-instruct", - "hfSlug": "Qwen/Qwen2.5-7B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-10-16T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Qwen2.5 7B Instruct (free)", - "shortName": "Qwen2.5 7B Instruct (free)", - "author": "qwen", - "description": "Qwen2.5 7B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", - "modelVersionGroupId": null, - "contextLength": 32768, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", - "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "qwen/qwen-2.5-7b-instruct", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "8a40a3b8-80d8-4f75-bb8a-c0c710b31eb5", - "name": "Nineteen | qwen/qwen-2.5-7b-instruct:free", - "contextLength": 32768, - "model": { - "slug": "qwen/qwen-2.5-7b-instruct", - "hfSlug": "Qwen/Qwen2.5-7B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-10-16T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Qwen2.5 7B Instruct", - "shortName": "Qwen2.5 7B Instruct", - "author": "qwen", - "description": "Qwen2.5 7B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:\n\n- Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.\n\n- Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.\n\n- Long-context Support up to 128K tokens and can generate up to 8K tokens.\n\n- Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", - "modelVersionGroupId": null, - "contextLength": 131072, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", - "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" + "seed", + "top_k", + "min_p", + "repetition_penalty" ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "qwen/qwen-2.5-7b-instruct", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "qwen/qwen-2.5-7b-instruct:free", - "modelVariantPermaslug": "qwen/qwen-2.5-7b-instruct:free", - "providerName": "Nineteen", - "providerInfo": { - "name": "Nineteen", - "displayName": "Nineteen", - "slug": "nineteen", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://nineteen.ai/tos", - "paidModels": { - "training": false - } - }, - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "Nineteen", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nineteen.ai/&size=256" - } + "inputCost": 0.12, + "outputCost": 0.35, + "throughput": 29.689, + "latency": 749 }, - "providerDisplayName": "Nineteen", - "providerModelId": "Qwen/Qwen2.5-7B-Instruct", - "providerGroup": "Nineteen", - "quantization": "bf16", - "variant": "free", - "isFree": true, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 32768, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://nineteen.ai/tos", - "paidModels": { - "training": false + { + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/llama-3.3-70b-instruct", + "pricing": { + "prompt": "0.00000013", + "completion": "0.00000039", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 }, - "training": false - }, - "pricing": { - "prompt": "0", - "completion": "0", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" + ], + "inputCost": 0.13, + "outputCost": 0.39, + "throughput": 76.6895, + "latency": 783 }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 23, - "newest": 188, - "throughputHighToLow": 11, - "latencyLowToHigh": 18, - "pricingLowToHigh": 60, - "pricingHighToLow": 222 - }, - "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": 16384, - "providerModelId": "Qwen/Qwen2.5-7B-Instruct", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Llama-3.3-70B-Instruct", "pricing": { - "prompt": "0.00000005", - "completion": "0.0000001", + "prompt": "0.00000013", + "completion": "0.0000004", "image": "0", "request": "0", "webSearch": "0", @@ -42425,25 +44152,91 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.13, + "outputCost": 0.4, + "throughput": 38.8, + "latency": 660 + }, + { + "name": "Parasail", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.parasail.io/&size=256", + "slug": "parasail", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "parasail-llama-33-70b-fp8", + "pricing": { + "prompt": "0.00000028", + "completion": "0.00000078", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "presence_penalty", + "frequency_penalty", "repetition_penalty", - "response_format", + "top_k" + ], + "inputCost": 0.28, + "outputCost": 0.78, + "throughput": 81.678, + "latency": 542 + }, + { + "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", + "slug": "cloudflare", + "quantization": "fp8", + "context": 24000, + "maxCompletionTokens": null, + "providerModelId": "@cf/meta/llama-3.3-70b-instruct-fp8-fast", + "pricing": { + "prompt": "0.00000029", + "completion": "0.00000225", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", "top_k", "seed", - "min_p" + "repetition_penalty", + "frequency_penalty", + "presence_penalty" ], - "inputCost": 0.05, - "outputCost": 0.1 + "inputCost": 0.29, + "outputCost": 2.25, + "throughput": 35.689, + "latency": 658 }, { - "name": "nCompass", - "slug": "nCompass", + "name": "CentML", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://centml.ai/&size=256", + "slug": "centMl", "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": 32768, - "providerModelId": "Qwen/Qwen2.5-7B-Instruct", + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "meta-llama/Llama-3.3-70B-Instruct", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", + "prompt": "0.00000035", + "completion": "0.00000035", "image": "0", "request": "0", "webSearch": "0", @@ -42454,6 +44247,8 @@ export default { "max_tokens", "temperature", "top_p", + "structured_outputs", + "response_format", "stop", "frequency_penalty", "presence_penalty", @@ -42462,19 +44257,22 @@ export default { "min_p", "repetition_penalty" ], - "inputCost": 0.2, - "outputCost": 0.2 + "inputCost": 0.35, + "outputCost": 0.35, + "throughput": 70.0325, + "latency": 734.5 }, { - "name": "Together", - "slug": "together", + "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", + "slug": "hyperbolic", "quantization": "fp8", - "context": 32768, - "maxCompletionTokens": 2048, - "providerModelId": "Qwen/Qwen2.5-7B-Instruct-Turbo", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Llama-3.3-70B-Instruct", "pricing": { - "prompt": "0.0000003", - "completion": "0.0000003", + "prompt": "0.0000004", + "completion": "0.0000004", "image": "0", "request": "0", "webSearch": "0", @@ -42488,203 +44286,233 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "top_k", - "repetition_penalty", + "logprobs", + "top_logprobs", + "seed", "logit_bias", + "top_k", "min_p", - "response_format" + "repetition_penalty" ], - "inputCost": 0.3, - "outputCost": 0.3 - } - ] - }, - { - "slug": "undi95/remm-slerp-l2-13b", - "hfSlug": "Undi95/ReMM-SLERP-L2-13B", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-07-22T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "ReMM SLERP 13B", - "shortName": "ReMM SLERP 13B", - "author": "undi95", - "description": "A recreation trial of the original MythoMax-L2-B13 but with updated models. #merge", - "modelVersionGroupId": null, - "contextLength": 6144, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Llama2", - "instructType": "alpaca", - "defaultSystem": null, - "defaultStops": [ - "###", - "" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "undi95/remm-slerp-l2-13b", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "989a7c1b-f4c8-4d78-b0e3-caf0adb687c5", - "name": "Mancer | undi95/remm-slerp-l2-13b", - "contextLength": 6144, - "model": { - "slug": "undi95/remm-slerp-l2-13b", - "hfSlug": "Undi95/ReMM-SLERP-L2-13B", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-07-22T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "ReMM SLERP 13B", - "shortName": "ReMM SLERP 13B", - "author": "undi95", - "description": "A recreation trial of the original MythoMax-L2-B13 but with updated models. #merge", - "modelVersionGroupId": null, - "contextLength": 4096, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" + "inputCost": 0.4, + "outputCost": 0.4, + "throughput": 36.439, + "latency": 1267 + }, + { + "name": "Atoma", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://atoma.network/&size=256", + "slug": "atoma", + "quantization": "fp8", + "context": 104962, + "maxCompletionTokens": 100000, + "providerModelId": "Infermatic/Llama-3.3-70B-Instruct-FP8-Dynamic", + "pricing": { + "prompt": "0.0000004", + "completion": "0.0000004", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p" ], - "hasTextOutput": true, - "group": "Llama2", - "instructType": "alpaca", - "defaultSystem": null, - "defaultStops": [ - "###", - "" + "inputCost": 0.4, + "outputCost": 0.4, + "throughput": 28.218, + "latency": 936 + }, + { + "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", + "slug": "groq", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 32768, + "providerModelId": "llama-3.3-70b-versatile", + "pricing": { + "prompt": "0.00000059", + "completion": "0.00000079", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "top_logprobs", + "logprobs", + "logit_bias", + "seed" ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "undi95/remm-slerp-l2-13b", - "reasoningConfig": null, - "features": {} + "inputCost": 0.59, + "outputCost": 0.79, + "throughput": 343.6785, + "latency": 463 }, - "modelVariantSlug": "undi95/remm-slerp-l2-13b", - "modelVariantPermaslug": "undi95/remm-slerp-l2-13b", - "providerName": "Mancer", - "providerInfo": { - "name": "Mancer", - "displayName": "Mancer", - "slug": "mancer", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://mancer.tech/terms", - "privacyPolicyUrl": "https://mancer.tech/privacy", - "paidModels": { - "training": true, - "retainsPrompts": true - } + { + "name": "Friendli", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://friendli.ai/&size=256", + "slug": "friendli", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "meta-llama-3.3-70b-instruct", + "pricing": { + "prompt": "0.0000006", + "completion": "0.0000006", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 }, - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "Mancer", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256" - } + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "response_format", + "structured_outputs" + ], + "inputCost": 0.6, + "outputCost": 0.6, + "throughput": 95.6525, + "latency": 818 }, - "providerDisplayName": "Mancer", - "providerModelId": "remm-slerp", - "providerGroup": "Mancer", - "quantization": "unknown", - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 1024, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "logit_bias", - "top_k", - "min_p", - "seed", - "top_a" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://mancer.tech/terms", - "privacyPolicyUrl": "https://mancer.tech/privacy", - "paidModels": { - "training": true, - "retainsPrompts": true + { + "name": "NextBit", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nextbit256.com/&size=256", + "slug": "nextBit", + "quantization": "fp8", + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "llama3.3:70b", + "pricing": { + "prompt": "0.0000006", + "completion": "0.00000075", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 }, - "training": true, - "retainsPrompts": true + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "structured_outputs" + ], + "inputCost": 0.6, + "outputCost": 0.75, + "throughput": 32.64, + "latency": 2383 }, - "pricing": { - "prompt": "0.0000005625", - "completion": "0.000001125", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0.25 + { + "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", + "slug": "sambaNova", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": 3072, + "providerModelId": "Meta-Llama-3.3-70B-Instruct", + "pricing": { + "prompt": "0.0000006", + "completion": "0.0000012", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 0.6, + "outputCost": 1.2, + "throughput": 309.927, + "latency": 2159.5 }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null + { + "name": "Cerebras", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cerebras.ai/&size=256", + "slug": "cerebras", + "quantization": "fp16", + "context": 32000, + "maxCompletionTokens": 32000, + "providerModelId": "llama-3.3-70b", + "pricing": { + "prompt": "0.00000085", + "completion": "0.0000012", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "structured_outputs", + "response_format", + "stop", + "seed", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.85, + "outputCost": 1.2, + "throughput": 4775.985, + "latency": 279 }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 141, - "newest": 312, - "throughputHighToLow": 201, - "latencyLowToHigh": 144, - "pricingLowToHigh": 193, - "pricingHighToLow": 126 - }, - "providers": [ { - "name": "Mancer", - "slug": "mancer", - "quantization": "unknown", - "context": 6144, - "maxCompletionTokens": 1024, - "providerModelId": "remm-slerp", + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": 2048, + "providerModelId": "meta-llama/Llama-3.3-70B-Instruct-Turbo", "pricing": { - "prompt": "0.0000005625", - "completion": "0.000001125", + "prompt": "0.00000088", + "completion": "0.00000088", "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", - "discount": 0.25 + "discount": 0 }, "supportedParameters": [ "max_tokens", @@ -42693,26 +44521,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "top_k", "repetition_penalty", "logit_bias", - "top_k", "min_p", - "seed", - "top_a" + "response_format" ], - "inputCost": 0.56, - "outputCost": 1.13 + "inputCost": 0.88, + "outputCost": 0.88, + "throughput": 94.488, + "latency": 559 }, { - "name": "Mancer (private)", - "slug": "mancer (private)", - "quantization": "unknown", - "context": 6144, - "maxCompletionTokens": 1024, - "providerModelId": "remm-slerp", + "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", + "slug": "fireworks", + "quantization": "fp16", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "accounts/fireworks/models/llama-v3p3-70b-instruct", "pricing": { - "prompt": "0.00000075", - "completion": "0.0000015", + "prompt": "0.0000009", + "completion": "0.0000009", "image": "0", "request": "0", "webSearch": "0", @@ -42726,26 +44556,30 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "top_k", "repetition_penalty", + "response_format", + "structured_outputs", "logit_bias", - "top_k", - "min_p", - "seed", - "top_a" + "logprobs", + "top_logprobs" ], - "inputCost": 0.75, - "outputCost": 1.5 + "inputCost": 0.9, + "outputCost": 0.9, + "throughput": 109.5785, + "latency": 3019.5 }, { - "name": "Featherless", - "slug": "featherless", + "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", + "slug": "inferenceNet", "quantization": "fp8", - "context": 4096, - "maxCompletionTokens": 4096, - "providerModelId": "Undi95/ReMM-SLERP-L2-13B", + "context": 128000, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/llama-3.3-70b-instruct/fp-8", "pricing": { - "prompt": "0.0000008", - "completion": "0.0000012", + "prompt": "0.0000001", + "completion": "0.00000025", "image": "0", "request": "0", "webSearch": "0", @@ -42756,112 +44590,110 @@ export default { "max_tokens", "temperature", "top_p", + "structured_outputs", + "response_format", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", + "seed", "top_k", "min_p", - "seed" + "logit_bias", + "top_logprobs" ], - "inputCost": 0.8, - "outputCost": 1.2 + "inputCost": 0.1, + "outputCost": 0.25, + "throughput": 13.164, + "latency": 1625 } ] }, { - "slug": "meta-llama/llama-3.2-90b-vision-instruct", - "hfSlug": "meta-llama/Llama-3.2-90B-Vision-Instruct", + "slug": "mistralai/ministral-8b", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-25T00:00:00+00:00", + "createdAt": "2024-10-17T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.2 90B Vision Instruct", - "shortName": "Llama 3.2 90B Vision Instruct", - "author": "meta-llama", - "description": "The Llama 90B Vision model is a top-tier, 90-billion-parameter multimodal model designed for the most challenging visual reasoning and language tasks. It offers unparalleled accuracy in image captioning, visual question answering, and advanced image-text comprehension. Pre-trained on vast multimodal datasets and fine-tuned with human feedback, the Llama 90B Vision is engineered to handle the most demanding image-based AI tasks.\n\nThis model is perfect for industries requiring cutting-edge multimodal AI capabilities, particularly those dealing with complex, real-time visual and textual analysis.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "name": "Mistral: Ministral 8B", + "shortName": "Ministral 8B", + "author": "mistralai", + "description": "Ministral 8B is an 8B parameter model featuring a unique interleaved sliding-window attention pattern for faster, memory-efficient inference. Designed for edge use cases, it supports up to 128k context length and excels in knowledge and reasoning tasks. It outperforms peers in the sub-10B category, making it perfect for low-latency, privacy-first applications.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Mistral", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.2-90b-vision-instruct", + "permaslug": "mistralai/ministral-8b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "6fa97646-3631-4446-beaa-8a24dd07ddb1", - "name": "Together | meta-llama/llama-3.2-90b-vision-instruct", - "contextLength": 131072, + "id": "c70696c7-73c0-47c3-96b4-20c44b17101b", + "name": "Mistral | mistralai/ministral-8b", + "contextLength": 128000, "model": { - "slug": "meta-llama/llama-3.2-90b-vision-instruct", - "hfSlug": "meta-llama/Llama-3.2-90B-Vision-Instruct", + "slug": "mistralai/ministral-8b", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-25T00:00:00+00:00", + "createdAt": "2024-10-17T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.2 90B Vision Instruct", - "shortName": "Llama 3.2 90B Vision Instruct", - "author": "meta-llama", - "description": "The Llama 90B Vision model is a top-tier, 90-billion-parameter multimodal model designed for the most challenging visual reasoning and language tasks. It offers unparalleled accuracy in image captioning, visual question answering, and advanced image-text comprehension. Pre-trained on vast multimodal datasets and fine-tuned with human feedback, the Llama 90B Vision is engineered to handle the most demanding image-based AI tasks.\n\nThis model is perfect for industries requiring cutting-edge multimodal AI capabilities, particularly those dealing with complex, real-time visual and textual analysis.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "name": "Mistral: Ministral 8B", + "shortName": "Ministral 8B", + "author": "mistralai", + "description": "Ministral 8B is an 8B parameter model featuring a unique interleaved sliding-window attention pattern for faster, memory-efficient inference. Designed for edge use cases, it supports up to 128k context length and excels in knowledge and reasoning tasks. It outperforms peers in the sub-10B category, making it perfect for low-latency, privacy-first applications.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Mistral", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.2-90b-vision-instruct", + "permaslug": "mistralai/ministral-8b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "meta-llama/llama-3.2-90b-vision-instruct", - "modelVariantPermaslug": "meta-llama/llama-3.2-90b-vision-instruct", - "providerName": "Together", + "modelVariantSlug": "mistralai/ministral-8b", + "modelVariantPermaslug": "mistralai/ministral-8b", + "providerName": "Mistral", "providerInfo": { - "name": "Together", - "displayName": "Together", - "slug": "together", + "name": "Mistral", + "displayName": "Mistral", + "slug": "mistral", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 } }, - "headquarters": "US", + "headquarters": "FR", "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, + "hasCompletions": false, + "isAbortable": false, "moderationRequired": false, - "group": "Together", + "group": "Mistral", "editors": [], "owners": [], "isMultipartSupported": true, @@ -42869,49 +44701,51 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" + "url": "/images/icons/Mistral.png" } }, - "providerDisplayName": "Together", - "providerModelId": "meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo", - "providerGroup": "Together", - "quantization": "fp8", + "providerDisplayName": "Mistral", + "providerModelId": "ministral-8b-2410", + "providerGroup": "Mistral", + "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 2048, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "response_format", + "structured_outputs", + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.0000012", - "completion": "0.0000012", - "image": "0.001734", + "prompt": "0.0000001", + "completion": "0.0000001", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -42921,12 +44755,12 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -42934,221 +44768,210 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 142, - "newest": 201, - "throughputHighToLow": 258, - "latencyLowToHigh": 155, - "pricingLowToHigh": 228, - "pricingHighToLow": 90 + "topWeekly": 138, + "newest": 186, + "throughputHighToLow": 58, + "latencyLowToHigh": 2, + "pricingLowToHigh": 115, + "pricingHighToLow": 203 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { - "name": "Together", - "slug": "together", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 2048, - "providerModelId": "meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo", - "pricing": { - "prompt": "0.0000012", - "completion": "0.0000012", - "image": "0.001734", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" - ], - "inputCost": 1.2, - "outputCost": 1.2 - }, - { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/Llama-3.2-90B-Vision-Instruct", + "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", + "slug": "mistral", + "quantization": "unknown", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "ministral-8b-2410", "pricing": { - "prompt": "0.00000035", - "completion": "0.0000004", - "image": "0.0005058", + "prompt": "0.0000001", + "completion": "0.0000001", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", "response_format", - "top_k", - "seed", - "min_p" + "structured_outputs", + "seed" ], - "inputCost": 0.35, - "outputCost": 0.4 + "inputCost": 0.1, + "outputCost": 0.1, + "throughput": 120.4205, + "latency": 134 } ] }, { - "slug": "anthropic/claude-3.5-sonnet-20240620", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-06-20T00:00:00+00:00", + "slug": "deepseek/deepseek-r1-distill-llama-8b", + "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Llama-8B", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-02-07T14:15:18.18663+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3.5 Sonnet (2024-06-20)", - "shortName": "Claude 3.5 Sonnet (2024-06-20)", - "author": "anthropic", - "description": "Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\nFor the latest version (2024-10-23), check out [Claude 3.5 Sonnet](/anthropic/claude-3.5-sonnet).\n\n#multimodal", - "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", - "contextLength": 200000, + "name": "DeepSeek: R1 Distill Llama 8B", + "shortName": "R1 Distill Llama 8B", + "author": "deepseek", + "description": "DeepSeek R1 Distill Llama 8B is a distilled large language model based on [Llama-3.1-8B-Instruct](/meta-llama/llama-3.1-8b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 50.4\n- MATH-500 pass@1: 89.1\n- CodeForces Rating: 1205\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.\n\nHugging Face: \n- [Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) \n- [DeepSeek-R1-Distill-Llama-8B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B) |", + "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", + "contextLength": 32000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", - "instructType": null, + "group": "Llama3", + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3.5-sonnet-20240620", - "reasoningConfig": null, - "features": {}, + "permaslug": "deepseek/deepseek-r1-distill-llama-8b", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "20a76230-7394-42f9-bcac-76081ff24cd8", - "name": "Anthropic | anthropic/claude-3.5-sonnet-20240620", - "contextLength": 200000, + "id": "fbc63cef-f2e8-4e4b-a135-9b45485e0eaf", + "name": "Novita | deepseek/deepseek-r1-distill-llama-8b", + "contextLength": 32000, "model": { - "slug": "anthropic/claude-3.5-sonnet-20240620", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-06-20T00:00:00+00:00", + "slug": "deepseek/deepseek-r1-distill-llama-8b", + "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Llama-8B", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-02-07T14:15:18.18663+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3.5 Sonnet (2024-06-20)", - "shortName": "Claude 3.5 Sonnet (2024-06-20)", - "author": "anthropic", - "description": "Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\nFor the latest version (2024-10-23), check out [Claude 3.5 Sonnet](/anthropic/claude-3.5-sonnet).\n\n#multimodal", - "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", - "contextLength": 200000, + "name": "DeepSeek: R1 Distill Llama 8B", + "shortName": "R1 Distill Llama 8B", + "author": "deepseek", + "description": "DeepSeek R1 Distill Llama 8B is a distilled large language model based on [Llama-3.1-8B-Instruct](/meta-llama/llama-3.1-8b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 50.4\n- MATH-500 pass@1: 89.1\n- CodeForces Rating: 1205\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.\n\nHugging Face: \n- [Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) \n- [DeepSeek-R1-Distill-Llama-8B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B) |", + "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", + "contextLength": 0, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", - "instructType": null, + "group": "Llama3", + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3.5-sonnet-20240620", - "reasoningConfig": null, - "features": {} + "permaslug": "deepseek/deepseek-r1-distill-llama-8b", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "anthropic/claude-3.5-sonnet-20240620", - "modelVariantPermaslug": "anthropic/claude-3.5-sonnet-20240620", - "providerName": "Anthropic", + "modelVariantSlug": "deepseek/deepseek-r1-distill-llama-8b", + "modelVariantPermaslug": "deepseek/deepseek-r1-distill-llama-8b", + "providerName": "Novita", "providerInfo": { - "name": "Anthropic", - "displayName": "Anthropic", - "slug": "anthropic", + "name": "Novita", + "displayName": "NovitaAI", + "slug": "novita", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "training": false + } }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "Anthropic", + "moderationRequired": false, + "group": "Novita", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.anthropic.com/", + "statusPageUrl": "https://status.novita.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Anthropic.svg" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "invertRequired": true } }, - "providerDisplayName": "Anthropic", - "providerModelId": "claude-3-5-sonnet-20240620", - "providerGroup": "Anthropic", + "providerDisplayName": "NovitaAI", + "providerModelId": "deepseek/deepseek-r1-distill-llama-8b", + "providerGroup": "Novita", "quantization": "unknown", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 8192, + "maxCompletionTokens": 32000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", "top_k", - "stop" + "min_p", + "repetition_penalty", + "logit_bias" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, "pricing": { - "prompt": "0.000003", - "completion": "0.000015", - "image": "0.0048", + "prompt": "0.00000004", + "completion": "0.00000004", + "image": "0", "request": "0", - "inputCacheRead": "0.0000003", - "inputCacheWrite": "0.00000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -43157,8 +44980,8 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, + "supportsToolParameters": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, @@ -43170,88 +44993,66 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 143, - "newest": 244, - "throughputHighToLow": 134, - "latencyLowToHigh": 273, - "pricingLowToHigh": 274, - "pricingHighToLow": 46 + "topWeekly": 139, + "newest": 117, + "throughputHighToLow": 194, + "latencyLowToHigh": 208, + "pricingLowToHigh": 87, + "pricingHighToLow": 230 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [ { - "name": "Anthropic", - "slug": "anthropic", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 8192, - "providerModelId": "claude-3-5-sonnet-20240620", - "pricing": { - "prompt": "0.000003", - "completion": "0.000015", - "image": "0.0048", - "request": "0", - "inputCacheRead": "0.0000003", - "inputCacheWrite": "0.00000375", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "top_k", - "stop" - ], - "inputCost": 3, - "outputCost": 15 - }, - { - "name": "Google Vertex", - "slug": "vertex", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 8192, - "providerModelId": "claude-3-5-sonnet@20240620", + "context": 32000, + "maxCompletionTokens": 32000, + "providerModelId": "deepseek/deepseek-r1-distill-llama-8b", "pricing": { - "prompt": "0.000003", - "completion": "0.000015", - "image": "0.0048", + "prompt": "0.00000004", + "completion": "0.00000004", + "image": "0", "request": "0", - "inputCacheRead": "0.0000003", - "inputCacheWrite": "0.00000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", "top_k", - "stop" + "min_p", + "repetition_penalty", + "logit_bias" ], - "inputCost": 3, - "outputCost": 15 + "inputCost": 0.04, + "outputCost": 0.04, + "throughput": 43.8155, + "latency": 1510 } ] }, { - "slug": "meta-llama/llama-3.3-70b-instruct", - "hfSlug": "meta-llama/Llama-3.3-70B-Instruct", + "slug": "undi95/remm-slerp-l2-13b", + "hfSlug": "Undi95/ReMM-SLERP-L2-13B", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-06T17:28:57.828422+00:00", + "createdAt": "2023-07-22T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.3 70B Instruct (free)", - "shortName": "Llama 3.3 70B Instruct (free)", - "author": "meta-llama", - "description": "The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.\n\nSupported languages: English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.\n\n[Model Card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/MODEL_CARD.md)", - "modelVersionGroupId": "397604e2-45fa-454e-a85d-9921f5138747", - "contextLength": 8000, + "name": "ReMM SLERP 13B", + "shortName": "ReMM SLERP 13B", + "author": "undi95", + "description": "A recreation trial of the original MythoMax-L2-B13 but with updated models. #merge", + "modelVersionGroupId": null, + "contextLength": 6144, "inputModalities": [ "text" ], @@ -43259,35 +45060,35 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Llama2", + "instructType": "alpaca", "defaultSystem": null, "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" + "###", + "" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.3-70b-instruct", + "permaslug": "undi95/remm-slerp-l2-13b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "4625027b-9188-4d61-b21a-ae6808bd4e5d", - "name": "Crusoe | meta-llama/llama-3.3-70b-instruct:free", - "contextLength": 8000, + "id": "989a7c1b-f4c8-4d78-b0e3-caf0adb687c5", + "name": "Mancer | undi95/remm-slerp-l2-13b", + "contextLength": 6144, "model": { - "slug": "meta-llama/llama-3.3-70b-instruct", - "hfSlug": "meta-llama/Llama-3.3-70B-Instruct", + "slug": "undi95/remm-slerp-l2-13b", + "hfSlug": "Undi95/ReMM-SLERP-L2-13B", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-06T17:28:57.828422+00:00", + "createdAt": "2023-07-22T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.3 70B Instruct", - "shortName": "Llama 3.3 70B Instruct", - "author": "meta-llama", - "description": "The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.\n\nSupported languages: English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.\n\n[Model Card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/MODEL_CARD.md)", - "modelVersionGroupId": "397604e2-45fa-454e-a85d-9921f5138747", - "contextLength": 131072, + "name": "ReMM SLERP 13B", + "shortName": "ReMM SLERP 13B", + "author": "undi95", + "description": "A recreation trial of the original MythoMax-L2-B13 but with updated models. #merge", + "modelVersionGroupId": null, + "contextLength": 4096, "inputModalities": [ "text" ], @@ -43295,41 +45096,41 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Llama2", + "instructType": "alpaca", "defaultSystem": null, "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" + "###", + "" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.3-70b-instruct", + "permaslug": "undi95/remm-slerp-l2-13b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "meta-llama/llama-3.3-70b-instruct:free", - "modelVariantPermaslug": "meta-llama/llama-3.3-70b-instruct:free", - "providerName": "Crusoe", + "modelVariantSlug": "undi95/remm-slerp-l2-13b", + "modelVariantPermaslug": "undi95/remm-slerp-l2-13b", + "providerName": "Mancer", "providerInfo": { - "name": "Crusoe", - "displayName": "Crusoe", - "slug": "crusoe", + "name": "Mancer", + "displayName": "Mancer", + "slug": "mancer", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://legal.crusoe.ai/open-router#cloud-privacy-policy", - "termsOfServiceUrl": "https://legal.crusoe.ai/open-router#managed-inference-tos-open-router", + "termsOfServiceUrl": "https://mancer.tech/terms", + "privacyPolicyUrl": "https://mancer.tech/privacy", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Crusoe", + "group": "Mancer", "editors": [], "owners": [], "isMultipartSupported": true, @@ -43337,18 +45138,18 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://crusoe.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256" } }, - "providerDisplayName": "Crusoe", - "providerModelId": "meta-llama/Llama-3.3-70B-Instruct", - "providerGroup": "Crusoe", - "quantization": "bf16", - "variant": "free", - "isFree": true, + "providerDisplayName": "Mancer", + "providerModelId": "remm-slerp", + "providerGroup": "Mancer", + "quantization": "unknown", + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 8000, + "maxCompletionTokens": 1024, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -43358,26 +45159,33 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed" + "repetition_penalty", + "logit_bias", + "top_k", + "min_p", + "seed", + "top_a" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://legal.crusoe.ai/open-router#cloud-privacy-policy", - "termsOfServiceUrl": "https://legal.crusoe.ai/open-router#managed-inference-tos-open-router", + "termsOfServiceUrl": "https://mancer.tech/terms", + "privacyPolicyUrl": "https://mancer.tech/privacy", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true }, - "training": false + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.0000005625", + "completion": "0.00000084375", "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", - "discount": 0 + "discount": 0.25 }, "variablePricings": [], "isHidden": false, @@ -43396,33 +45204,33 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 8, - "newest": 157, - "throughputHighToLow": 231, - "latencyLowToHigh": 19, - "pricingLowToHigh": 56, - "pricingHighToLow": 208 + "topWeekly": 140, + "newest": 312, + "throughputHighToLow": 200, + "latencyLowToHigh": 154, + "pricingLowToHigh": 187, + "pricingHighToLow": 131 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/Llama-3.3-70B-Instruct-Turbo", + "name": "Mancer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", + "slug": "mancer", + "quantization": "unknown", + "context": 6144, + "maxCompletionTokens": 1024, + "providerModelId": "remm-slerp", "pricing": { - "prompt": "0.00000008", - "completion": "0.00000025", + "prompt": "0.0000005625", + "completion": "0.00000084375", "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", - "discount": 0 + "discount": 0.25 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -43430,47 +45238,28 @@ export default { "frequency_penalty", "presence_penalty", "repetition_penalty", - "response_format", + "logit_bias", "top_k", + "min_p", "seed", - "min_p" + "top_a" ], - "inputCost": 0.08, - "outputCost": 0.25 - }, - { - "name": "kluster.ai", - "slug": "klusterAi", - "quantization": "fp8", - "context": 131000, - "maxCompletionTokens": 131000, - "providerModelId": "klusterai/Meta-Llama-3.3-70B-Instruct-Turbo", - "pricing": { - "prompt": "0.00000008", - "completion": "0.00000035", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature" - ], - "inputCost": 0.08, - "outputCost": 0.35 + "inputCost": 0.56, + "outputCost": 0.84, + "throughput": 41.7145, + "latency": 888.5 }, { - "name": "inference.net", - "slug": "inferenceNet", + "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", + "slug": "featherless", "quantization": "fp8", - "context": 128000, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/llama-3.3-70b-instruct/fp-8", + "context": 4096, + "maxCompletionTokens": 4096, + "providerModelId": "Undi95/ReMM-SLERP-L2-13B", "pricing": { - "prompt": "0.0000001", - "completion": "0.00000025", + "prompt": "0.0000008", + "completion": "0.0000012", "image": "0", "request": "0", "webSearch": "0", @@ -43481,30 +45270,30 @@ export default { "max_tokens", "temperature", "top_p", - "structured_outputs", - "response_format", "stop", "frequency_penalty", "presence_penalty", - "seed", + "repetition_penalty", "top_k", "min_p", - "logit_bias", - "top_logprobs" + "seed" ], - "inputCost": 0.1, - "outputCost": 0.25 + "inputCost": 0.8, + "outputCost": 1.2, + "throughput": 15.738, + "latency": 1279 }, { - "name": "Lambda", - "slug": "lambda", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "llama3.3-70b-instruct-fp8", + "name": "Mancer (private)", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", + "slug": "mancer (private)", + "quantization": "unknown", + "context": 6144, + "maxCompletionTokens": 1024, + "providerModelId": "remm-slerp", "pricing": { - "prompt": "0.00000012", - "completion": "0.0000003", + "prompt": "0.000001", + "completion": "0.0000015", "image": "0", "request": "0", "webSearch": "0", @@ -43518,92 +45307,447 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", + "repetition_penalty", "logit_bias", - "logprobs", - "top_logprobs", - "response_format" + "top_k", + "min_p", + "seed", + "top_a" ], - "inputCost": 0.12, - "outputCost": 0.3 + "inputCost": 1, + "outputCost": 1.5, + "throughput": 42.253, + "latency": 722 + } + ] + }, + { + "slug": "meta-llama/llama-3.2-90b-vision-instruct", + "hfSlug": "meta-llama/Llama-3.2-90B-Vision-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-09-25T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Meta: Llama 3.2 90B Vision Instruct", + "shortName": "Llama 3.2 90B Vision Instruct", + "author": "meta-llama", + "description": "The Llama 90B Vision model is a top-tier, 90-billion-parameter multimodal model designed for the most challenging visual reasoning and language tasks. It offers unparalleled accuracy in image captioning, visual question answering, and advanced image-text comprehension. Pre-trained on vast multimodal datasets and fine-tuned with human feedback, the Llama 90B Vision is engineered to handle the most demanding image-based AI tasks.\n\nThis model is perfect for industries requiring cutting-edge multimodal AI capabilities, particularly those dealing with complex, real-time visual and textual analysis.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "modelVersionGroupId": null, + "contextLength": 131072, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Llama3", + "instructType": "llama3", + "defaultSystem": null, + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "meta-llama/llama-3.2-90b-vision-instruct", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "6fa97646-3631-4446-beaa-8a24dd07ddb1", + "name": "Together | meta-llama/llama-3.2-90b-vision-instruct", + "contextLength": 131072, + "model": { + "slug": "meta-llama/llama-3.2-90b-vision-instruct", + "hfSlug": "meta-llama/Llama-3.2-90B-Vision-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-09-25T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Meta: Llama 3.2 90B Vision Instruct", + "shortName": "Llama 3.2 90B Vision Instruct", + "author": "meta-llama", + "description": "The Llama 90B Vision model is a top-tier, 90-billion-parameter multimodal model designed for the most challenging visual reasoning and language tasks. It offers unparalleled accuracy in image captioning, visual question answering, and advanced image-text comprehension. Pre-trained on vast multimodal datasets and fine-tuned with human feedback, the Llama 90B Vision is engineered to handle the most demanding image-based AI tasks.\n\nThis model is perfect for industries requiring cutting-edge multimodal AI capabilities, particularly those dealing with complex, real-time visual and textual analysis.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "modelVersionGroupId": null, + "contextLength": 131072, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Llama3", + "instructType": "llama3", + "defaultSystem": null, + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "meta-llama/llama-3.2-90b-vision-instruct", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "meta-llama/llama-3.2-90b-vision-instruct", + "modelVariantPermaslug": "meta-llama/llama-3.2-90b-vision-instruct", + "providerName": "Together", + "providerInfo": { + "name": "Together", + "displayName": "Together", + "slug": "together", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", + "paidModels": { + "training": false, + "retainsPrompts": false + } + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "Together", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" + } + }, + "providerDisplayName": "Together", + "providerModelId": "meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo", + "providerGroup": "Together", + "quantization": "fp8", + "variant": "standard", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": 2048, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", + "paidModels": { + "training": false, + "retainsPrompts": false + }, + "training": false, + "retainsPrompts": false + }, + "pricing": { + "prompt": "0.0000012", + "completion": "0.0000012", + "image": "0.001734", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 141, + "newest": 201, + "throughputHighToLow": 261, + "latencyLowToHigh": 98, + "pricingLowToHigh": 228, + "pricingHighToLow": 90 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", + "providers": [ { - "name": "Phala", - "slug": "phala", - "quantization": null, + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": "fp8", "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "phala/llama-3.3-70b-instruct", + "maxCompletionTokens": 2048, + "providerModelId": "meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo", "pricing": { - "prompt": "0.00000012", - "completion": "0.00000035", - "image": "0", + "prompt": "0.0000012", + "completion": "0.0000012", + "image": "0.001734", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "seed", "top_k", + "repetition_penalty", + "logit_bias", "min_p", - "repetition_penalty" + "response_format" ], - "inputCost": 0.12, - "outputCost": 0.35 + "inputCost": 1.2, + "outputCost": 1.2, + "throughput": 25.1865, + "latency": 723 }, { - "name": "NovitaAI", - "slug": "novitaAi", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/llama-3.3-70b-instruct", + "context": 32768, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/Llama-3.2-90B-Vision-Instruct", "pricing": { - "prompt": "0.00000013", - "completion": "0.00000039", - "image": "0", + "prompt": "0.00000035", + "completion": "0.0000004", + "image": "0.0005058", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "min_p", "repetition_penalty", - "logit_bias" + "response_format", + "top_k", + "seed", + "min_p" ], - "inputCost": 0.13, - "outputCost": 0.39 + "inputCost": 0.35, + "outputCost": 0.4, + "throughput": 19.3805, + "latency": 1697.5 + } + ] + }, + { + "slug": "x-ai/grok-2-1212", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-12-15T03:20:14.161318+00:00", + "hfUpdatedAt": null, + "name": "xAI: Grok 2 1212", + "shortName": "Grok 2 1212", + "author": "x-ai", + "description": "Grok 2 1212 introduces significant enhancements to accuracy, instruction adherence, and multilingual support, making it a powerful and flexible choice for developers seeking a highly steerable, intelligent model.", + "modelVersionGroupId": null, + "contextLength": 131072, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Grok", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "x-ai/grok-2-1212", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "9d94b368-75f7-44b1-a056-f8a1240da545", + "name": "xAI | x-ai/grok-2-1212", + "contextLength": 131072, + "model": { + "slug": "x-ai/grok-2-1212", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-12-15T03:20:14.161318+00:00", + "hfUpdatedAt": null, + "name": "xAI: Grok 2 1212", + "shortName": "Grok 2 1212", + "author": "x-ai", + "description": "Grok 2 1212 introduces significant enhancements to accuracy, instruction adherence, and multilingual support, making it a powerful and flexible choice for developers seeking a highly steerable, intelligent model.", + "modelVersionGroupId": null, + "contextLength": 131072, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Grok", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "x-ai/grok-2-1212", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "x-ai/grok-2-1212", + "modelVariantPermaslug": "x-ai/grok-2-1212", + "providerName": "xAI", + "providerInfo": { + "name": "xAI", + "displayName": "xAI", + "slug": "xai", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://x.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://x.ai/legal/privacy-policy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + } + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "xAI", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": "https://status.x.ai/", + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://x.ai/&size=256" + } + }, + "providerDisplayName": "xAI", + "providerModelId": "grok-2-1212", + "providerGroup": "xAI", + "quantization": null, + "variant": "standard", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "logprobs", + "top_logprobs", + "response_format" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://x.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://x.ai/legal/privacy-policy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "pricing": { + "prompt": "0.000002", + "completion": "0.00001", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": true, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 142, + "newest": 154, + "throughputHighToLow": 92, + "latencyLowToHigh": 34, + "pricingLowToHigh": 255, + "pricingHighToLow": 64 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://x.ai/\\u0026size=256", + "providers": [ { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", + "name": "xAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://x.ai/&size=256", + "slug": "xAi", + "quantization": null, "context": 131072, "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-3.3-70B-Instruct", + "providerModelId": "grok-2-1212", "pricing": { - "prompt": "0.00000013", - "completion": "0.0000004", + "prompt": "0.000002", + "completion": "0.00001", "image": "0", "request": "0", "webSearch": "0", @@ -43611,6 +45755,8 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", @@ -43618,115 +45764,189 @@ export default { "frequency_penalty", "presence_penalty", "seed", - "top_k", - "logit_bias", "logprobs", - "top_logprobs" + "top_logprobs", + "response_format" ], - "inputCost": 0.13, - "outputCost": 0.4 - }, - { - "name": "Parasail", - "slug": "parasail", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "parasail-llama-33-70b-fp8", - "pricing": { - "prompt": "0.00000028", - "completion": "0.00000078", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "presence_penalty", - "frequency_penalty", - "repetition_penalty", - "top_k" + "inputCost": 2, + "outputCost": 10, + "throughput": 87.404, + "latency": 323 + } + ] + }, + { + "slug": "qwen/qwen-vl-max", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-01T18:25:04.223655+00:00", + "hfUpdatedAt": null, + "name": "Qwen: Qwen VL Max", + "shortName": "Qwen VL Max", + "author": "qwen", + "description": "Qwen VL Max is a visual understanding model with 7500 tokens context length. It excels in delivering optimal performance for a broader spectrum of complex tasks.\n", + "modelVersionGroupId": null, + "contextLength": 7500, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Qwen", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "qwen/qwen-vl-max-2025-01-25", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "40a33c72-3801-49f3-ac2a-966d0b249981", + "name": "Alibaba | qwen/qwen-vl-max-2025-01-25", + "contextLength": 7500, + "model": { + "slug": "qwen/qwen-vl-max", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-01T18:25:04.223655+00:00", + "hfUpdatedAt": null, + "name": "Qwen: Qwen VL Max", + "shortName": "Qwen VL Max", + "author": "qwen", + "description": "Qwen VL Max is a visual understanding model with 7500 tokens context length. It excels in delivering optimal performance for a broader spectrum of complex tasks.\n", + "modelVersionGroupId": null, + "contextLength": 7500, + "inputModalities": [ + "text", + "image" ], - "inputCost": 0.28, - "outputCost": 0.78 + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Qwen", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "qwen/qwen-vl-max-2025-01-25", + "reasoningConfig": null, + "features": {} }, - { - "name": "Cloudflare", - "slug": "cloudflare", - "quantization": "fp8", - "context": 24000, - "maxCompletionTokens": null, - "providerModelId": "@cf/meta/llama-3.3-70b-instruct-fp8-fast", - "pricing": { - "prompt": "0.00000029", - "completion": "0.00000225", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 + "modelVariantSlug": "qwen/qwen-vl-max", + "modelVariantPermaslug": "qwen/qwen-vl-max-2025-01-25", + "providerName": "Alibaba", + "providerInfo": { + "name": "Alibaba", + "displayName": "Alibaba", + "slug": "alibaba", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", + "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", + "paidModels": { + "training": false + } }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "top_k", - "seed", - "repetition_penalty", - "frequency_penalty", - "presence_penalty" - ], - "inputCost": 0.29, - "outputCost": 2.25 + "headquarters": "CN", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": false, + "moderationRequired": false, + "group": "Alibaba", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256" + } }, - { - "name": "CentML", - "slug": "centMl", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "meta-llama/Llama-3.3-70B-Instruct", - "pricing": { - "prompt": "0.00000035", - "completion": "0.00000035", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 + "providerDisplayName": "Alibaba", + "providerModelId": "qwen-vl-max", + "providerGroup": "Alibaba", + "quantization": null, + "variant": "standard", + "isFree": false, + "canAbort": false, + "maxPromptTokens": 6000, + "maxCompletionTokens": 1500, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "seed", + "response_format", + "presence_penalty" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", + "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", + "paidModels": { + "training": false }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "structured_outputs", - "response_format", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty" - ], - "inputCost": 0.35, - "outputCost": 0.35 + "training": false + }, + "pricing": { + "prompt": "0.0000008", + "completion": "0.0000032", + "image": "0.001024", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": false, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 143, + "newest": 123, + "throughputHighToLow": 204, + "latencyLowToHigh": 243, + "pricingLowToHigh": 217, + "pricingHighToLow": 100 + }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", + "providers": [ { - "name": "Hyperbolic", - "slug": "hyperbolic", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-3.3-70B-Instruct", + "name": "Alibaba", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256", + "slug": "alibaba", + "quantization": null, + "context": 7500, + "maxCompletionTokens": 1500, + "providerModelId": "qwen-vl-max", "pricing": { - "prompt": "0.0000004", - "completion": "0.0000004", - "image": "0", + "prompt": "0.0000008", + "completion": "0.0000032", + "image": "0.001024", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -43736,56 +45956,203 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "logprobs", - "top_logprobs", "seed", - "logit_bias", - "top_k", - "min_p", - "repetition_penalty" + "response_format", + "presence_penalty" ], - "inputCost": 0.4, - "outputCost": 0.4 + "inputCost": 0.8, + "outputCost": 3.2, + "throughput": 38.802, + "latency": 1905.5 + } + ] + }, + { + "slug": "anthropic/claude-3.5-sonnet-20240620", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-06-20T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Anthropic: Claude 3.5 Sonnet (2024-06-20)", + "shortName": "Claude 3.5 Sonnet (2024-06-20)", + "author": "anthropic", + "description": "Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\nFor the latest version (2024-10-23), check out [Claude 3.5 Sonnet](/anthropic/claude-3.5-sonnet).\n\n#multimodal", + "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", + "contextLength": 200000, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Claude", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "anthropic/claude-3.5-sonnet-20240620", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "20a76230-7394-42f9-bcac-76081ff24cd8", + "name": "Anthropic | anthropic/claude-3.5-sonnet-20240620", + "contextLength": 200000, + "model": { + "slug": "anthropic/claude-3.5-sonnet-20240620", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-06-20T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Anthropic: Claude 3.5 Sonnet (2024-06-20)", + "shortName": "Claude 3.5 Sonnet (2024-06-20)", + "author": "anthropic", + "description": "Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\nFor the latest version (2024-10-23), check out [Claude 3.5 Sonnet](/anthropic/claude-3.5-sonnet).\n\n#multimodal", + "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", + "contextLength": 200000, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Claude", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "anthropic/claude-3.5-sonnet-20240620", + "reasoningConfig": null, + "features": {} }, - { - "name": "Atoma", - "slug": "atoma", - "quantization": "fp8", - "context": 104962, - "maxCompletionTokens": 100000, - "providerModelId": "Infermatic/Llama-3.3-70B-Instruct-FP8-Dynamic", - "pricing": { - "prompt": "0.0000004", - "completion": "0.0000004", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 + "modelVariantSlug": "anthropic/claude-3.5-sonnet-20240620", + "modelVariantPermaslug": "anthropic/claude-3.5-sonnet-20240620", + "providerName": "Anthropic", + "providerInfo": { + "name": "Anthropic", + "displayName": "Anthropic", + "slug": "anthropic", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p" - ], - "inputCost": 0.4, - "outputCost": 0.4 + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": true, + "group": "Anthropic", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": "https://status.anthropic.com/", + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/Anthropic.svg" + } + }, + "providerDisplayName": "Anthropic", + "providerModelId": "claude-3-5-sonnet-20240620", + "providerGroup": "Anthropic", + "quantization": "unknown", + "variant": "standard", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": 8192, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "isByok": false, + "moderationRequired": true, + "dataPolicy": { + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "pricing": { + "prompt": "0.000003", + "completion": "0.000015", + "image": "0.0048", + "request": "0", + "inputCacheRead": "0.0000003", + "inputCacheWrite": "0.00000375", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": true, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 144, + "newest": 244, + "throughputHighToLow": 134, + "latencyLowToHigh": 280, + "pricingLowToHigh": 274, + "pricingHighToLow": 46 + }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", + "providers": [ { - "name": "Groq", - "slug": "groq", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 32768, - "providerModelId": "llama-3.3-70b-versatile", + "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", + "slug": "anthropic", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 8192, + "providerModelId": "claude-3-5-sonnet-20240620", "pricing": { - "prompt": "0.00000059", - "completion": "0.00000079", - "image": "0", + "prompt": "0.000003", + "completion": "0.000015", + "image": "0.0048", "request": "0", + "inputCacheRead": "0.0000003", + "inputCacheWrite": "0.00000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -43796,30 +46163,29 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "top_logprobs", - "logprobs", - "logit_bias", - "seed" + "top_k", + "stop" ], - "inputCost": 0.59, - "outputCost": 0.79 + "inputCost": 3, + "outputCost": 15, + "throughput": 60.469, + "latency": 2826 }, { - "name": "Friendli", - "slug": "friendli", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "meta-llama-3.3-70b-instruct", + "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", + "slug": "vertex", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 8192, + "providerModelId": "claude-3-5-sonnet@20240620", "pricing": { - "prompt": "0.0000006", - "completion": "0.0000006", - "image": "0", + "prompt": "0.000003", + "completion": "0.000015", + "image": "0.0048", "request": "0", + "inputCacheRead": "0.0000003", + "inputCacheWrite": "0.00000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -43830,171 +46196,421 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", "top_k", - "min_p", - "repetition_penalty", - "response_format", - "structured_outputs" + "stop" ], - "inputCost": 0.6, - "outputCost": 0.6 + "inputCost": 3, + "outputCost": 15, + "throughput": 62.276, + "latency": 994 + } + ] + }, + { + "slug": "nousresearch/deephermes-3-mistral-24b-preview", + "hfSlug": "NousResearch/DeepHermes-3-Mistral-24B-Preview", + "updatedAt": "2025-05-09T23:02:10.58499+00:00", + "createdAt": "2025-05-09T22:48:24.165453+00:00", + "hfUpdatedAt": null, + "name": "Nous: DeepHermes 3 Mistral 24B Preview (free)", + "shortName": "DeepHermes 3 Mistral 24B Preview (free)", + "author": "nousresearch", + "description": "DeepHermes 3 (Mistral 24B Preview) is an instruction-tuned language model by Nous Research based on Mistral-Small-24B, designed for chat, function calling, and advanced multi-turn reasoning. It introduces a dual-mode system that toggles between intuitive chat responses and structured “deep reasoning” mode using special system prompts. Fine-tuned via distillation from R1, it supports structured output (JSON mode) and function call syntax for agent-based applications.\n\nDeepHermes 3 supports a **reasoning toggle via system prompt**, allowing users to switch between fast, intuitive responses and deliberate, multi-step reasoning. When activated with the following specific system instruction, the model enters a *\"deep thinking\"* mode—generating extended chains of thought wrapped in `` tags before delivering a final answer. \n\nSystem Prompt: You are a deep thinking AI, you may use extremely long chains of thought to deeply consider the problem and deliberate with yourself via systematic reasoning processes to help come to a correct solution prior to answering. You should enclose your thoughts and internal monologue inside tags, and then provide your solution or response to the problem.\n", + "modelVersionGroupId": null, + "contextLength": 32768, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Other", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "nousresearch/deephermes-3-mistral-24b-preview", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "90728530-42b4-4bcf-bc67-167eee5bbac4", + "name": "Chutes | nousresearch/deephermes-3-mistral-24b-preview:free", + "contextLength": 32768, + "model": { + "slug": "nousresearch/deephermes-3-mistral-24b-preview", + "hfSlug": "NousResearch/DeepHermes-3-Mistral-24B-Preview", + "updatedAt": "2025-05-09T23:02:10.58499+00:00", + "createdAt": "2025-05-09T22:48:24.165453+00:00", + "hfUpdatedAt": null, + "name": "Nous: DeepHermes 3 Mistral 24B Preview", + "shortName": "DeepHermes 3 Mistral 24B Preview", + "author": "nousresearch", + "description": "DeepHermes 3 (Mistral 24B Preview) is an instruction-tuned language model by Nous Research based on Mistral-Small-24B, designed for chat, function calling, and advanced multi-turn reasoning. It introduces a dual-mode system that toggles between intuitive chat responses and structured “deep reasoning” mode using special system prompts. Fine-tuned via distillation from R1, it supports structured output (JSON mode) and function call syntax for agent-based applications.\n\nDeepHermes 3 supports a **reasoning toggle via system prompt**, allowing users to switch between fast, intuitive responses and deliberate, multi-step reasoning. When activated with the following specific system instruction, the model enters a *\"deep thinking\"* mode—generating extended chains of thought wrapped in `` tags before delivering a final answer. \n\nSystem Prompt: You are a deep thinking AI, you may use extremely long chains of thought to deeply consider the problem and deliberate with yourself via systematic reasoning processes to help come to a correct solution prior to answering. You should enclose your thoughts and internal monologue inside tags, and then provide your solution or response to the problem.\n", + "modelVersionGroupId": null, + "contextLength": 32768, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Other", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "nousresearch/deephermes-3-mistral-24b-preview", + "reasoningConfig": null, + "features": {} }, - { - "name": "NextBit", - "slug": "nextBit", - "quantization": "fp8", - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "llama3.3:70b", - "pricing": { - "prompt": "0.0000006", - "completion": "0.00000075", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 + "modelVariantSlug": "nousresearch/deephermes-3-mistral-24b-preview:free", + "modelVariantPermaslug": "nousresearch/deephermes-3-mistral-24b-preview:free", + "providerName": "Chutes", + "providerInfo": { + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://chutes.ai/tos", + "paidModels": { + "training": true, + "retainsPrompts": true + } }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "structured_outputs" - ], - "inputCost": 0.6, - "outputCost": 0.75 + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "Chutes", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + } }, - { - "name": "SambaNova", - "slug": "sambaNova", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": 3072, - "providerModelId": "Meta-Llama-3.3-70B-Instruct", - "pricing": { - "prompt": "0.0000006", - "completion": "0.0000012", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 + "providerDisplayName": "Chutes", + "providerModelId": "NousResearch/DeepHermes-3-Mistral-24B-Preview", + "providerGroup": "Chutes", + "quantization": null, + "variant": "free", + "isFree": true, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://chutes.ai/tos", + "paidModels": { + "training": true, + "retainsPrompts": true }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "top_k", - "stop" + "training": true, + "retainsPrompts": true + }, + "pricing": { + "prompt": "0", + "completion": "0", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": true, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 145, + "newest": 0, + "throughputHighToLow": 99, + "latencyLowToHigh": 175, + "pricingLowToHigh": 0, + "pricingHighToLow": 248 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nousresearch.com/\\u0026size=256", + "providers": [] + }, + { + "slug": "anthropic/claude-3-haiku", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-03-13T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Anthropic: Claude 3 Haiku (self-moderated)", + "shortName": "Claude 3 Haiku (self-moderated)", + "author": "anthropic", + "description": "Claude 3 Haiku is Anthropic's fastest and most compact model for\nnear-instant responsiveness. Quick and accurate targeted performance.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-haiku)\n\n#multimodal", + "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", + "contextLength": 200000, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Claude", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "anthropic/claude-3-haiku", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "8661a1db-b0cf-4eb2-ba04-c2a79f698682", + "name": "Anthropic | anthropic/claude-3-haiku:beta", + "contextLength": 200000, + "model": { + "slug": "anthropic/claude-3-haiku", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-03-13T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Anthropic: Claude 3 Haiku", + "shortName": "Claude 3 Haiku", + "author": "anthropic", + "description": "Claude 3 Haiku is Anthropic's fastest and most compact model for\nnear-instant responsiveness. Quick and accurate targeted performance.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-haiku)\n\n#multimodal", + "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", + "contextLength": 200000, + "inputModalities": [ + "text", + "image" ], - "inputCost": 0.6, - "outputCost": 1.2 + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Claude", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "anthropic/claude-3-haiku", + "reasoningConfig": null, + "features": {} }, - { - "name": "Cerebras", - "slug": "cerebras", - "quantization": "fp16", - "context": 32000, - "maxCompletionTokens": 32000, - "providerModelId": "llama-3.3-70b", - "pricing": { - "prompt": "0.00000085", - "completion": "0.0000012", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 + "modelVariantSlug": "anthropic/claude-3-haiku:beta", + "modelVariantPermaslug": "anthropic/claude-3-haiku:beta", + "providerName": "Anthropic", + "providerInfo": { + "name": "Anthropic", + "displayName": "Anthropic", + "slug": "anthropic", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "structured_outputs", - "response_format", - "stop", - "seed", - "logprobs", - "top_logprobs" - ], - "inputCost": 0.85, - "outputCost": 1.2 + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": true, + "group": "Anthropic", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": "https://status.anthropic.com/", + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/Anthropic.svg" + } + }, + "providerDisplayName": "Anthropic", + "providerModelId": "claude-3-haiku-20240307", + "providerGroup": "Anthropic", + "quantization": "unknown", + "variant": "beta", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": 4096, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, + "pricing": { + "prompt": "0.00000025", + "completion": "0.00000125", + "image": "0.0004", + "request": "0", + "inputCacheRead": "0.00000003", + "inputCacheWrite": "0.0000003", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": true, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 62, + "newest": 277, + "throughputHighToLow": 28, + "latencyLowToHigh": 133, + "pricingLowToHigh": 168, + "pricingHighToLow": 149 + }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", + "providers": [ { - "name": "Together", - "slug": "together", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 2048, - "providerModelId": "meta-llama/Llama-3.3-70B-Instruct-Turbo", + "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", + "slug": "anthropic", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 4096, + "providerModelId": "claude-3-haiku-20240307", "pricing": { - "prompt": "0.00000088", - "completion": "0.00000088", - "image": "0", + "prompt": "0.00000025", + "completion": "0.00000125", + "image": "0.0004", "request": "0", + "inputCacheRead": "0.00000003", + "inputCacheWrite": "0.0000003", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "stop" ], - "inputCost": 0.88, - "outputCost": 0.88 + "inputCost": 0.25, + "outputCost": 1.25, + "throughput": 152.853, + "latency": 780.5 }, { - "name": "Fireworks", - "slug": "fireworks", - "quantization": "fp16", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/llama-v3p3-70b-instruct", + "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", + "slug": "vertex", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 4096, + "providerModelId": "claude-3-haiku@20240307", "pricing": { - "prompt": "0.0000009", - "completion": "0.0000009", - "image": "0", + "prompt": "0.00000025", + "completion": "0.00000125", + "image": "0.0004", "request": "0", + "inputCacheRead": "0.00000003", + "inputCacheWrite": "0.0000003", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", "top_k", - "repetition_penalty", - "response_format", - "structured_outputs", - "logit_bias", - "logprobs", - "top_logprobs" + "stop" ], - "inputCost": 0.9, - "outputCost": 0.9 + "inputCost": 0.25, + "outputCost": 1.25, + "throughput": 152.9625, + "latency": 1510 } ] }, @@ -44179,16 +46795,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 102, + "topWeekly": 91, "newest": 26, - "throughputHighToLow": 63, - "latencyLowToHigh": 171, + "throughputHighToLow": 64, + "latencyLowToHigh": 168, "pricingLowToHigh": 11, "pricingHighToLow": 212 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "fp8", "context": 40960, @@ -44219,10 +46837,13 @@ export default { "min_p" ], "inputCost": 0.07, - "outputCost": 0.24 + "outputCost": 0.24, + "throughput": 75.656, + "latency": 1075 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "fp8", "context": 128000, @@ -44249,156 +46870,639 @@ export default { "seed", "top_k", "min_p", - "repetition_penalty", - "logit_bias" + "repetition_penalty", + "logit_bias" + ], + "inputCost": 0.07, + "outputCost": 0.28, + "throughput": 57.333, + "latency": 787 + }, + { + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 40960, + "maxCompletionTokens": null, + "providerModelId": "Qwen/Qwen3-14B", + "pricing": { + "prompt": "0.00000008", + "completion": "0.00000024", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.08, + "outputCost": 0.24, + "throughput": 93.2925, + "latency": 273 + } + ] + }, + { + "slug": "anthropic/claude-3.5-sonnet-20240620", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-06-20T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Anthropic: Claude 3.5 Sonnet (2024-06-20) (self-moderated)", + "shortName": "Claude 3.5 Sonnet (2024-06-20) (self-moderated)", + "author": "anthropic", + "description": "Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\nFor the latest version (2024-10-23), check out [Claude 3.5 Sonnet](/anthropic/claude-3.5-sonnet).\n\n#multimodal", + "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", + "contextLength": 200000, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Claude", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "anthropic/claude-3.5-sonnet-20240620", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "20a76230-7394-42f9-bcac-76081ff24cd8", + "name": "Anthropic | anthropic/claude-3.5-sonnet-20240620:beta", + "contextLength": 200000, + "model": { + "slug": "anthropic/claude-3.5-sonnet-20240620", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-06-20T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Anthropic: Claude 3.5 Sonnet (2024-06-20)", + "shortName": "Claude 3.5 Sonnet (2024-06-20)", + "author": "anthropic", + "description": "Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\nFor the latest version (2024-10-23), check out [Claude 3.5 Sonnet](/anthropic/claude-3.5-sonnet).\n\n#multimodal", + "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", + "contextLength": 200000, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Claude", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "anthropic/claude-3.5-sonnet-20240620", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "anthropic/claude-3.5-sonnet-20240620:beta", + "modelVariantPermaslug": "anthropic/claude-3.5-sonnet-20240620:beta", + "providerName": "Anthropic", + "providerInfo": { + "name": "Anthropic", + "displayName": "Anthropic", + "slug": "anthropic", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": true, + "group": "Anthropic", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": "https://status.anthropic.com/", + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/Anthropic.svg" + } + }, + "providerDisplayName": "Anthropic", + "providerModelId": "claude-3-5-sonnet-20240620", + "providerGroup": "Anthropic", + "quantization": "unknown", + "variant": "beta", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": 8192, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "pricing": { + "prompt": "0.000003", + "completion": "0.000015", + "image": "0.0048", + "request": "0", + "inputCacheRead": "0.0000003", + "inputCacheWrite": "0.00000375", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": true, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 144, + "newest": 244, + "throughputHighToLow": 134, + "latencyLowToHigh": 280, + "pricingLowToHigh": 274, + "pricingHighToLow": 46 + }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", + "providers": [ + { + "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", + "slug": "anthropic", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 8192, + "providerModelId": "claude-3-5-sonnet-20240620", + "pricing": { + "prompt": "0.000003", + "completion": "0.000015", + "image": "0.0048", + "request": "0", + "inputCacheRead": "0.0000003", + "inputCacheWrite": "0.00000375", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 3, + "outputCost": 15, + "throughput": 60.469, + "latency": 2826 + }, + { + "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", + "slug": "vertex", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 8192, + "providerModelId": "claude-3-5-sonnet@20240620", + "pricing": { + "prompt": "0.000003", + "completion": "0.000015", + "image": "0.0048", + "request": "0", + "inputCacheRead": "0.0000003", + "inputCacheWrite": "0.00000375", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 3, + "outputCost": 15, + "throughput": 62.276, + "latency": 994 + } + ] + }, + { + "slug": "mistralai/pixtral-12b", + "hfSlug": "mistralai/Pixtral-12B-2409", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-09-10T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Mistral: Pixtral 12B", + "shortName": "Pixtral 12B", + "author": "mistralai", + "description": "The first multi-modal, text+image-to-text model from Mistral AI. Its weights were launched via torrent: https://x.com/mistralai/status/1833758285167722836.", + "modelVersionGroupId": null, + "contextLength": 32768, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Mistral", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "mistralai/pixtral-12b", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "b550f7af-571a-45fd-b442-b3327afaf38c", + "name": "Hyperbolic | mistralai/pixtral-12b", + "contextLength": 32768, + "model": { + "slug": "mistralai/pixtral-12b", + "hfSlug": "mistralai/Pixtral-12B-2409", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-09-10T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Mistral: Pixtral 12B", + "shortName": "Pixtral 12B", + "author": "mistralai", + "description": "The first multi-modal, text+image-to-text model from Mistral AI. Its weights were launched via torrent: https://x.com/mistralai/status/1833758285167722836.", + "modelVersionGroupId": null, + "contextLength": 4096, + "inputModalities": [ + "text", + "image" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Mistral", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "mistralai/pixtral-12b", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "mistralai/pixtral-12b", + "modelVariantPermaslug": "mistralai/pixtral-12b", + "providerName": "Hyperbolic", + "providerInfo": { + "name": "Hyperbolic", + "displayName": "Hyperbolic", + "slug": "hyperbolic", + "baseUrl": "url", + "dataPolicy": { + "privacyPolicyUrl": "https://hyperbolic.xyz/privacy", + "termsOfServiceUrl": "https://hyperbolic.xyz/terms", + "paidModels": { + "training": false + } + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "Hyperbolic", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256" + } + }, + "providerDisplayName": "Hyperbolic", + "providerModelId": "mistralai/Pixtral-12B-2409", + "providerGroup": "Hyperbolic", + "quantization": "bf16", + "variant": "standard", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "logprobs", + "top_logprobs", + "seed", + "logit_bias", + "top_k", + "min_p", + "repetition_penalty" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "privacyPolicyUrl": "https://hyperbolic.xyz/privacy", + "termsOfServiceUrl": "https://hyperbolic.xyz/terms", + "paidModels": { + "training": false + }, + "training": false + }, + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000001", + "image": "0.0001445", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 149, + "newest": 211, + "throughputHighToLow": 225, + "latencyLowToHigh": 227, + "pricingLowToHigh": 116, + "pricingHighToLow": 204 + }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", + "providers": [ + { + "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", + "slug": "hyperbolic", + "quantization": "bf16", + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "mistralai/Pixtral-12B-2409", + "pricing": { + "prompt": "0.0000001", + "completion": "0.0000001", + "image": "0.0001445", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "logprobs", + "top_logprobs", + "seed", + "logit_bias", + "top_k", + "min_p", + "repetition_penalty" ], - "inputCost": 0.07, - "outputCost": 0.28 + "inputCost": 0.1, + "outputCost": 0.1, + "throughput": 56.568, + "latency": 1575 }, { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 40960, + "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", + "slug": "mistral", + "quantization": "unknown", + "context": 131072, "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen3-14B", + "providerModelId": "pixtral-12b-2409", "pricing": { - "prompt": "0.00000008", - "completion": "0.00000024", - "image": "0", + "prompt": "0.00000015", + "completion": "0.00000015", + "image": "0.0002168", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" + "response_format", + "structured_outputs", + "seed" ], - "inputCost": 0.08, - "outputCost": 0.24 + "inputCost": 0.15, + "outputCost": 0.15, + "throughput": 82.949, + "latency": 392 } ] }, { - "slug": "anthropic/claude-3.5-sonnet-20240620", + "slug": "mistralai/mistral-large", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-06-20T00:00:00+00:00", + "createdAt": "2024-02-26T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3.5 Sonnet (2024-06-20) (self-moderated)", - "shortName": "Claude 3.5 Sonnet (2024-06-20) (self-moderated)", - "author": "anthropic", - "description": "Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\nFor the latest version (2024-10-23), check out [Claude 3.5 Sonnet](/anthropic/claude-3.5-sonnet).\n\n#multimodal", - "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", - "contextLength": 200000, + "name": "Mistral Large", + "shortName": "Mistral Large", + "author": "mistralai", + "description": "This is Mistral AI's flagship model, Mistral Large 2 (version `mistral-large-2407`). It's a proprietary weights-available model and excels at reasoning, code, JSON, chat, and more. Read the launch announcement [here](https://mistral.ai/news/mistral-large-2407/).\n\nIt supports dozens of languages including French, German, Spanish, Italian, Portuguese, Arabic, Hindi, Russian, Chinese, Japanese, and Korean, along with 80+ coding languages including Python, Java, C, C++, JavaScript, and Bash. Its long context window allows precise information recall from large documents.", + "modelVersionGroupId": "83129748-0564-4485-982a-d7a37a1ef3ec", + "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", + "group": "Mistral", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3.5-sonnet-20240620", + "permaslug": "mistralai/mistral-large", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "20a76230-7394-42f9-bcac-76081ff24cd8", - "name": "Anthropic | anthropic/claude-3.5-sonnet-20240620:beta", - "contextLength": 200000, + "id": "f1a57233-f872-4fa0-ad37-66c9a6b00469", + "name": "Mistral | mistralai/mistral-large", + "contextLength": 128000, "model": { - "slug": "anthropic/claude-3.5-sonnet-20240620", + "slug": "mistralai/mistral-large", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-06-20T00:00:00+00:00", + "createdAt": "2024-02-26T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3.5 Sonnet (2024-06-20)", - "shortName": "Claude 3.5 Sonnet (2024-06-20)", - "author": "anthropic", - "description": "Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\nFor the latest version (2024-10-23), check out [Claude 3.5 Sonnet](/anthropic/claude-3.5-sonnet).\n\n#multimodal", - "modelVersionGroupId": "30636d20-cda3-4a59-aa0c-1a5b6efba072", - "contextLength": 200000, + "name": "Mistral Large", + "shortName": "Mistral Large", + "author": "mistralai", + "description": "This is Mistral AI's flagship model, Mistral Large 2 (version `mistral-large-2407`). It's a proprietary weights-available model and excels at reasoning, code, JSON, chat, and more. Read the launch announcement [here](https://mistral.ai/news/mistral-large-2407/).\n\nIt supports dozens of languages including French, German, Spanish, Italian, Portuguese, Arabic, Hindi, Russian, Chinese, Japanese, and Korean, along with 80+ coding languages including Python, Java, C, C++, JavaScript, and Bash. Its long context window allows precise information recall from large documents.", + "modelVersionGroupId": "83129748-0564-4485-982a-d7a37a1ef3ec", + "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", + "group": "Mistral", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3.5-sonnet-20240620", + "permaslug": "mistralai/mistral-large", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "anthropic/claude-3.5-sonnet-20240620:beta", - "modelVariantPermaslug": "anthropic/claude-3.5-sonnet-20240620:beta", - "providerName": "Anthropic", + "modelVariantSlug": "mistralai/mistral-large", + "modelVariantPermaslug": "mistralai/mistral-large", + "providerName": "Mistral", "providerInfo": { - "name": "Anthropic", - "displayName": "Anthropic", - "slug": "anthropic", + "name": "Mistral", + "displayName": "Mistral", + "slug": "mistral", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 - }, - "requiresUserIds": true + } }, - "headquarters": "US", + "headquarters": "FR", "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": true, - "group": "Anthropic", + "hasCompletions": false, + "isAbortable": false, + "moderationRequired": false, + "group": "Mistral", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.anthropic.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Anthropic.svg" + "url": "/images/icons/Mistral.png" } }, - "providerDisplayName": "Anthropic", - "providerModelId": "claude-3-5-sonnet-20240620", - "providerGroup": "Anthropic", + "providerDisplayName": "Mistral", + "providerModelId": "mistral-large-2407", + "providerGroup": "Mistral", "quantization": "unknown", - "variant": "beta", + "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 8192, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -44407,31 +47511,32 @@ export default { "max_tokens", "temperature", "top_p", - "top_k", - "stop" + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "structured_outputs", + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 }, - "requiresUserIds": true, "training": false, "retainsPrompts": true, "retentionDays": 30 }, "pricing": { - "prompt": "0.000003", - "completion": "0.000015", - "image": "0.0048", + "prompt": "0.000002", + "completion": "0.000006", + "image": "0", "request": "0", - "inputCacheRead": "0.0000003", - "inputCacheWrite": "0.00000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -44445,7 +47550,7 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -44453,28 +47558,28 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 143, - "newest": 244, - "throughputHighToLow": 134, - "latencyLowToHigh": 273, - "pricingLowToHigh": 274, - "pricingHighToLow": 46 + "topWeekly": 150, + "newest": 284, + "throughputHighToLow": 199, + "latencyLowToHigh": 68, + "pricingLowToHigh": 246, + "pricingHighToLow": 75 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { - "name": "Anthropic", - "slug": "anthropic", + "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", + "slug": "mistral", "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 8192, - "providerModelId": "claude-3-5-sonnet-20240620", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "mistral-large-2407", "pricing": { - "prompt": "0.000003", - "completion": "0.000015", - "image": "0.0048", + "prompt": "0.000002", + "completion": "0.000006", + "image": "0", "request": "0", - "inputCacheRead": "0.0000003", - "inputCacheWrite": "0.00000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -44485,26 +47590,31 @@ export default { "max_tokens", "temperature", "top_p", - "top_k", - "stop" + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "structured_outputs", + "seed" ], - "inputCost": 3, - "outputCost": 15 + "inputCost": 2, + "outputCost": 6, + "throughput": 43.599, + "latency": 477.5 }, { - "name": "Google Vertex", - "slug": "vertex", + "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", + "slug": "azure", "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 8192, - "providerModelId": "claude-3-5-sonnet@20240620", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "mistral-large", "pricing": { "prompt": "0.000003", - "completion": "0.000015", - "image": "0.0048", + "completion": "0.000009", + "image": "0", "request": "0", - "inputCacheRead": "0.0000003", - "inputCacheWrite": "0.00000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -44515,155 +47625,152 @@ export default { "max_tokens", "temperature", "top_p", - "top_k", - "stop" + "response_format", + "stop", + "seed" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 9, + "throughput": 39.9965, + "latency": 790.5 } ] }, { - "slug": "anthropic/claude-3-haiku", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-03-13T00:00:00+00:00", + "slug": "nvidia/llama-3.3-nemotron-super-49b-v1", + "hfSlug": "nvidia/Llama-3_3-Nemotron-Super-49B-v1", + "updatedAt": "2025-04-08T14:53:49.407401+00:00", + "createdAt": "2025-04-08T13:38:14.544231+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3 Haiku (self-moderated)", - "shortName": "Claude 3 Haiku (self-moderated)", - "author": "anthropic", - "description": "Claude 3 Haiku is Anthropic's fastest and most compact model for\nnear-instant responsiveness. Quick and accurate targeted performance.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-haiku)\n\n#multimodal", - "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", - "contextLength": 200000, + "name": "NVIDIA: Llama 3.3 Nemotron Super 49B v1 (free)", + "shortName": "Llama 3.3 Nemotron Super 49B v1 (free)", + "author": "nvidia", + "description": "Llama-3.3-Nemotron-Super-49B-v1 is a large language model (LLM) optimized for advanced reasoning, conversational interactions, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta's Llama-3.3-70B-Instruct, it employs a Neural Architecture Search (NAS) approach, significantly enhancing efficiency and reducing memory requirements. This allows the model to support a context length of up to 128K tokens and fit efficiently on single high-performance GPUs, such as NVIDIA H200.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.", + "modelVersionGroupId": null, + "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-haiku", + "permaslug": "nvidia/llama-3.3-nemotron-super-49b-v1", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "8661a1db-b0cf-4eb2-ba04-c2a79f698682", - "name": "Anthropic | anthropic/claude-3-haiku:beta", - "contextLength": 200000, + "id": "d4a7aa46-e0f4-47ce-945a-47f2ed715cbb", + "name": "Chutes | nvidia/llama-3.3-nemotron-super-49b-v1:free", + "contextLength": 131072, "model": { - "slug": "anthropic/claude-3-haiku", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-03-13T00:00:00+00:00", + "slug": "nvidia/llama-3.3-nemotron-super-49b-v1", + "hfSlug": "nvidia/Llama-3_3-Nemotron-Super-49B-v1", + "updatedAt": "2025-04-08T14:53:49.407401+00:00", + "createdAt": "2025-04-08T13:38:14.544231+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3 Haiku", - "shortName": "Claude 3 Haiku", - "author": "anthropic", - "description": "Claude 3 Haiku is Anthropic's fastest and most compact model for\nnear-instant responsiveness. Quick and accurate targeted performance.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-haiku)\n\n#multimodal", - "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", - "contextLength": 200000, + "name": "NVIDIA: Llama 3.3 Nemotron Super 49B v1", + "shortName": "Llama 3.3 Nemotron Super 49B v1", + "author": "nvidia", + "description": "Llama-3.3-Nemotron-Super-49B-v1 is a large language model (LLM) optimized for advanced reasoning, conversational interactions, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta's Llama-3.3-70B-Instruct, it employs a Neural Architecture Search (NAS) approach, significantly enhancing efficiency and reducing memory requirements. This allows the model to support a context length of up to 128K tokens and fit efficiently on single high-performance GPUs, such as NVIDIA H200.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.", + "modelVersionGroupId": null, + "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-haiku", + "permaslug": "nvidia/llama-3.3-nemotron-super-49b-v1", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "anthropic/claude-3-haiku:beta", - "modelVariantPermaslug": "anthropic/claude-3-haiku:beta", - "providerName": "Anthropic", + "modelVariantSlug": "nvidia/llama-3.3-nemotron-super-49b-v1:free", + "modelVariantPermaslug": "nvidia/llama-3.3-nemotron-super-49b-v1:free", + "providerName": "Chutes", "providerInfo": { - "name": "Anthropic", - "displayName": "Anthropic", - "slug": "anthropic", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "training": true, + "retainsPrompts": true + } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "Anthropic", + "moderationRequired": false, + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.anthropic.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Anthropic.svg" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "Anthropic", - "providerModelId": "claude-3-haiku-20240307", - "providerGroup": "Anthropic", - "quantization": "unknown", - "variant": "beta", - "isFree": false, + "providerDisplayName": "Chutes", + "providerModelId": "nvidia/Llama-3_3-Nemotron-Super-49B-v1", + "providerGroup": "Chutes", + "quantization": "bf16", + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", "top_k", - "stop" + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.00000025", - "completion": "0.00000125", - "image": "0.0004", + "prompt": "0", + "completion": "0", + "image": "0", "request": "0", - "inputCacheRead": "0.00000003", - "inputCacheWrite": "0.0000003", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -44672,7 +47779,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -44680,93 +47787,70 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 58, - "newest": 277, - "throughputHighToLow": 36, - "latencyLowToHigh": 134, - "pricingLowToHigh": 168, - "pricingHighToLow": 149 + "topWeekly": 151, + "newest": 58, + "throughputHighToLow": 172, + "latencyLowToHigh": 20, + "pricingLowToHigh": 24, + "pricingHighToLow": 182 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nvidia.com/\\u0026size=256", "providers": [ { - "name": "Anthropic", - "slug": "anthropic", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 4096, - "providerModelId": "claude-3-haiku-20240307", - "pricing": { - "prompt": "0.00000025", - "completion": "0.00000125", - "image": "0.0004", - "request": "0", - "inputCacheRead": "0.00000003", - "inputCacheWrite": "0.0000003", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "top_k", - "stop" - ], - "inputCost": 0.25, - "outputCost": 1.25 - }, - { - "name": "Google Vertex", - "slug": "vertex", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 4096, - "providerModelId": "claude-3-haiku@20240307", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "nvidia/Llama-3_3-Nemotron-Super-49B-v1", "pricing": { - "prompt": "0.00000025", - "completion": "0.00000125", - "image": "0.0004", + "prompt": "0.00000013", + "completion": "0.0000004", + "image": "0", "request": "0", - "inputCacheRead": "0.00000003", - "inputCacheWrite": "0.0000003", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", "top_k", - "stop" + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 0.25, - "outputCost": 1.25 + "inputCost": 0.13, + "outputCost": 0.4, + "throughput": 47.2755, + "latency": 326 } ] }, { - "slug": "mistralai/mistral-large", - "hfSlug": null, + "slug": "qwen/qwen-max", + "hfSlug": "", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-02-26T00:00:00+00:00", + "createdAt": "2025-02-01T09:31:29.206961+00:00", "hfUpdatedAt": null, - "name": "Mistral Large", - "shortName": "Mistral Large", - "author": "mistralai", - "description": "This is Mistral AI's flagship model, Mistral Large 2 (version `mistral-large-2407`). It's a proprietary weights-available model and excels at reasoning, code, JSON, chat, and more. Read the launch announcement [here](https://mistral.ai/news/mistral-large-2407/).\n\nIt supports dozens of languages including French, German, Spanish, Italian, Portuguese, Arabic, Hindi, Russian, Chinese, Japanese, and Korean, along with 80+ coding languages including Python, Java, C, C++, JavaScript, and Bash. Its long context window allows precise information recall from large documents.", - "modelVersionGroupId": "83129748-0564-4485-982a-d7a37a1ef3ec", - "contextLength": 128000, + "name": "Qwen: Qwen-Max ", + "shortName": "Qwen-Max ", + "author": "qwen", + "description": "Qwen-Max, based on Qwen2.5, provides the best inference performance among [Qwen models](/qwen), especially for complex multi-step tasks. It's a large-scale MoE model that has been pretrained on over 20 trillion tokens and further post-trained with curated Supervised Fine-Tuning (SFT) and Reinforcement Learning from Human Feedback (RLHF) methodologies. The parameter count is unknown.", + "modelVersionGroupId": null, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -44774,32 +47858,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "Qwen", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-large", + "permaslug": "qwen/qwen-max-2025-01-25", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "f1a57233-f872-4fa0-ad37-66c9a6b00469", - "name": "Mistral | mistralai/mistral-large", - "contextLength": 128000, + "id": "8cf90481-400f-473d-b188-1bed01001a01", + "name": "Alibaba | qwen/qwen-max-2025-01-25", + "contextLength": 32768, "model": { - "slug": "mistralai/mistral-large", - "hfSlug": null, + "slug": "qwen/qwen-max", + "hfSlug": "", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-02-26T00:00:00+00:00", + "createdAt": "2025-02-01T09:31:29.206961+00:00", "hfUpdatedAt": null, - "name": "Mistral Large", - "shortName": "Mistral Large", - "author": "mistralai", - "description": "This is Mistral AI's flagship model, Mistral Large 2 (version `mistral-large-2407`). It's a proprietary weights-available model and excels at reasoning, code, JSON, chat, and more. Read the launch announcement [here](https://mistral.ai/news/mistral-large-2407/).\n\nIt supports dozens of languages including French, German, Spanish, Italian, Portuguese, Arabic, Hindi, Russian, Chinese, Japanese, and Korean, along with 80+ coding languages including Python, Java, C, C++, JavaScript, and Bash. Its long context window allows precise information recall from large documents.", - "modelVersionGroupId": "83129748-0564-4485-982a-d7a37a1ef3ec", - "contextLength": 128000, + "name": "Qwen: Qwen-Max ", + "shortName": "Qwen-Max ", + "author": "qwen", + "description": "Qwen-Max, based on Qwen2.5, provides the best inference performance among [Qwen models](/qwen), especially for complex multi-step tasks. It's a large-scale MoE model that has been pretrained on over 20 trillion tokens and further post-trained with curated Supervised Fine-Tuning (SFT) and Reinforcement Learning from Human Feedback (RLHF) methodologies. The parameter count is unknown.", + "modelVersionGroupId": null, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -44807,40 +47891,38 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "Qwen", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-large", + "permaslug": "qwen/qwen-max-2025-01-25", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "mistralai/mistral-large", - "modelVariantPermaslug": "mistralai/mistral-large", - "providerName": "Mistral", + "modelVariantSlug": "qwen/qwen-max", + "modelVariantPermaslug": "qwen/qwen-max-2025-01-25", + "providerName": "Alibaba", "providerInfo": { - "name": "Mistral", - "displayName": "Mistral", - "slug": "mistral", + "name": "Alibaba", + "displayName": "Alibaba", + "slug": "alibaba", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", + "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false } }, - "headquarters": "FR", + "headquarters": "CN", "hasChatCompletions": true, - "hasCompletions": false, + "hasCompletions": true, "isAbortable": false, "moderationRequired": false, - "group": "Mistral", + "group": "Alibaba", "editors": [], "owners": [], "isMultipartSupported": true, @@ -44848,18 +47930,18 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Mistral.png" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256" } }, - "providerDisplayName": "Mistral", - "providerModelId": "mistral-large-2407", - "providerGroup": "Mistral", - "quantization": "unknown", + "providerDisplayName": "Alibaba", + "providerModelId": "qwen-max", + "providerGroup": "Alibaba", + "quantization": null, "variant": "standard", "isFree": false, "canAbort": false, - "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxPromptTokens": 30720, + "maxCompletionTokens": 8192, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -44868,30 +47950,23 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", + "seed", "response_format", - "structured_outputs", - "seed" + "presence_penalty" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", + "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, "pricing": { - "prompt": "0.000002", - "completion": "0.000006", + "prompt": "0.0000016", + "completion": "0.0000064", "image": "0", "request": "0", "webSearch": "0", @@ -44907,7 +47982,7 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -44915,56 +47990,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 148, - "newest": 284, - "throughputHighToLow": 188, - "latencyLowToHigh": 57, - "pricingLowToHigh": 246, - "pricingHighToLow": 75 + "topWeekly": 152, + "newest": 128, + "throughputHighToLow": 212, + "latencyLowToHigh": 214, + "pricingLowToHigh": 241, + "pricingHighToLow": 77 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { - "name": "Mistral", - "slug": "mistral", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "mistral-large-2407", - "pricing": { - "prompt": "0.000002", - "completion": "0.000006", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "structured_outputs", - "seed" - ], - "inputCost": 2, - "outputCost": 6 - }, - { - "name": "Azure", - "slug": "azure", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "mistral-large", + "name": "Alibaba", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256", + "slug": "alibaba", + "quantization": null, + "context": 32768, + "maxCompletionTokens": 8192, + "providerModelId": "qwen-max", "pricing": { - "prompt": "0.000003", - "completion": "0.000009", + "prompt": "0.0000016", + "completion": "0.0000064", "image": "0", "request": "0", "webSearch": "0", @@ -44977,12 +48022,14 @@ export default { "max_tokens", "temperature", "top_p", + "seed", "response_format", - "stop", - "seed" + "presence_penalty" ], - "inputCost": 3, - "outputCost": 9 + "inputCost": 1.6, + "outputCost": 6.4, + "throughput": 40.26, + "latency": 1445 } ] }, @@ -45164,16 +48211,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 149, + "topWeekly": 153, "newest": 10, - "throughputHighToLow": 30, - "latencyLowToHigh": 122, + "throughputHighToLow": 26, + "latencyLowToHigh": 120, "pricingLowToHigh": 1, "pricingHighToLow": 206 }, + "authorIcon": "https://openrouter.ai/images/icons/Microsoft.svg", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "bf16", "context": 32768, @@ -45204,22 +48253,24 @@ export default { "min_p" ], "inputCost": 0.07, - "outputCost": 0.35 + "outputCost": 0.35, + "throughput": 46.1485, + "latency": 714 } ] }, { - "slug": "nvidia/llama-3.3-nemotron-super-49b-v1", - "hfSlug": "nvidia/Llama-3_3-Nemotron-Super-49B-v1", - "updatedAt": "2025-04-08T14:53:49.407401+00:00", - "createdAt": "2025-04-08T13:38:14.544231+00:00", + "slug": "qwen/qwen3-4b", + "hfSlug": "Qwen/Qwen3-4B", + "updatedAt": "2025-05-12T00:35:27.640755+00:00", + "createdAt": "2025-04-30T16:38:24.032465+00:00", "hfUpdatedAt": null, - "name": "NVIDIA: Llama 3.3 Nemotron Super 49B v1 (free)", - "shortName": "Llama 3.3 Nemotron Super 49B v1 (free)", - "author": "nvidia", - "description": "Llama-3.3-Nemotron-Super-49B-v1 is a large language model (LLM) optimized for advanced reasoning, conversational interactions, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta's Llama-3.3-70B-Instruct, it employs a Neural Architecture Search (NAS) approach, significantly enhancing efficiency and reducing memory requirements. This allows the model to support a context length of up to 128K tokens and fit efficiently on single high-performance GPUs, such as NVIDIA H200.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.", + "name": "Qwen: Qwen3 4B (free)", + "shortName": "Qwen3 4B (free)", + "author": "qwen", + "description": "Qwen3-4B is a 4 billion parameter dense language model from the Qwen3 series, designed to support both general-purpose and reasoning-intensive tasks. It introduces a dual-mode architecture—thinking and non-thinking—allowing dynamic switching between high-precision logical reasoning and efficient dialogue generation. This makes it well-suited for multi-turn chat, instruction following, and complex agent workflows.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -45227,32 +48278,43 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "nvidia/llama-3.3-nemotron-super-49b-v1", - "reasoningConfig": null, - "features": {}, + "permaslug": "qwen/qwen3-4b-04-28", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "d4a7aa46-e0f4-47ce-945a-47f2ed715cbb", - "name": "Chutes | nvidia/llama-3.3-nemotron-super-49b-v1:free", - "contextLength": 131072, + "id": "b1d65b01-eded-4bfd-92ac-0deaac309a5e", + "name": "Novita | qwen/qwen3-4b-04-28:free", + "contextLength": 128000, "model": { - "slug": "nvidia/llama-3.3-nemotron-super-49b-v1", - "hfSlug": "nvidia/Llama-3_3-Nemotron-Super-49B-v1", - "updatedAt": "2025-04-08T14:53:49.407401+00:00", - "createdAt": "2025-04-08T13:38:14.544231+00:00", + "slug": "qwen/qwen3-4b", + "hfSlug": "Qwen/Qwen3-4B", + "updatedAt": "2025-05-12T00:35:27.640755+00:00", + "createdAt": "2025-04-30T16:38:24.032465+00:00", "hfUpdatedAt": null, - "name": "NVIDIA: Llama 3.3 Nemotron Super 49B v1", - "shortName": "Llama 3.3 Nemotron Super 49B v1", - "author": "nvidia", - "description": "Llama-3.3-Nemotron-Super-49B-v1 is a large language model (LLM) optimized for advanced reasoning, conversational interactions, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta's Llama-3.3-70B-Instruct, it employs a Neural Architecture Search (NAS) approach, significantly enhancing efficiency and reducing memory requirements. This allows the model to support a context length of up to 128K tokens and fit efficiently on single high-performance GPUs, such as NVIDIA H200.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.", + "name": "Qwen: Qwen3 4B", + "shortName": "Qwen3 4B", + "author": "qwen", + "description": "Qwen3-4B is a 4 billion parameter dense language model from the Qwen3 series, designed to support both general-purpose and reasoning-intensive tasks. It introduces a dual-mode architecture—thinking and non-thinking—allowing dynamic switching between high-precision logical reasoning and efficient dialogue generation. This makes it well-suited for multi-turn chat, instruction following, and complex agent workflows.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -45260,51 +48322,64 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "nvidia/llama-3.3-nemotron-super-49b-v1", - "reasoningConfig": null, - "features": {} + "permaslug": "qwen/qwen3-4b-04-28", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "nvidia/llama-3.3-nemotron-super-49b-v1:free", - "modelVariantPermaslug": "nvidia/llama-3.3-nemotron-super-49b-v1:free", - "providerName": "Chutes", + "modelVariantSlug": "qwen/qwen3-4b:free", + "modelVariantPermaslug": "qwen/qwen3-4b-04-28:free", + "providerName": "Novita", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "Novita", + "displayName": "NovitaAI", + "slug": "novita", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false } }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Chutes", + "group": "Novita", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.novita.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "invertRequired": true } }, - "providerDisplayName": "Chutes", - "providerModelId": "nvidia/Llama-3_3-Nemotron-Super-49B-v1", - "providerGroup": "Chutes", - "quantization": "bf16", + "providerDisplayName": "NovitaAI", + "providerModelId": "qwen/qwen3-4b-fp8", + "providerGroup": "Novita", + "quantization": "fp8", "variant": "free", "isFree": true, "canAbort": true, @@ -45316,6 +48391,8 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -45323,20 +48400,17 @@ export default { "top_k", "min_p", "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" + "logit_bias" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false }, - "training": true, - "retainsPrompts": true + "training": false }, "pricing": { "prompt": "0", @@ -45352,7 +48426,7 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, @@ -45365,60 +48439,28 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 150, - "newest": 58, - "throughputHighToLow": 181, - "latencyLowToHigh": 21, - "pricingLowToHigh": 24, - "pricingHighToLow": 179 + "topWeekly": 154, + "newest": 16, + "throughputHighToLow": 34, + "latencyLowToHigh": 103, + "pricingLowToHigh": 5, + "pricingHighToLow": 253 }, - "providers": [ - { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "nvidia/Llama-3_3-Nemotron-Super-49B-v1", - "pricing": { - "prompt": "0.00000013", - "completion": "0.0000004", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" - ], - "inputCost": 0.13, - "outputCost": 0.4 - } - ] + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", + "providers": [] }, { - "slug": "qwen/qwen3-4b", - "hfSlug": "Qwen/Qwen3-4B", - "updatedAt": "2025-05-12T00:35:27.640755+00:00", - "createdAt": "2025-04-30T16:38:24.032465+00:00", + "slug": "agentica-org/deepcoder-14b-preview", + "hfSlug": "agentica-org/DeepCoder-14B-Preview", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-04-13T14:43:15.533943+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 4B (free)", - "shortName": "Qwen3 4B (free)", - "author": "qwen", - "description": "Qwen3-4B is a 4 billion parameter dense language model from the Qwen3 series, designed to support both general-purpose and reasoning-intensive tasks. It introduces a dual-mode architecture—thinking and non-thinking—allowing dynamic switching between high-precision logical reasoning and efficient dialogue generation. This makes it well-suited for multi-turn chat, instruction following, and complex agent workflows.", + "name": "Agentica: Deepcoder 14B Preview (free)", + "shortName": "Deepcoder 14B Preview (free)", + "author": "agentica-org", + "description": "DeepCoder-14B-Preview is a 14B parameter code generation model fine-tuned from DeepSeek-R1-Distill-Qwen-14B using reinforcement learning with GRPO+ and iterative context lengthening. It is optimized for long-context program synthesis and achieves strong performance across coding benchmarks, including 60.6% on LiveCodeBench v5, competitive with models like o3-Mini", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 96000, "inputModalities": [ "text" ], @@ -45426,17 +48468,17 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Other", + "instructType": "deepseek-r1", "defaultSystem": null, "defaultStops": [ - "<|im_start|>", - "<|im_end|>" + "<｜User｜>", + "<｜end▁of▁sentence｜>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen3-4b-04-28", + "permaslug": "agentica-org/deepcoder-14b-preview", "reasoningConfig": { "startToken": "", "endToken": "" @@ -45448,21 +48490,21 @@ export default { } }, "endpoint": { - "id": "b1d65b01-eded-4bfd-92ac-0deaac309a5e", - "name": "Novita | qwen/qwen3-4b-04-28:free", - "contextLength": 128000, + "id": "491302ef-8014-45e5-9c32-5c9a02ebc055", + "name": "Chutes | agentica-org/deepcoder-14b-preview:free", + "contextLength": 96000, "model": { - "slug": "qwen/qwen3-4b", - "hfSlug": "Qwen/Qwen3-4B", - "updatedAt": "2025-05-12T00:35:27.640755+00:00", - "createdAt": "2025-04-30T16:38:24.032465+00:00", + "slug": "agentica-org/deepcoder-14b-preview", + "hfSlug": "agentica-org/DeepCoder-14B-Preview", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-04-13T14:43:15.533943+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 4B", - "shortName": "Qwen3 4B", - "author": "qwen", - "description": "Qwen3-4B is a 4 billion parameter dense language model from the Qwen3 series, designed to support both general-purpose and reasoning-intensive tasks. It introduces a dual-mode architecture—thinking and non-thinking—allowing dynamic switching between high-precision logical reasoning and efficient dialogue generation. This makes it well-suited for multi-turn chat, instruction following, and complex agent workflows.", + "name": "Agentica: Deepcoder 14B Preview", + "shortName": "Deepcoder 14B Preview", + "author": "agentica-org", + "description": "DeepCoder-14B-Preview is a 14B parameter code generation model fine-tuned from DeepSeek-R1-Distill-Qwen-14B using reinforcement learning with GRPO+ and iterative context lengthening. It is optimized for long-context program synthesis and achieves strong performance across coding benchmarks, including 60.6% on LiveCodeBench v5, competitive with models like o3-Mini", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 96000, "inputModalities": [ "text" ], @@ -45470,17 +48512,17 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Other", + "instructType": "deepseek-r1", "defaultSystem": null, "defaultStops": [ - "<|im_start|>", - "<|im_end|>" + "<｜User｜>", + "<｜end▁of▁sentence｜>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen3-4b-04-28", + "permaslug": "agentica-org/deepcoder-14b-preview", "reasoningConfig": { "startToken": "", "endToken": "" @@ -45492,42 +48534,40 @@ export default { } } }, - "modelVariantSlug": "qwen/qwen3-4b:free", - "modelVariantPermaslug": "qwen/qwen3-4b-04-28:free", - "providerName": "Novita", + "modelVariantSlug": "agentica-org/deepcoder-14b-preview:free", + "modelVariantPermaslug": "agentica-org/deepcoder-14b-preview:free", + "providerName": "Chutes", "providerInfo": { - "name": "Novita", - "displayName": "NovitaAI", - "slug": "novita", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Novita", + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.novita.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "NovitaAI", - "providerModelId": "qwen/qwen3-4b-fp8", - "providerGroup": "Novita", - "quantization": "fp8", + "providerDisplayName": "Chutes", + "providerModelId": "agentica-org/DeepCoder-14B-Preview", + "providerGroup": "Chutes", + "quantization": null, "variant": "free", "isFree": true, "canAbort": true, @@ -45548,17 +48588,20 @@ export default { "top_k", "min_p", "repetition_penalty", - "logit_bias" + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true }, - "training": false + "training": true, + "retainsPrompts": true }, "pricing": { "prompt": "0", @@ -45587,27 +48630,28 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 151, - "newest": 16, - "throughputHighToLow": 94, - "latencyLowToHigh": 140, - "pricingLowToHigh": 5, - "pricingHighToLow": 253 + "topWeekly": 155, + "newest": 54, + "throughputHighToLow": 100, + "latencyLowToHigh": 234, + "pricingLowToHigh": 22, + "pricingHighToLow": 270 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { - "slug": "openai/gpt-4.5-preview", + "slug": "perplexity/sonar", "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-27T20:23:30.841555+00:00", + "updatedAt": "2025-05-05T14:39:01.632671+00:00", + "createdAt": "2025-01-27T21:36:48.666939+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4.5 (Preview)", - "shortName": "GPT-4.5 (Preview)", - "author": "openai", - "description": "GPT-4.5 (Preview) is a research preview of OpenAI’s latest language model, designed to advance capabilities in reasoning, creativity, and multi-turn conversation. It builds on previous iterations with improvements in world knowledge, contextual coherence, and the ability to follow user intent more effectively.\n\nThe model demonstrates enhanced performance in tasks that require open-ended thinking, problem-solving, and communication. Early testing suggests it is better at generating nuanced responses, maintaining long-context coherence, and reducing hallucinations compared to earlier versions.\n\nThis research preview is intended to help evaluate GPT-4.5’s strengths and limitations in real-world use cases as OpenAI continues to refine and develop future models. Read more at the [blog post here.](https://openai.com/index/introducing-gpt-4-5/)", + "name": "Perplexity: Sonar", + "shortName": "Sonar", + "author": "perplexity", + "description": "Sonar is lightweight, affordable, fast, and simple to use — now featuring citations and the ability to customize sources. It is designed for companies seeking to integrate lightweight question-and-answer features optimized for speed.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 127072, "inputModalities": [ "text", "image" @@ -45616,32 +48660,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "openai/gpt-4.5-preview-2025-02-27", + "warningMessage": "", + "permaslug": "perplexity/sonar", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "dc592ff9-2fc9-4a5e-b910-27a57bb7bfdd", - "name": "OpenAI | openai/gpt-4.5-preview-2025-02-27", - "contextLength": 128000, + "id": "5f831e7a-c555-4d3a-b228-88286347558a", + "name": "Perplexity | perplexity/sonar", + "contextLength": 127072, "model": { - "slug": "openai/gpt-4.5-preview", + "slug": "perplexity/sonar", "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-27T20:23:30.841555+00:00", + "updatedAt": "2025-05-05T14:39:01.632671+00:00", + "createdAt": "2025-01-27T21:36:48.666939+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4.5 (Preview)", - "shortName": "GPT-4.5 (Preview)", - "author": "openai", - "description": "GPT-4.5 (Preview) is a research preview of OpenAI’s latest language model, designed to advance capabilities in reasoning, creativity, and multi-turn conversation. It builds on previous iterations with improvements in world knowledge, contextual coherence, and the ability to follow user intent more effectively.\n\nThe model demonstrates enhanced performance in tasks that require open-ended thinking, problem-solving, and communication. Early testing suggests it is better at generating nuanced responses, maintaining long-context coherence, and reducing hallucinations compared to earlier versions.\n\nThis research preview is intended to help evaluate GPT-4.5’s strengths and limitations in real-world use cases as OpenAI continues to refine and develop future models. Read more at the [blog post here.](https://openai.com/index/introducing-gpt-4-5/)", + "name": "Perplexity: Sonar", + "shortName": "Sonar", + "author": "perplexity", + "description": "Sonar is lightweight, affordable, fast, and simple to use — now featuring citations and the ability to customize sources. It is designed for companies seeking to integrate lightweight question-and-answer features optimized for speed.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 127072, "inputModalities": [ "text", "image" @@ -45650,116 +48694,113 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "openai/gpt-4.5-preview-2025-02-27", + "warningMessage": "", + "permaslug": "perplexity/sonar", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/gpt-4.5-preview", - "modelVariantPermaslug": "openai/gpt-4.5-preview-2025-02-27", - "providerName": "OpenAI", + "modelVariantSlug": "perplexity/sonar", + "modelVariantPermaslug": "perplexity/sonar", + "providerName": "Perplexity", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Perplexity", + "displayName": "Perplexity", + "slug": "perplexity", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", + "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "training": false + } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", + "hasCompletions": false, + "isAbortable": false, + "moderationRequired": false, + "group": "Perplexity", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": "https://status.perplexity.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true + "url": "/images/icons/Perplexity.svg" } }, - "providerDisplayName": "OpenAI", - "providerModelId": "gpt-4.5-preview-2025-02-27", - "providerGroup": "OpenAI", - "quantization": null, + "providerDisplayName": "Perplexity", + "providerModelId": "sonar", + "providerGroup": "Perplexity", + "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 16384, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "stop", + "web_search_options", + "top_k", "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" + "presence_penalty" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", + "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, "pricing": { - "prompt": "0.000075", - "completion": "0.00015", - "image": "0.108375", - "request": "0", - "inputCacheRead": "0.0000375", + "prompt": "0.000001", + "completion": "0.000001", + "image": "0", + "request": "0.005", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [], + "variablePricings": [ + { + "type": "search-threshold", + "threshold": "high", + "request": "0.012" + }, + { + "type": "search-threshold", + "threshold": "medium", + "request": "0.008" + }, + { + "type": "search-threshold", + "threshold": "low", + "request": "0.005" + } + ], "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -45767,64 +48808,60 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 152, - "newest": 106, - "throughputHighToLow": 256, - "latencyLowToHigh": 178, - "pricingLowToHigh": 317, - "pricingHighToLow": 1 + "topWeekly": 156, + "newest": 138, + "throughputHighToLow": 62, + "latencyLowToHigh": 252, + "pricingLowToHigh": 287, + "pricingHighToLow": 30 }, + "authorIcon": "https://openrouter.ai/images/icons/Perplexity.svg", "providers": [ { - "name": "OpenAI", - "slug": "openAi", - "quantization": null, - "context": 128000, - "maxCompletionTokens": 16384, - "providerModelId": "gpt-4.5-preview-2025-02-27", + "name": "Perplexity", + "icon": "https://openrouter.ai/images/icons/Perplexity.svg", + "slug": "perplexity", + "quantization": "unknown", + "context": 127072, + "maxCompletionTokens": null, + "providerModelId": "sonar", "pricing": { - "prompt": "0.000075", - "completion": "0.00015", - "image": "0.108375", - "request": "0", - "inputCacheRead": "0.0000375", + "prompt": "0.000001", + "completion": "0.000001", + "image": "0", + "request": "0.005", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "stop", + "web_search_options", + "top_k", "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" + "presence_penalty" ], - "inputCost": 75, - "outputCost": 150 + "inputCost": 1, + "outputCost": 1, + "throughput": 108.181, + "latency": 1948 } ] }, { - "slug": "agentica-org/deepcoder-14b-preview", - "hfSlug": "agentica-org/DeepCoder-14B-Preview", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-04-13T14:43:15.533943+00:00", + "slug": "microsoft/phi-4-reasoning-plus", + "hfSlug": "microsoft/Phi-4-reasoning-plus", + "updatedAt": "2025-05-02T05:32:14.23708+00:00", + "createdAt": "2025-05-01T20:22:41.235613+00:00", "hfUpdatedAt": null, - "name": "Agentica: Deepcoder 14B Preview (free)", - "shortName": "Deepcoder 14B Preview (free)", - "author": "agentica-org", - "description": "DeepCoder-14B-Preview is a 14B parameter code generation model fine-tuned from DeepSeek-R1-Distill-Qwen-14B using reinforcement learning with GRPO+ and iterative context lengthening. It is optimized for long-context program synthesis and achieves strong performance across coding benchmarks, including 60.6% on LiveCodeBench v5, competitive with models like o3-Mini", + "name": "Microsoft: Phi 4 Reasoning Plus (free)", + "shortName": "Phi 4 Reasoning Plus (free)", + "author": "microsoft", + "description": "Phi-4-reasoning-plus is an enhanced 14B parameter model from Microsoft, fine-tuned from Phi-4 with additional reinforcement learning to boost accuracy on math, science, and code reasoning tasks. It uses the same dense decoder-only transformer architecture as Phi-4, but generates longer, more comprehensive outputs structured into a step-by-step reasoning trace and final answer.\n\nWhile it offers improved benchmark scores over Phi-4-reasoning across tasks like AIME, OmniMath, and HumanEvalPlus, its responses are typically ~50% longer, resulting in higher latency. Designed for English-only applications, it is well-suited for structured reasoning workflows where output quality takes priority over response speed.", "modelVersionGroupId": null, - "contextLength": 96000, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -45833,16 +48870,13 @@ export default { ], "hasTextOutput": true, "group": "Other", - "instructType": "deepseek-r1", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "agentica-org/deepcoder-14b-preview", + "permaslug": "microsoft/phi-4-reasoning-plus-04-30", "reasoningConfig": { "startToken": "", "endToken": "" @@ -45854,21 +48888,21 @@ export default { } }, "endpoint": { - "id": "491302ef-8014-45e5-9c32-5c9a02ebc055", - "name": "Chutes | agentica-org/deepcoder-14b-preview:free", - "contextLength": 96000, + "id": "bbc205d3-b5e1-457f-8b74-771855206a98", + "name": "Chutes | microsoft/phi-4-reasoning-plus-04-30:free", + "contextLength": 32768, "model": { - "slug": "agentica-org/deepcoder-14b-preview", - "hfSlug": "agentica-org/DeepCoder-14B-Preview", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-04-13T14:43:15.533943+00:00", + "slug": "microsoft/phi-4-reasoning-plus", + "hfSlug": "microsoft/Phi-4-reasoning-plus", + "updatedAt": "2025-05-02T05:32:14.23708+00:00", + "createdAt": "2025-05-01T20:22:41.235613+00:00", "hfUpdatedAt": null, - "name": "Agentica: Deepcoder 14B Preview", - "shortName": "Deepcoder 14B Preview", - "author": "agentica-org", - "description": "DeepCoder-14B-Preview is a 14B parameter code generation model fine-tuned from DeepSeek-R1-Distill-Qwen-14B using reinforcement learning with GRPO+ and iterative context lengthening. It is optimized for long-context program synthesis and achieves strong performance across coding benchmarks, including 60.6% on LiveCodeBench v5, competitive with models like o3-Mini", + "name": "Microsoft: Phi 4 Reasoning Plus", + "shortName": "Phi 4 Reasoning Plus", + "author": "microsoft", + "description": "Phi-4-reasoning-plus is an enhanced 14B parameter model from Microsoft, fine-tuned from Phi-4 with additional reinforcement learning to boost accuracy on math, science, and code reasoning tasks. It uses the same dense decoder-only transformer architecture as Phi-4, but generates longer, more comprehensive outputs structured into a step-by-step reasoning trace and final answer.\n\nWhile it offers improved benchmark scores over Phi-4-reasoning across tasks like AIME, OmniMath, and HumanEvalPlus, its responses are typically ~50% longer, resulting in higher latency. Designed for English-only applications, it is well-suited for structured reasoning workflows where output quality takes priority over response speed.", "modelVersionGroupId": null, - "contextLength": 96000, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -45877,16 +48911,13 @@ export default { ], "hasTextOutput": true, "group": "Other", - "instructType": "deepseek-r1", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "agentica-org/deepcoder-14b-preview", + "permaslug": "microsoft/phi-4-reasoning-plus-04-30", "reasoningConfig": { "startToken": "", "endToken": "" @@ -45898,8 +48929,8 @@ export default { } } }, - "modelVariantSlug": "agentica-org/deepcoder-14b-preview:free", - "modelVariantPermaslug": "agentica-org/deepcoder-14b-preview:free", + "modelVariantSlug": "microsoft/phi-4-reasoning-plus:free", + "modelVariantPermaslug": "microsoft/phi-4-reasoning-plus-04-30:free", "providerName": "Chutes", "providerInfo": { "name": "Chutes", @@ -45929,7 +48960,7 @@ export default { } }, "providerDisplayName": "Chutes", - "providerModelId": "agentica-org/DeepCoder-14B-Preview", + "providerModelId": "microsoft/Phi-4-reasoning-plus", "providerGroup": "Chutes", "quantization": null, "variant": "free", @@ -45995,117 +49026,163 @@ export default { }, "sorting": { "topWeekly": 153, - "newest": 54, - "throughputHighToLow": 101, - "latencyLowToHigh": 225, - "pricingLowToHigh": 22, - "pricingHighToLow": 270 + "newest": 10, + "throughputHighToLow": 26, + "latencyLowToHigh": 120, + "pricingLowToHigh": 1, + "pricingHighToLow": 206 }, - "providers": [] + "authorIcon": "https://openrouter.ai/images/icons/Microsoft.svg", + "providers": [ + { + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "microsoft/phi-4-reasoning-plus", + "pricing": { + "prompt": "0.00000007", + "completion": "0.00000035", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" + ], + "inputCost": 0.07, + "outputCost": 0.35, + "throughput": 46.1485, + "latency": 714 + } + ] }, { - "slug": "qwen/qwen-max", - "hfSlug": "", + "slug": "openai/gpt-4-turbo", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-01T09:31:29.206961+00:00", + "createdAt": "2024-04-09T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen-Max ", - "shortName": "Qwen-Max ", - "author": "qwen", - "description": "Qwen-Max, based on Qwen2.5, provides the best inference performance among [Qwen models](/qwen), especially for complex multi-step tasks. It's a large-scale MoE model that has been pretrained on over 20 trillion tokens and further post-trained with curated Supervised Fine-Tuning (SFT) and Reinforcement Learning from Human Feedback (RLHF) methodologies. The parameter count is unknown.", + "name": "OpenAI: GPT-4 Turbo", + "shortName": "GPT-4 Turbo", + "author": "openai", + "description": "The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.\n\nTraining data: up to December 2023.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 128000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-max-2025-01-25", + "permaslug": "openai/gpt-4-turbo", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "8cf90481-400f-473d-b188-1bed01001a01", - "name": "Alibaba | qwen/qwen-max-2025-01-25", - "contextLength": 32768, + "id": "da16824f-3ba0-43a1-86f8-a6131837f457", + "name": "OpenAI | openai/gpt-4-turbo", + "contextLength": 128000, "model": { - "slug": "qwen/qwen-max", - "hfSlug": "", + "slug": "openai/gpt-4-turbo", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-01T09:31:29.206961+00:00", + "createdAt": "2024-04-09T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen-Max ", - "shortName": "Qwen-Max ", - "author": "qwen", - "description": "Qwen-Max, based on Qwen2.5, provides the best inference performance among [Qwen models](/qwen), especially for complex multi-step tasks. It's a large-scale MoE model that has been pretrained on over 20 trillion tokens and further post-trained with curated Supervised Fine-Tuning (SFT) and Reinforcement Learning from Human Feedback (RLHF) methodologies. The parameter count is unknown.", + "name": "OpenAI: GPT-4 Turbo", + "shortName": "GPT-4 Turbo", + "author": "openai", + "description": "The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.\n\nTraining data: up to December 2023.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 128000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-max-2025-01-25", + "permaslug": "openai/gpt-4-turbo", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "qwen/qwen-max", - "modelVariantPermaslug": "qwen/qwen-max-2025-01-25", - "providerName": "Alibaba", + "modelVariantSlug": "openai/gpt-4-turbo", + "modelVariantPermaslug": "openai/gpt-4-turbo", + "providerName": "OpenAI", "providerInfo": { - "name": "Alibaba", - "displayName": "Alibaba", - "slug": "alibaba", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", - "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { - "training": false - } + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, - "headquarters": "CN", + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, - "moderationRequired": false, - "group": "Alibaba", + "isAbortable": true, + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Alibaba", - "providerModelId": "qwen-max", - "providerGroup": "Alibaba", - "quantization": null, + "providerDisplayName": "OpenAI", + "providerModelId": "gpt-4-turbo", + "providerGroup": "OpenAI", + "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": false, - "maxPromptTokens": 30720, - "maxCompletionTokens": 8192, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -46114,24 +49191,35 @@ export default { "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "seed", - "response_format", - "presence_penalty" + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", - "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { - "training": false + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "training": false + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.0000016", - "completion": "0.0000064", - "image": "0", + "prompt": "0.00001", + "completion": "0.00003", + "image": "0.01445", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -46154,25 +49242,27 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 154, - "newest": 128, - "throughputHighToLow": 213, - "latencyLowToHigh": 175, - "pricingLowToHigh": 241, - "pricingHighToLow": 77 + "topWeekly": 158, + "newest": 271, + "throughputHighToLow": 229, + "latencyLowToHigh": 136, + "pricingLowToHigh": 302, + "pricingHighToLow": 14 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "Alibaba", - "slug": "alibaba", - "quantization": null, - "context": 32768, - "maxCompletionTokens": 8192, - "providerModelId": "qwen-max", + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": "unknown", + "context": 128000, + "maxCompletionTokens": 4096, + "providerModelId": "gpt-4-turbo", "pricing": { - "prompt": "0.0000016", - "completion": "0.0000064", - "image": "0", + "prompt": "0.00001", + "completion": "0.00003", + "image": "0.01445", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -46184,27 +49274,34 @@ export default { "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "seed", - "response_format", - "presence_penalty" + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" ], - "inputCost": 1.6, - "outputCost": 6.4 + "inputCost": 10, + "outputCost": 30, + "throughput": 31.894, + "latency": 842 } ] }, { - "slug": "perplexity/sonar", + "slug": "openai/gpt-4.5-preview", "hfSlug": "", - "updatedAt": "2025-05-05T14:39:01.632671+00:00", - "createdAt": "2025-01-27T21:36:48.666939+00:00", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-27T20:23:30.841555+00:00", "hfUpdatedAt": null, - "name": "Perplexity: Sonar", - "shortName": "Sonar", - "author": "perplexity", - "description": "Sonar is lightweight, affordable, fast, and simple to use — now featuring citations and the ability to customize sources. It is designed for companies seeking to integrate lightweight question-and-answer features optimized for speed.", + "name": "OpenAI: GPT-4.5 (Preview)", + "shortName": "GPT-4.5 (Preview)", + "author": "openai", + "description": "GPT-4.5 (Preview) is a research preview of OpenAI’s latest language model, designed to advance capabilities in reasoning, creativity, and multi-turn conversation. It builds on previous iterations with improvements in world knowledge, contextual coherence, and the ability to follow user intent more effectively.\n\nThe model demonstrates enhanced performance in tasks that require open-ended thinking, problem-solving, and communication. Early testing suggests it is better at generating nuanced responses, maintaining long-context coherence, and reducing hallucinations compared to earlier versions.\n\nThis research preview is intended to help evaluate GPT-4.5’s strengths and limitations in real-world use cases as OpenAI continues to refine and develop future models. Read more at the [blog post here.](https://openai.com/index/introducing-gpt-4-5/)", "modelVersionGroupId": null, - "contextLength": 127072, + "contextLength": 128000, "inputModalities": [ "text", "image" @@ -46213,32 +49310,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "perplexity/sonar", + "warningMessage": null, + "permaslug": "openai/gpt-4.5-preview-2025-02-27", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "5f831e7a-c555-4d3a-b228-88286347558a", - "name": "Perplexity | perplexity/sonar", - "contextLength": 127072, + "id": "dc592ff9-2fc9-4a5e-b910-27a57bb7bfdd", + "name": "OpenAI | openai/gpt-4.5-preview-2025-02-27", + "contextLength": 128000, "model": { - "slug": "perplexity/sonar", + "slug": "openai/gpt-4.5-preview", "hfSlug": "", - "updatedAt": "2025-05-05T14:39:01.632671+00:00", - "createdAt": "2025-01-27T21:36:48.666939+00:00", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-27T20:23:30.841555+00:00", "hfUpdatedAt": null, - "name": "Perplexity: Sonar", - "shortName": "Sonar", - "author": "perplexity", - "description": "Sonar is lightweight, affordable, fast, and simple to use — now featuring citations and the ability to customize sources. It is designed for companies seeking to integrate lightweight question-and-answer features optimized for speed.", + "name": "OpenAI: GPT-4.5 (Preview)", + "shortName": "GPT-4.5 (Preview)", + "author": "openai", + "description": "GPT-4.5 (Preview) is a research preview of OpenAI’s latest language model, designed to advance capabilities in reasoning, creativity, and multi-turn conversation. It builds on previous iterations with improvements in world knowledge, contextual coherence, and the ability to follow user intent more effectively.\n\nThe model demonstrates enhanced performance in tasks that require open-ended thinking, problem-solving, and communication. Early testing suggests it is better at generating nuanced responses, maintaining long-context coherence, and reducing hallucinations compared to earlier versions.\n\nThis research preview is intended to help evaluate GPT-4.5’s strengths and limitations in real-world use cases as OpenAI continues to refine and develop future models. Read more at the [blog post here.](https://openai.com/index/introducing-gpt-4-5/)", "modelVersionGroupId": null, - "contextLength": 127072, + "contextLength": 128000, "inputModalities": [ "text", "image" @@ -46247,113 +49344,116 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "perplexity/sonar", + "warningMessage": null, + "permaslug": "openai/gpt-4.5-preview-2025-02-27", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "perplexity/sonar", - "modelVariantPermaslug": "perplexity/sonar", - "providerName": "Perplexity", + "modelVariantSlug": "openai/gpt-4.5-preview", + "modelVariantPermaslug": "openai/gpt-4.5-preview-2025-02-27", + "providerName": "OpenAI", "providerInfo": { - "name": "Perplexity", - "displayName": "Perplexity", - "slug": "perplexity", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", - "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { - "training": false - } + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, - "moderationRequired": false, - "group": "Perplexity", + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.perplexity.ai/", + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Perplexity.svg" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Perplexity", - "providerModelId": "sonar", - "providerGroup": "Perplexity", - "quantization": "unknown", + "providerDisplayName": "OpenAI", + "providerModelId": "gpt-4.5-preview-2025-02-27", + "providerGroup": "OpenAI", + "quantization": null, "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 16384, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "web_search_options", - "top_k", + "stop", "frequency_penalty", - "presence_penalty" + "presence_penalty", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", - "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { - "training": false + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "training": false + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.000001", - "completion": "0.000001", - "image": "0", - "request": "0.005", + "prompt": "0.000075", + "completion": "0.00015", + "image": "0.108375", + "request": "0", + "inputCacheRead": "0.0000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [ - { - "type": "search-threshold", - "threshold": "high", - "request": "0.012" - }, - { - "type": "search-threshold", - "threshold": "medium", - "request": "0.008" - }, - { - "type": "search-threshold", - "threshold": "low", - "request": "0.005" - } - ], + "variablePricings": [], "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -46361,56 +49461,68 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 155, - "newest": 138, - "throughputHighToLow": 85, - "latencyLowToHigh": 238, - "pricingLowToHigh": 287, - "pricingHighToLow": 30 + "topWeekly": 159, + "newest": 106, + "throughputHighToLow": 256, + "latencyLowToHigh": 190, + "pricingLowToHigh": 317, + "pricingHighToLow": 1 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "Perplexity", - "slug": "perplexity", - "quantization": "unknown", - "context": 127072, - "maxCompletionTokens": null, - "providerModelId": "sonar", + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": null, + "context": 128000, + "maxCompletionTokens": 16384, + "providerModelId": "gpt-4.5-preview-2025-02-27", "pricing": { - "prompt": "0.000001", - "completion": "0.000001", - "image": "0", - "request": "0.005", + "prompt": "0.000075", + "completion": "0.00015", + "image": "0.108375", + "request": "0", + "inputCacheRead": "0.0000375", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "web_search_options", - "top_k", + "stop", "frequency_penalty", - "presence_penalty" + "presence_penalty", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" ], - "inputCost": 1, - "outputCost": 1 + "inputCost": 75, + "outputCost": 150, + "throughput": 27.5085, + "latency": 1211 } ] }, { - "slug": "microsoft/phi-4-reasoning-plus", - "hfSlug": "microsoft/Phi-4-reasoning-plus", - "updatedAt": "2025-05-02T05:32:14.23708+00:00", - "createdAt": "2025-05-01T20:22:41.235613+00:00", + "slug": "qwen/qwen-plus", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-01T11:37:20.886831+00:00", "hfUpdatedAt": null, - "name": "Microsoft: Phi 4 Reasoning Plus (free)", - "shortName": "Phi 4 Reasoning Plus (free)", - "author": "microsoft", - "description": "Phi-4-reasoning-plus is an enhanced 14B parameter model from Microsoft, fine-tuned from Phi-4 with additional reinforcement learning to boost accuracy on math, science, and code reasoning tasks. It uses the same dense decoder-only transformer architecture as Phi-4, but generates longer, more comprehensive outputs structured into a step-by-step reasoning trace and final answer.\n\nWhile it offers improved benchmark scores over Phi-4-reasoning across tasks like AIME, OmniMath, and HumanEvalPlus, its responses are typically ~50% longer, resulting in higher latency. Designed for English-only applications, it is well-suited for structured reasoning workflows where output quality takes priority over response speed.", + "name": "Qwen: Qwen-Plus", + "shortName": "Qwen-Plus", + "author": "qwen", + "description": "Qwen-Plus, based on the Qwen2.5 foundation model, is a 131K context model with a balanced performance, speed, and cost combination.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -46418,40 +49530,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "Qwen", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "microsoft/phi-4-reasoning-plus-04-30", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "warningMessage": "", + "permaslug": "qwen/qwen-plus-2025-01-25", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "bbc205d3-b5e1-457f-8b74-771855206a98", - "name": "Chutes | microsoft/phi-4-reasoning-plus-04-30:free", - "contextLength": 32768, + "id": "47e88ae3-94ed-4d6a-96bb-a150dd354853", + "name": "Alibaba | qwen/qwen-plus-2025-01-25", + "contextLength": 131072, "model": { - "slug": "microsoft/phi-4-reasoning-plus", - "hfSlug": "microsoft/Phi-4-reasoning-plus", - "updatedAt": "2025-05-02T05:32:14.23708+00:00", - "createdAt": "2025-05-01T20:22:41.235613+00:00", + "slug": "qwen/qwen-plus", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-01T11:37:20.886831+00:00", "hfUpdatedAt": null, - "name": "Microsoft: Phi 4 Reasoning Plus", - "shortName": "Phi 4 Reasoning Plus", - "author": "microsoft", - "description": "Phi-4-reasoning-plus is an enhanced 14B parameter model from Microsoft, fine-tuned from Phi-4 with additional reinforcement learning to boost accuracy on math, science, and code reasoning tasks. It uses the same dense decoder-only transformer architecture as Phi-4, but generates longer, more comprehensive outputs structured into a step-by-step reasoning trace and final answer.\n\nWhile it offers improved benchmark scores over Phi-4-reasoning across tasks like AIME, OmniMath, and HumanEvalPlus, its responses are typically ~50% longer, resulting in higher latency. Designed for English-only applications, it is well-suited for structured reasoning workflows where output quality takes priority over response speed.", + "name": "Qwen: Qwen-Plus", + "shortName": "Qwen-Plus", + "author": "qwen", + "description": "Qwen-Plus, based on the Qwen2.5 foundation model, is a 131K context model with a balanced performance, speed, and cost combination.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -46459,45 +49563,38 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "Qwen", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "microsoft/phi-4-reasoning-plus-04-30", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "warningMessage": "", + "permaslug": "qwen/qwen-plus-2025-01-25", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "microsoft/phi-4-reasoning-plus:free", - "modelVariantPermaslug": "microsoft/phi-4-reasoning-plus-04-30:free", - "providerName": "Chutes", + "modelVariantSlug": "qwen/qwen-plus", + "modelVariantPermaslug": "qwen/qwen-plus-2025-01-25", + "providerName": "Alibaba", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "Alibaba", + "displayName": "Alibaba", + "slug": "alibaba", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", + "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false } }, + "headquarters": "CN", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": true, + "isAbortable": false, "moderationRequired": false, - "group": "Chutes", + "group": "Alibaba", "editors": [], "owners": [], "isMultipartSupported": true, @@ -46505,51 +49602,43 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256" } }, - "providerDisplayName": "Chutes", - "providerModelId": "microsoft/Phi-4-reasoning-plus", - "providerGroup": "Chutes", + "providerDisplayName": "Alibaba", + "providerModelId": "qwen-plus", + "providerGroup": "Alibaba", "quantization": null, - "variant": "free", - "isFree": true, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": null, + "variant": "standard", + "isFree": false, + "canAbort": false, + "maxPromptTokens": 129024, + "maxCompletionTokens": 8192, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", "seed", - "top_k", - "min_p", - "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" + "response_format", + "presence_penalty" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", + "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false }, - "training": true, - "retainsPrompts": true + "training": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.0000004", + "completion": "0.0000012", "image": "0", "request": "0", "webSearch": "0", @@ -46560,38 +49649,39 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": true, + "supportsToolParameters": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 149, - "newest": 10, - "throughputHighToLow": 30, - "latencyLowToHigh": 122, - "pricingLowToHigh": 1, - "pricingHighToLow": 206 + "topWeekly": 160, + "newest": 127, + "throughputHighToLow": 231, + "latencyLowToHigh": 189, + "pricingLowToHigh": 175, + "pricingHighToLow": 142 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "microsoft/phi-4-reasoning-plus", + "name": "Alibaba", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256", + "slug": "alibaba", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 8192, + "providerModelId": "qwen-plus", "pricing": { - "prompt": "0.00000007", - "completion": "0.00000035", + "prompt": "0.0000004", + "completion": "0.0000012", "image": "0", "request": "0", "webSearch": "0", @@ -46599,118 +49689,106 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", "seed", - "min_p" + "response_format", + "presence_penalty" ], - "inputCost": 0.07, - "outputCost": 0.35 + "inputCost": 0.4, + "outputCost": 1.2, + "throughput": 33.081, + "latency": 994 } ] }, { - "slug": "google/gemma-2-27b-it", - "hfSlug": "google/gemma-2-27b-it", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-07-13T00:00:00+00:00", + "slug": "opengvlab/internvl3-14b", + "hfSlug": "OpenGVLab/InternVL3-14B", + "updatedAt": "2025-04-30T14:14:32.387955+00:00", + "createdAt": "2025-04-30T13:55:55.014183+00:00", "hfUpdatedAt": null, - "name": "Google: Gemma 2 27B", - "shortName": "Gemma 2 27B", - "author": "google", - "description": "Gemma 2 27B by Google is an open model built from the same research and technology used to create the [Gemini models](/models?q=gemini).\n\nGemma models are well-suited for a variety of text generation tasks, including question answering, summarization, and reasoning.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).", + "name": "OpenGVLab: InternVL3 14B (free)", + "shortName": "InternVL3 14B (free)", + "author": "opengvlab", + "description": "The 14b version of the InternVL3 series. An advanced multimodal large language model (MLLM) series that demonstrates superior overall performance. Compared to InternVL 2.5, InternVL3 exhibits superior multimodal perception and reasoning capabilities, while further extending its multimodal capabilities to encompass tool usage, GUI agents, industrial image analysis, 3D vision perception, and more.", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 32000, "inputModalities": [ + "image", "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Gemini", - "instructType": "gemma", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "", - "", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "google/gemma-2-27b-it", + "permaslug": "opengvlab/internvl3-14b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "2ddecefa-fddf-44bc-b8db-fa904f993eef", - "name": "Nebius | google/gemma-2-27b-it", - "contextLength": 8192, + "id": "3be27401-83ab-4e46-9418-462abbc1c8af", + "name": "Nineteen | opengvlab/internvl3-14b:free", + "contextLength": 32000, "model": { - "slug": "google/gemma-2-27b-it", - "hfSlug": "google/gemma-2-27b-it", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-07-13T00:00:00+00:00", + "slug": "opengvlab/internvl3-14b", + "hfSlug": "OpenGVLab/InternVL3-14B", + "updatedAt": "2025-04-30T14:14:32.387955+00:00", + "createdAt": "2025-04-30T13:55:55.014183+00:00", "hfUpdatedAt": null, - "name": "Google: Gemma 2 27B", - "shortName": "Gemma 2 27B", - "author": "google", - "description": "Gemma 2 27B by Google is an open model built from the same research and technology used to create the [Gemini models](/models?q=gemini).\n\nGemma models are well-suited for a variety of text generation tasks, including question answering, summarization, and reasoning.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).", + "name": "OpenGVLab: InternVL3 14B", + "shortName": "InternVL3 14B", + "author": "opengvlab", + "description": "The 14b version of the InternVL3 series. An advanced multimodal large language model (MLLM) series that demonstrates superior overall performance. Compared to InternVL 2.5, InternVL3 exhibits superior multimodal perception and reasoning capabilities, while further extending its multimodal capabilities to encompass tool usage, GUI agents, industrial image analysis, 3D vision perception, and more.", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 32000, "inputModalities": [ + "image", "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Gemini", - "instructType": "gemma", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "", - "", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "google/gemma-2-27b-it", + "permaslug": "opengvlab/internvl3-14b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "google/gemma-2-27b-it", - "modelVariantPermaslug": "google/gemma-2-27b-it", - "providerName": "Nebius", + "modelVariantSlug": "opengvlab/internvl3-14b:free", + "modelVariantPermaslug": "opengvlab/internvl3-14b:free", + "providerName": "Nineteen", "providerInfo": { - "name": "Nebius", - "displayName": "Nebius AI Studio", - "slug": "nebius", + "name": "Nineteen", + "displayName": "Nineteen", + "slug": "nineteen", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "termsOfServiceUrl": "https://nineteen.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": false + "training": false } }, - "headquarters": "DE", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, + "isAbortable": true, "moderationRequired": false, - "group": "Nebius", + "group": "Nineteen", "editors": [], "owners": [], "isMultipartSupported": true, @@ -46718,17 +49796,16 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nineteen.ai/&size=256" } }, - "providerDisplayName": "Nebius AI Studio", - "providerModelId": "google/gemma-2-27b-it", - "providerGroup": "Nebius", - "quantization": "fp8", - "variant": "standard", - "isFree": false, - "canAbort": false, + "providerDisplayName": "Nineteen", + "providerModelId": "OpenGVLab/InternVL3-14B", + "providerGroup": "Nineteen", + "quantization": "bf16", + "variant": "free", + "isFree": true, + "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, "maxPromptImages": null, @@ -46736,31 +49813,20 @@ export default { "supportedParameters": [ "max_tokens", "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" + "top_p" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "termsOfServiceUrl": "https://nineteen.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": false + "training": false }, - "training": false, - "retainsPrompts": false + "training": false }, "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -46785,164 +49851,104 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 157, - "newest": 238, - "throughputHighToLow": 192, - "latencyLowToHigh": 77, - "pricingLowToHigh": 124, - "pricingHighToLow": 195 + "topWeekly": 161, + "newest": 17, + "throughputHighToLow": 25, + "latencyLowToHigh": 176, + "pricingLowToHigh": 6, + "pricingHighToLow": 254 }, - "providers": [ - { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 8192, - "maxCompletionTokens": null, - "providerModelId": "google/gemma-2-27b-it", - "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" - ], - "inputCost": 0.1, - "outputCost": 0.3 - }, - { - "name": "Together", - "slug": "together", - "quantization": "unknown", - "context": 8192, - "maxCompletionTokens": 2048, - "providerModelId": "google/gemma-2-27b-it", - "pricing": { - "prompt": "0.0000008", - "completion": "0.0000008", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" - ], - "inputCost": 0.8, - "outputCost": 0.8 - } - ] + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [] }, { - "slug": "qwen/qwen-plus", - "hfSlug": "", + "slug": "microsoft/phi-4-multimodal-instruct", + "hfSlug": "microsoft/Phi-4-multimodal-instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-01T11:37:20.886831+00:00", + "createdAt": "2025-03-08T01:11:24.652063+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen-Plus", - "shortName": "Qwen-Plus", - "author": "qwen", - "description": "Qwen-Plus, based on the Qwen2.5 foundation model, is a 131K context model with a balanced performance, speed, and cost combination.", + "name": "Microsoft: Phi 4 Multimodal Instruct", + "shortName": "Phi 4 Multimodal Instruct", + "author": "microsoft", + "description": "Phi-4 Multimodal Instruct is a versatile 5.6B parameter foundation model that combines advanced reasoning and instruction-following capabilities across both text and visual inputs, providing accurate text outputs. The unified architecture enables efficient, low-latency inference, suitable for edge and mobile deployments. Phi-4 Multimodal Instruct supports text inputs in multiple languages including Arabic, Chinese, English, French, German, Japanese, Spanish, and more, with visual input optimized primarily for English. It delivers impressive performance on multimodal tasks involving mathematical, scientific, and document reasoning, providing developers and enterprises a powerful yet compact model for sophisticated interactive applications. For more information, see the [Phi-4 Multimodal blog post](https://azure.microsoft.com/en-us/blog/empowering-innovation-the-next-generation-of-the-phi-family/).\n", "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "qwen/qwen-plus-2025-01-25", + "warningMessage": null, + "permaslug": "microsoft/phi-4-multimodal-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "47e88ae3-94ed-4d6a-96bb-a150dd354853", - "name": "Alibaba | qwen/qwen-plus-2025-01-25", + "id": "345f416d-3f33-4530-8033-614ac900a8fd", + "name": "DeepInfra | microsoft/phi-4-multimodal-instruct", "contextLength": 131072, "model": { - "slug": "qwen/qwen-plus", - "hfSlug": "", + "slug": "microsoft/phi-4-multimodal-instruct", + "hfSlug": "microsoft/Phi-4-multimodal-instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-01T11:37:20.886831+00:00", + "createdAt": "2025-03-08T01:11:24.652063+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen-Plus", - "shortName": "Qwen-Plus", - "author": "qwen", - "description": "Qwen-Plus, based on the Qwen2.5 foundation model, is a 131K context model with a balanced performance, speed, and cost combination.", + "name": "Microsoft: Phi 4 Multimodal Instruct", + "shortName": "Phi 4 Multimodal Instruct", + "author": "microsoft", + "description": "Phi-4 Multimodal Instruct is a versatile 5.6B parameter foundation model that combines advanced reasoning and instruction-following capabilities across both text and visual inputs, providing accurate text outputs. The unified architecture enables efficient, low-latency inference, suitable for edge and mobile deployments. Phi-4 Multimodal Instruct supports text inputs in multiple languages including Arabic, Chinese, English, French, German, Japanese, Spanish, and more, with visual input optimized primarily for English. It delivers impressive performance on multimodal tasks involving mathematical, scientific, and document reasoning, providing developers and enterprises a powerful yet compact model for sophisticated interactive applications. For more information, see the [Phi-4 Multimodal blog post](https://azure.microsoft.com/en-us/blog/empowering-innovation-the-next-generation-of-the-phi-family/).\n", "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "qwen/qwen-plus-2025-01-25", + "warningMessage": null, + "permaslug": "microsoft/phi-4-multimodal-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "qwen/qwen-plus", - "modelVariantPermaslug": "qwen/qwen-plus-2025-01-25", - "providerName": "Alibaba", + "modelVariantSlug": "microsoft/phi-4-multimodal-instruct", + "modelVariantPermaslug": "microsoft/phi-4-multimodal-instruct", + "providerName": "DeepInfra", "providerInfo": { - "name": "Alibaba", - "displayName": "Alibaba", - "slug": "alibaba", + "name": "DeepInfra", + "displayName": "DeepInfra", + "slug": "deepinfra", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", - "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false } }, - "headquarters": "CN", + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, + "isAbortable": true, "moderationRequired": false, - "group": "Alibaba", + "group": "DeepInfra", "editors": [], "owners": [], "isMultipartSupported": true, @@ -46950,44 +49956,50 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256" + "url": "/images/icons/DeepInfra.webp" } }, - "providerDisplayName": "Alibaba", - "providerModelId": "qwen-plus", - "providerGroup": "Alibaba", - "quantization": null, + "providerDisplayName": "DeepInfra", + "providerModelId": "microsoft/Phi-4-multimodal-instruct", + "providerGroup": "DeepInfra", + "quantization": "bf16", "variant": "standard", "isFree": false, - "canAbort": false, - "maxPromptTokens": 129024, - "maxCompletionTokens": 8192, - "maxPromptImages": null, - "maxTokensPerImage": null, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": 1, + "maxTokensPerImage": 3537, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "seed", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", "response_format", - "presence_penalty" + "top_k", + "seed", + "min_p" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", - "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false }, - "training": false + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0.0000004", - "completion": "0.0000012", - "image": "0", + "prompt": "0.00000005", + "completion": "0.0000001", + "image": "0.00017685", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -46997,7 +50009,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -47010,57 +50022,64 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 158, - "newest": 127, - "throughputHighToLow": 255, - "latencyLowToHigh": 123, - "pricingLowToHigh": 175, - "pricingHighToLow": 142 + "topWeekly": 162, + "newest": 97, + "throughputHighToLow": 259, + "latencyLowToHigh": 265, + "pricingLowToHigh": 98, + "pricingHighToLow": 221 }, + "authorIcon": "https://openrouter.ai/images/icons/Microsoft.svg", "providers": [ { - "name": "Alibaba", - "slug": "alibaba", - "quantization": null, + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", "context": 131072, - "maxCompletionTokens": 8192, - "providerModelId": "qwen-plus", + "maxCompletionTokens": null, + "providerModelId": "microsoft/Phi-4-multimodal-instruct", "pricing": { - "prompt": "0.0000004", - "completion": "0.0000012", - "image": "0", + "prompt": "0.00000005", + "completion": "0.0000001", + "image": "0.00017685", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "seed", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", "response_format", - "presence_penalty" + "top_k", + "seed", + "min_p" ], - "inputCost": 0.4, - "outputCost": 1.2 + "inputCost": 0.05, + "outputCost": 0.1, + "throughput": 11.834, + "latency": 3108 } ] }, { - "slug": "perplexity/llama-3.1-sonar-small-128k-online", - "hfSlug": null, + "slug": "mistralai/mixtral-8x22b-instruct", + "hfSlug": "mistralai/Mixtral-8x22B-Instruct-v0.1", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-01T00:00:00+00:00", + "createdAt": "2024-04-17T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Perplexity: Llama 3.1 Sonar 8B Online", - "shortName": "Llama 3.1 Sonar 8B Online", - "author": "perplexity", - "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3.1-sonar-small-128k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", - "modelVersionGroupId": "9e4cc81b-5f14-4987-9e40-74db1c86ecda", - "contextLength": 127072, + "name": "Mistral: Mixtral 8x22B Instruct", + "shortName": "Mixtral 8x22B Instruct", + "author": "mistralai", + "description": "Mistral's official instruct fine-tuned version of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b). It uses 39B active parameters out of 141B, offering unparalleled cost efficiency for its size. Its strengths include:\n- strong math, coding, and reasoning\n- large context length (64k)\n- fluency in English, French, Italian, German, and Spanish\n\nSee benchmarks on the launch announcement [here](https://mistral.ai/news/mixtral-8x22b/).\n#moe", + "modelVersionGroupId": null, + "contextLength": 65536, "inputModalities": [ "text" ], @@ -47068,32 +50087,35 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": null, + "group": "Mistral", + "instructType": "mistral", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "[INST]", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "perplexity/llama-3.1-sonar-small-128k-online", + "permaslug": "mistralai/mixtral-8x22b-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "48b44b03-a839-42e9-863e-1b6c07aa42ae", - "name": "Perplexity | perplexity/llama-3.1-sonar-small-128k-online", - "contextLength": 127072, + "id": "c26356d7-ec87-4537-b049-61870b15cf3c", + "name": "Nebius | mistralai/mixtral-8x22b-instruct", + "contextLength": 65536, "model": { - "slug": "perplexity/llama-3.1-sonar-small-128k-online", - "hfSlug": null, + "slug": "mistralai/mixtral-8x22b-instruct", + "hfSlug": "mistralai/Mixtral-8x22B-Instruct-v0.1", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-01T00:00:00+00:00", + "createdAt": "2024-04-17T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Perplexity: Llama 3.1 Sonar 8B Online", - "shortName": "Llama 3.1 Sonar 8B Online", - "author": "perplexity", - "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3.1-sonar-small-128k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", - "modelVersionGroupId": "9e4cc81b-5f14-4987-9e40-74db1c86ecda", - "contextLength": 127072, + "name": "Mistral: Mixtral 8x22B Instruct", + "shortName": "Mixtral 8x22B Instruct", + "author": "mistralai", + "description": "Mistral's official instruct fine-tuned version of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b). It uses 39B active parameters out of 141B, offering unparalleled cost efficiency for its size. Its strengths include:\n- strong math, coding, and reasoning\n- large context length (64k)\n- fluency in English, French, Italian, German, and Spanish\n\nSee benchmarks on the launch announcement [here](https://mistral.ai/news/mixtral-8x22b/).\n#moe", + "modelVersionGroupId": null, + "contextLength": 65536, "inputModalities": [ "text" ], @@ -47101,52 +50123,57 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": null, + "group": "Mistral", + "instructType": "mistral", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "[INST]", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "perplexity/llama-3.1-sonar-small-128k-online", + "permaslug": "mistralai/mixtral-8x22b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "perplexity/llama-3.1-sonar-small-128k-online", - "modelVariantPermaslug": "perplexity/llama-3.1-sonar-small-128k-online", - "providerName": "Perplexity", + "modelVariantSlug": "mistralai/mixtral-8x22b-instruct", + "modelVariantPermaslug": "mistralai/mixtral-8x22b-instruct", + "providerName": "Nebius", "providerInfo": { - "name": "Perplexity", - "displayName": "Perplexity", - "slug": "perplexity", + "name": "Nebius", + "displayName": "Nebius AI Studio", + "slug": "nebius", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", - "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false } }, - "headquarters": "US", + "headquarters": "DE", "hasChatCompletions": true, - "hasCompletions": false, + "hasCompletions": true, "isAbortable": false, "moderationRequired": false, - "group": "Perplexity", + "group": "Nebius", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.perplexity.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Perplexity.svg" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "invertRequired": true } }, - "providerDisplayName": "Perplexity", - "providerModelId": "llama-3.1-sonar-small-128k-online", - "providerGroup": "Perplexity", - "quantization": "unknown", + "providerDisplayName": "Nebius AI Studio", + "providerModelId": "mistralai/Mixtral-8x22B-Instruct-v0.1", + "providerGroup": "Nebius", + "quantization": "fp8", "variant": "standard", "isFree": false, "canAbort": false, @@ -47158,25 +50185,32 @@ export default { "max_tokens", "temperature", "top_p", - "top_k", + "stop", "frequency_penalty", - "presence_penalty" + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", - "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false }, - "training": false + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", + "prompt": "0.0000004", + "completion": "0.0000012", "image": "0", - "request": "0.005", + "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -47190,63 +50224,147 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 159, - "newest": 228, - "throughputHighToLow": 9, - "latencyLowToHigh": 185, - "pricingLowToHigh": 286, - "pricingHighToLow": 32 + "topWeekly": 163, + "newest": 268, + "throughputHighToLow": 188, + "latencyLowToHigh": 51, + "pricingLowToHigh": 176, + "pricingHighToLow": 143 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { - "name": "Perplexity", - "slug": "perplexity", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 65536, + "maxCompletionTokens": null, + "providerModelId": "mistralai/Mixtral-8x22B-Instruct-v0.1", + "pricing": { + "prompt": "0.0000004", + "completion": "0.0000012", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.4, + "outputCost": 1.2, + "throughput": 46.9865, + "latency": 409 + }, + { + "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", + "slug": "fireworks", "quantization": "unknown", - "context": 127072, + "context": 65536, "maxCompletionTokens": null, - "providerModelId": "llama-3.1-sonar-small-128k-online", + "providerModelId": "accounts/fireworks/models/mixtral-8x22b-instruct", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", + "prompt": "0.0000009", + "completion": "0.0000009", "image": "0", - "request": "0.005", + "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "top_k", + "repetition_penalty", + "response_format", + "structured_outputs", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.9, + "outputCost": 0.9, + "throughput": 80.916, + "latency": 432 + }, + { + "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", + "slug": "mistral", + "quantization": "unknown", + "context": 65536, + "maxCompletionTokens": null, + "providerModelId": "open-mixtral-8x22b", + "pricing": { + "prompt": "0.000002", + "completion": "0.000006", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", "frequency_penalty", - "presence_penalty" + "presence_penalty", + "response_format", + "structured_outputs", + "seed" ], - "inputCost": 0.2, - "outputCost": 0.2 + "inputCost": 2, + "outputCost": 6, + "throughput": 63.068, + "latency": 359 } ] }, { - "slug": "thudm/glm-z1-rumination-32b", - "hfSlug": "THUDM/GLM-Z1-Rumination-32B-0414", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-04-25T17:18:15.30585+00:00", + "slug": "qwen/qwen-2-72b-instruct", + "hfSlug": "Qwen/Qwen2-72B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-06-07T00:00:00+00:00", "hfUpdatedAt": null, - "name": "THUDM: GLM Z1 Rumination 32B ", - "shortName": "GLM Z1 Rumination 32B ", - "author": "thudm", - "description": "THUDM: GLM Z1 Rumination 32B is a 32B-parameter deep reasoning model from the GLM-4-Z1 series, optimized for complex, open-ended tasks requiring prolonged deliberation. It builds upon glm-4-32b-0414 with additional reinforcement learning phases and multi-stage alignment strategies, introducing “rumination” capabilities designed to emulate extended cognitive processing. This includes iterative reasoning, multi-hop analysis, and tool-augmented workflows such as search, retrieval, and citation-aware synthesis.\n\nThe model excels in research-style writing, comparative analysis, and intricate question answering. It supports function calling for search and navigation primitives (`search`, `click`, `open`, `finish`), enabling use in agent-style pipelines. Rumination behavior is governed by multi-turn loops with rule-based reward shaping and delayed decision mechanisms, benchmarked against Deep Research frameworks such as OpenAI’s internal alignment stacks. This variant is suitable for scenarios requiring depth over speed.", + "name": "Qwen 2 72B Instruct", + "shortName": "Qwen 2 72B Instruct", + "author": "qwen", + "description": "Qwen2 72B is a transformer-based model that excels in language understanding, multilingual capabilities, coding, mathematics, and reasoning.\n\nIt features SwiGLU activation, attention QKV bias, and group query attention. It is pretrained on extensive data with supervised finetuning and direct preference optimization.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2/) and [GitHub repo](https://github.com/QwenLM/Qwen2).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", "modelVersionGroupId": null, - "contextLength": 32000, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -47254,43 +50372,36 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "deepseek-r1", + "group": "Qwen", + "instructType": "chatml", "defaultSystem": null, "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "thudm/glm-z1-rumination-32b-0414", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "permaslug": "qwen/qwen-2-72b-instruct", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "c61d1c8b-1e6d-4f60-97d8-893f51320223", - "name": "Novita | thudm/glm-z1-rumination-32b-0414", - "contextLength": 32000, + "id": "89dce4f1-1836-4539-88f3-ee86d594d7a5", + "name": "Together | qwen/qwen-2-72b-instruct", + "contextLength": 32768, "model": { - "slug": "thudm/glm-z1-rumination-32b", - "hfSlug": "THUDM/GLM-Z1-Rumination-32B-0414", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-04-25T17:18:15.30585+00:00", + "slug": "qwen/qwen-2-72b-instruct", + "hfSlug": "Qwen/Qwen2-72B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-06-07T00:00:00+00:00", "hfUpdatedAt": null, - "name": "THUDM: GLM Z1 Rumination 32B ", - "shortName": "GLM Z1 Rumination 32B ", - "author": "thudm", - "description": "THUDM: GLM Z1 Rumination 32B is a 32B-parameter deep reasoning model from the GLM-4-Z1 series, optimized for complex, open-ended tasks requiring prolonged deliberation. It builds upon glm-4-32b-0414 with additional reinforcement learning phases and multi-stage alignment strategies, introducing “rumination” capabilities designed to emulate extended cognitive processing. This includes iterative reasoning, multi-hop analysis, and tool-augmented workflows such as search, retrieval, and citation-aware synthesis.\n\nThe model excels in research-style writing, comparative analysis, and intricate question answering. It supports function calling for search and navigation primitives (`search`, `click`, `open`, `finish`), enabling use in agent-style pipelines. Rumination behavior is governed by multi-turn loops with rule-based reward shaping and delayed decision mechanisms, benchmarked against Deep Research frameworks such as OpenAI’s internal alignment stacks. This variant is suitable for scenarios requiring depth over speed.", + "name": "Qwen 2 72B Instruct", + "shortName": "Qwen 2 72B Instruct", + "author": "qwen", + "description": "Qwen2 72B is a transformer-based model that excels in language understanding, multilingual capabilities, coding, mathematics, and reasoning.\n\nIt features SwiGLU activation, attention QKV bias, and group query attention. It is pretrained on extensive data with supervised finetuning and direct preference optimization.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2/) and [GitHub repo](https://github.com/QwenLM/Qwen2).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", "modelVersionGroupId": null, - "contextLength": 32000, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -47298,41 +50409,35 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "deepseek-r1", + "group": "Qwen", + "instructType": "chatml", "defaultSystem": null, "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "thudm/glm-z1-rumination-32b-0414", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "permaslug": "qwen/qwen-2-72b-instruct", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "thudm/glm-z1-rumination-32b", - "modelVariantPermaslug": "thudm/glm-z1-rumination-32b-0414", - "providerName": "Novita", + "modelVariantSlug": "qwen/qwen-2-72b-instruct", + "modelVariantPermaslug": "qwen/qwen-2-72b-instruct", + "providerName": "Together", "providerInfo": { - "name": "Novita", - "displayName": "NovitaAI", - "slug": "novita", + "name": "Together", + "displayName": "Together", + "slug": "together", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false } }, "headquarters": "US", @@ -47340,57 +50445,56 @@ export default { "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Novita", + "group": "Together", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.novita.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" } }, - "providerDisplayName": "NovitaAI", - "providerModelId": "thudm/glm-z1-rumination-32b-0414", - "providerGroup": "Novita", - "quantization": null, + "providerDisplayName": "Together", + "providerModelId": "Qwen/Qwen2-72B-Instruct", + "providerGroup": "Together", + "quantization": "unknown", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "seed", "top_k", - "min_p", "repetition_penalty", - "logit_bias" + "logit_bias", + "min_p", + "response_format" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false }, - "training": false + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0.00000024", - "completion": "0.00000024", + "prompt": "0.0000009", + "completion": "0.0000009", "image": "0", "request": "0", "webSearch": "0", @@ -47402,37 +50506,38 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 160, - "newest": 33, - "throughputHighToLow": 275, - "latencyLowToHigh": 228, - "pricingLowToHigh": 156, - "pricingHighToLow": 160 + "topWeekly": 164, + "newest": 248, + "throughputHighToLow": 206, + "latencyLowToHigh": 93, + "pricingLowToHigh": 214, + "pricingHighToLow": 105 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": null, - "context": 32000, - "maxCompletionTokens": null, - "providerModelId": "thudm/glm-z1-rumination-32b-0414", + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": "unknown", + "context": 32768, + "maxCompletionTokens": 4096, + "providerModelId": "Qwen/Qwen2-72B-Instruct", "pricing": { - "prompt": "0.00000024", - "completion": "0.00000024", + "prompt": "0.0000009", + "completion": "0.0000009", "image": "0", "request": "0", "webSearch": "0", @@ -47443,34 +50548,34 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "seed", "top_k", - "min_p", "repetition_penalty", - "logit_bias" + "logit_bias", + "min_p", + "response_format" ], - "inputCost": 0.24, - "outputCost": 0.24 + "inputCost": 0.9, + "outputCost": 0.9, + "throughput": 38.908, + "latency": 703 } ] }, { - "slug": "nousresearch/deephermes-3-mistral-24b-preview", - "hfSlug": "NousResearch/DeepHermes-3-Mistral-24B-Preview", - "updatedAt": "2025-05-09T23:02:10.58499+00:00", - "createdAt": "2025-05-09T22:48:24.165453+00:00", + "slug": "nvidia/llama-3.3-nemotron-super-49b-v1", + "hfSlug": "nvidia/Llama-3_3-Nemotron-Super-49B-v1", + "updatedAt": "2025-04-08T14:53:49.407401+00:00", + "createdAt": "2025-04-08T13:38:14.544231+00:00", "hfUpdatedAt": null, - "name": "Nous: DeepHermes 3 Mistral 24B Preview (free)", - "shortName": "DeepHermes 3 Mistral 24B Preview (free)", - "author": "nousresearch", - "description": "DeepHermes 3 (Mistral 24B Preview) is an instruction-tuned language model by Nous Research based on Mistral-Small-24B, designed for chat, function calling, and advanced multi-turn reasoning. It introduces a dual-mode system that toggles between intuitive chat responses and structured “deep reasoning” mode using special system prompts. Fine-tuned via distillation from R1, it supports structured output (JSON mode) and function call syntax for agent-based applications.\n\nDeepHermes 3 supports a **reasoning toggle via system prompt**, allowing users to switch between fast, intuitive responses and deliberate, multi-step reasoning. When activated with the following specific system instruction, the model enters a *\"deep thinking\"* mode—generating extended chains of thought wrapped in `` tags before delivering a final answer. \n\nSystem Prompt: You are a deep thinking AI, you may use extremely long chains of thought to deeply consider the problem and deliberate with yourself via systematic reasoning processes to help come to a correct solution prior to answering. You should enclose your thoughts and internal monologue inside tags, and then provide your solution or response to the problem.\n", + "name": "NVIDIA: Llama 3.3 Nemotron Super 49B v1", + "shortName": "Llama 3.3 Nemotron Super 49B v1", + "author": "nvidia", + "description": "Llama-3.3-Nemotron-Super-49B-v1 is a large language model (LLM) optimized for advanced reasoning, conversational interactions, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta's Llama-3.3-70B-Instruct, it employs a Neural Architecture Search (NAS) approach, significantly enhancing efficiency and reducing memory requirements. This allows the model to support a context length of up to 128K tokens and fit efficiently on single high-performance GPUs, such as NVIDIA H200.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -47485,25 +50590,25 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "nousresearch/deephermes-3-mistral-24b-preview", + "permaslug": "nvidia/llama-3.3-nemotron-super-49b-v1", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "90728530-42b4-4bcf-bc67-167eee5bbac4", - "name": "Chutes | nousresearch/deephermes-3-mistral-24b-preview:free", - "contextLength": 32768, + "id": "1b7717f4-53d6-4a42-96ea-a58872ef08a0", + "name": "Nebius | nvidia/llama-3.3-nemotron-super-49b-v1", + "contextLength": 131072, "model": { - "slug": "nousresearch/deephermes-3-mistral-24b-preview", - "hfSlug": "NousResearch/DeepHermes-3-Mistral-24B-Preview", - "updatedAt": "2025-05-09T23:02:10.58499+00:00", - "createdAt": "2025-05-09T22:48:24.165453+00:00", + "slug": "nvidia/llama-3.3-nemotron-super-49b-v1", + "hfSlug": "nvidia/Llama-3_3-Nemotron-Super-49B-v1", + "updatedAt": "2025-04-08T14:53:49.407401+00:00", + "createdAt": "2025-04-08T13:38:14.544231+00:00", "hfUpdatedAt": null, - "name": "Nous: DeepHermes 3 Mistral 24B Preview", - "shortName": "DeepHermes 3 Mistral 24B Preview", - "author": "nousresearch", - "description": "DeepHermes 3 (Mistral 24B Preview) is an instruction-tuned language model by Nous Research based on Mistral-Small-24B, designed for chat, function calling, and advanced multi-turn reasoning. It introduces a dual-mode system that toggles between intuitive chat responses and structured “deep reasoning” mode using special system prompts. Fine-tuned via distillation from R1, it supports structured output (JSON mode) and function call syntax for agent-based applications.\n\nDeepHermes 3 supports a **reasoning toggle via system prompt**, allowing users to switch between fast, intuitive responses and deliberate, multi-step reasoning. When activated with the following specific system instruction, the model enters a *\"deep thinking\"* mode—generating extended chains of thought wrapped in `` tags before delivering a final answer. \n\nSystem Prompt: You are a deep thinking AI, you may use extremely long chains of thought to deeply consider the problem and deliberate with yourself via systematic reasoning processes to help come to a correct solution prior to answering. You should enclose your thoughts and internal monologue inside tags, and then provide your solution or response to the problem.\n", + "name": "NVIDIA: Llama 3.3 Nemotron Super 49B v1", + "shortName": "Llama 3.3 Nemotron Super 49B v1", + "author": "nvidia", + "description": "Llama-3.3-Nemotron-Super-49B-v1 is a large language model (LLM) optimized for advanced reasoning, conversational interactions, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta's Llama-3.3-70B-Instruct, it employs a Neural Architecture Search (NAS) approach, significantly enhancing efficiency and reducing memory requirements. This allows the model to support a context length of up to 128K tokens and fit efficiently on single high-performance GPUs, such as NVIDIA H200.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -47518,30 +50623,32 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "nousresearch/deephermes-3-mistral-24b-preview", + "permaslug": "nvidia/llama-3.3-nemotron-super-49b-v1", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "nousresearch/deephermes-3-mistral-24b-preview:free", - "modelVariantPermaslug": "nousresearch/deephermes-3-mistral-24b-preview:free", - "providerName": "Chutes", + "modelVariantSlug": "nvidia/llama-3.3-nemotron-super-49b-v1", + "modelVariantPermaslug": "nvidia/llama-3.3-nemotron-super-49b-v1", + "providerName": "Nebius", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "Nebius", + "displayName": "Nebius AI Studio", + "slug": "nebius", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false } }, + "headquarters": "DE", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": true, + "isAbortable": false, "moderationRequired": false, - "group": "Chutes", + "group": "Nebius", "editors": [], "owners": [], "isMultipartSupported": true, @@ -47549,16 +50656,17 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "invertRequired": true } }, - "providerDisplayName": "Chutes", - "providerModelId": "NousResearch/DeepHermes-3-Mistral-24B-Preview", - "providerGroup": "Chutes", - "quantization": null, - "variant": "free", - "isFree": true, - "canAbort": true, + "providerDisplayName": "Nebius AI Studio", + "providerModelId": "nvidia/Llama-3_3-Nemotron-Super-49B-v1", + "providerGroup": "Nebius", + "quantization": "fp8", + "variant": "standard", + "isFree": false, + "canAbort": false, "maxPromptTokens": null, "maxCompletionTokens": null, "maxPromptImages": null, @@ -47567,33 +50675,30 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", "seed", "top_k", - "min_p", - "repetition_penalty", - "logprobs", "logit_bias", + "logprobs", "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.00000013", + "completion": "0.0000004", "image": "0", "request": "0", "webSearch": "0", @@ -47605,7 +50710,7 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, @@ -47618,108 +50723,144 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 161, - "newest": 0, - "throughputHighToLow": 99, - "latencyLowToHigh": 181, - "pricingLowToHigh": 0, - "pricingHighToLow": 248 + "topWeekly": 151, + "newest": 58, + "throughputHighToLow": 172, + "latencyLowToHigh": 20, + "pricingLowToHigh": 24, + "pricingHighToLow": 182 }, - "providers": [] + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nvidia.com/\\u0026size=256", + "providers": [ + { + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "nvidia/Llama-3_3-Nemotron-Super-49B-v1", + "pricing": { + "prompt": "0.00000013", + "completion": "0.0000004", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.13, + "outputCost": 0.4, + "throughput": 47.2755, + "latency": 326 + } + ] }, { - "slug": "google/learnlm-1.5-pro-experimental", - "hfSlug": "", + "slug": "mistralai/mistral-7b-instruct", + "hfSlug": "mistralai/Mistral-7B-Instruct-v0.3", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-21T19:15:51.117686+00:00", + "createdAt": "2024-05-27T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Google: LearnLM 1.5 Pro Experimental (free)", - "shortName": "LearnLM 1.5 Pro Experimental (free)", - "author": "google", - "description": "An experimental version of [Gemini 1.5 Pro](/google/gemini-pro-1.5) from Google.", - "modelVersionGroupId": null, - "contextLength": 32767, + "name": "Mistral: Mistral 7B Instruct (free)", + "shortName": "Mistral 7B Instruct (free)", + "author": "mistralai", + "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*", + "modelVersionGroupId": "1d07cc56-c54d-4587-b785-5093496397a4", + "contextLength": 32768, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Gemini", - "instructType": null, + "group": "Mistral", + "instructType": "mistral", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "[INST]", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "google/learnlm-1.5-pro-experimental", + "permaslug": "mistralai/mistral-7b-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "bf41a79d-a83a-4b05-afad-b69bcc66dddc", - "name": "Google AI Studio | google/learnlm-1.5-pro-experimental:free", - "contextLength": 32767, + "id": "d09ade5a-2aaf-4a93-9881-3db42f1ba63a", + "name": "DeepInfra | mistralai/mistral-7b-instruct:free", + "contextLength": 32768, "model": { - "slug": "google/learnlm-1.5-pro-experimental", - "hfSlug": "", + "slug": "mistralai/mistral-7b-instruct", + "hfSlug": "mistralai/Mistral-7B-Instruct-v0.3", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-21T19:15:51.117686+00:00", + "createdAt": "2024-05-27T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Google: LearnLM 1.5 Pro Experimental", - "shortName": "LearnLM 1.5 Pro Experimental", - "author": "google", - "description": "An experimental version of [Gemini 1.5 Pro](/google/gemini-pro-1.5) from Google.", - "modelVersionGroupId": null, - "contextLength": 40960, + "name": "Mistral: Mistral 7B Instruct", + "shortName": "Mistral 7B Instruct", + "author": "mistralai", + "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*", + "modelVersionGroupId": "1d07cc56-c54d-4587-b785-5093496397a4", + "contextLength": 32768, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Gemini", - "instructType": null, + "group": "Mistral", + "instructType": "mistral", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "[INST]", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "google/learnlm-1.5-pro-experimental", + "permaslug": "mistralai/mistral-7b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "google/learnlm-1.5-pro-experimental:free", - "modelVariantPermaslug": "google/learnlm-1.5-pro-experimental:free", - "providerName": "Google AI Studio", + "modelVariantSlug": "mistralai/mistral-7b-instruct:free", + "modelVariantPermaslug": "mistralai/mistral-7b-instruct:free", + "providerName": "DeepInfra", "providerInfo": { - "name": "Google AI Studio", - "displayName": "Google AI Studio", - "slug": "google-ai-studio", + "name": "DeepInfra", + "displayName": "DeepInfra", + "slug": "deepinfra", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://cloud.google.com/terms/", - "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 55 - }, - "freeModels": { - "training": true, - "retainsPrompts": true, - "retentionDays": 55 + "retainsPrompts": false } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, + "hasCompletions": true, + "isAbortable": true, "moderationRequired": false, - "group": "Google AI Studio", + "group": "DeepInfra", "editors": [], "owners": [], "isMultipartSupported": true, @@ -47727,49 +50868,47 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/GoogleAIStudio.svg" + "url": "/images/icons/DeepInfra.webp" } }, - "providerDisplayName": "Google AI Studio", - "providerModelId": "learnlm-1.5-pro-experimental", - "providerGroup": "Google AI Studio", - "quantization": null, + "providerDisplayName": "DeepInfra", + "providerModelId": "mistralai/Mistral-7B-Instruct-v0.3", + "providerGroup": "DeepInfra", + "quantization": "bf16", "variant": "free", "isFree": true, - "canAbort": false, - "maxPromptTokens": 32767, - "maxCompletionTokens": 8192, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": 16384, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "seed", + "repetition_penalty", "response_format", - "structured_outputs" + "top_k", + "seed", + "min_p" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://cloud.google.com/terms/", - "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 55 - }, - "freeModels": { - "training": true, - "retainsPrompts": true, - "retentionDays": 55 + "retainsPrompts": false }, - "training": true, - "retainsPrompts": true, - "retentionDays": 55 + "training": false, + "retainsPrompts": false }, "pricing": { "prompt": "0", @@ -47784,43 +50923,216 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, - "limitRpd": 80, - "hasCompletions": false, + "limitRpd": null, + "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 162, - "newest": 164, - "throughputHighToLow": 309, - "latencyLowToHigh": 316, - "pricingLowToHigh": 58, - "pricingHighToLow": 306 + "topWeekly": 129, + "newest": 249, + "throughputHighToLow": 36, + "latencyLowToHigh": 74, + "pricingLowToHigh": 70, + "pricingHighToLow": 234 }, - "providers": [] + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", + "providers": [ + { + "name": "Enfer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://enfer.ai/&size=256", + "slug": "enfer", + "quantization": null, + "context": 32768, + "maxCompletionTokens": 16384, + "providerModelId": "mistralai/mistral-7b-instruct-v0-3", + "pricing": { + "prompt": "0.000000028", + "completion": "0.000000054", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "logit_bias", + "logprobs" + ], + "inputCost": 0.03, + "outputCost": 0.05, + "throughput": 113.0205, + "latency": 4302 + }, + { + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 32768, + "maxCompletionTokens": 16384, + "providerModelId": "mistralai/Mistral-7B-Instruct-v0.3", + "pricing": { + "prompt": "0.000000029", + "completion": "0.000000055", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" + ], + "inputCost": 0.03, + "outputCost": 0.06, + "throughput": 105.912, + "latency": 593.5 + }, + { + "name": "NextBit", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nextbit256.com/&size=256", + "slug": "nextBit", + "quantization": "bf16", + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "mistral:7b", + "pricing": { + "prompt": "0.000000029", + "completion": "0.000000058", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "structured_outputs" + ], + "inputCost": 0.03, + "outputCost": 0.06, + "throughput": 78.8245, + "latency": 1221 + }, + { + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": "unknown", + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "mistralai/mistral-7b-instruct", + "pricing": { + "prompt": "0.000000029", + "completion": "0.000000059", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" + ], + "inputCost": 0.03, + "outputCost": 0.06, + "throughput": 142.876, + "latency": 768 + }, + { + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": "unknown", + "context": 32768, + "maxCompletionTokens": 4096, + "providerModelId": "mistralai/Mistral-7B-Instruct-v0.3", + "pricing": { + "prompt": "0.0000002", + "completion": "0.0000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" + ], + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 211.92, + "latency": 283 + } + ] }, { - "slug": "openai/gpt-4-turbo", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-04-09T00:00:00+00:00", + "slug": "openai/gpt-4o-search-preview", + "hfSlug": "", + "updatedAt": "2025-04-22T22:33:59.221907+00:00", + "createdAt": "2025-03-12T22:19:09.996816+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4 Turbo", - "shortName": "GPT-4 Turbo", + "name": "OpenAI: GPT-4o Search Preview", + "shortName": "GPT-4o Search Preview", "author": "openai", - "description": "The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.\n\nTraining data: up to December 2023.", + "description": "GPT-4o Search Previewis a specialized model for web search in Chat Completions. It is trained to understand and execute web search queries.", "modelVersionGroupId": null, "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" @@ -47832,29 +51144,28 @@ export default { "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "openai/gpt-4-turbo", + "warningMessage": "", + "permaslug": "openai/gpt-4o-search-preview-2025-03-11", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "da16824f-3ba0-43a1-86f8-a6131837f457", - "name": "OpenAI | openai/gpt-4-turbo", + "id": "f37536d3-fa09-47a3-b63c-831a1965253e", + "name": "OpenAI | openai/gpt-4o-search-preview-2025-03-11", "contextLength": 128000, "model": { - "slug": "openai/gpt-4-turbo", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-04-09T00:00:00+00:00", + "slug": "openai/gpt-4o-search-preview", + "hfSlug": "", + "updatedAt": "2025-04-22T22:33:59.221907+00:00", + "createdAt": "2025-03-12T22:19:09.996816+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4 Turbo", - "shortName": "GPT-4 Turbo", + "name": "OpenAI: GPT-4o Search Preview", + "shortName": "GPT-4o Search Preview", "author": "openai", - "description": "The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.\n\nTraining data: up to December 2023.", + "description": "GPT-4o Search Previewis a specialized model for web search in Chat Completions. It is trained to understand and execute web search queries.", "modelVersionGroupId": null, "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" @@ -47866,13 +51177,13 @@ export default { "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "openai/gpt-4-turbo", + "warningMessage": "", + "permaslug": "openai/gpt-4o-search-preview-2025-03-11", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/gpt-4-turbo", - "modelVariantPermaslug": "openai/gpt-4-turbo", + "modelVariantSlug": "openai/gpt-4o-search-preview", + "modelVariantPermaslug": "openai/gpt-4o-search-preview-2025-03-11", "providerName": "OpenAI", "providerInfo": { "name": "OpenAI", @@ -47908,30 +51219,21 @@ export default { } }, "providerDisplayName": "OpenAI", - "providerModelId": "gpt-4-turbo", + "providerModelId": "gpt-4o-search-preview-2025-03-11", "providerGroup": "OpenAI", - "quantization": "unknown", + "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": 16384, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", + "web_search_options", "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format" + "response_format", + "structured_outputs" ], "isByok": false, "moderationRequired": true, @@ -47950,223 +51252,31 @@ export default { "retentionDays": 30 }, "pricing": { - "prompt": "0.00001", - "completion": "0.00003", - "image": "0.01445", - "request": "0", + "prompt": "0.0000025", + "completion": "0.00001", + "image": "0.003613", + "request": "0.035", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 163, - "newest": 270, - "throughputHighToLow": 234, - "latencyLowToHigh": 120, - "pricingLowToHigh": 302, - "pricingHighToLow": 14 - }, - "providers": [ - { - "name": "OpenAI", - "slug": "openAi", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 4096, - "providerModelId": "gpt-4-turbo", - "pricing": { - "prompt": "0.00001", - "completion": "0.00003", - "image": "0.01445", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 + "variablePricings": [ + { + "type": "search-threshold", + "threshold": "high", + "request": "0.05" }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format" - ], - "inputCost": 10, - "outputCost": 30 - } - ] - }, - { - "slug": "mistralai/mixtral-8x22b-instruct", - "hfSlug": "mistralai/Mixtral-8x22B-Instruct-v0.1", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-04-17T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Mistral: Mixtral 8x22B Instruct", - "shortName": "Mixtral 8x22B Instruct", - "author": "mistralai", - "description": "Mistral's official instruct fine-tuned version of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b). It uses 39B active parameters out of 141B, offering unparalleled cost efficiency for its size. Its strengths include:\n- strong math, coding, and reasoning\n- large context length (64k)\n- fluency in English, French, Italian, German, and Spanish\n\nSee benchmarks on the launch announcement [here](https://mistral.ai/news/mixtral-8x22b/).\n#moe", - "modelVersionGroupId": null, - "contextLength": 65536, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Mistral", - "instructType": "mistral", - "defaultSystem": null, - "defaultStops": [ - "[INST]", - "" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "mistralai/mixtral-8x22b-instruct", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "c26356d7-ec87-4537-b049-61870b15cf3c", - "name": "Nebius | mistralai/mixtral-8x22b-instruct", - "contextLength": 65536, - "model": { - "slug": "mistralai/mixtral-8x22b-instruct", - "hfSlug": "mistralai/Mixtral-8x22B-Instruct-v0.1", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-04-17T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Mistral: Mixtral 8x22B Instruct", - "shortName": "Mixtral 8x22B Instruct", - "author": "mistralai", - "description": "Mistral's official instruct fine-tuned version of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b). It uses 39B active parameters out of 141B, offering unparalleled cost efficiency for its size. Its strengths include:\n- strong math, coding, and reasoning\n- large context length (64k)\n- fluency in English, French, Italian, German, and Spanish\n\nSee benchmarks on the launch announcement [here](https://mistral.ai/news/mixtral-8x22b/).\n#moe", - "modelVersionGroupId": null, - "contextLength": 65536, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Mistral", - "instructType": "mistral", - "defaultSystem": null, - "defaultStops": [ - "[INST]", - "" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "mistralai/mixtral-8x22b-instruct", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "mistralai/mixtral-8x22b-instruct", - "modelVariantPermaslug": "mistralai/mixtral-8x22b-instruct", - "providerName": "Nebius", - "providerInfo": { - "name": "Nebius", - "displayName": "Nebius AI Studio", - "slug": "nebius", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", - "paidModels": { - "training": false, - "retainsPrompts": false - } + { + "type": "search-threshold", + "threshold": "medium", + "request": "0.035" }, - "headquarters": "DE", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": false, - "moderationRequired": false, - "group": "Nebius", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", - "invertRequired": true + { + "type": "search-threshold", + "threshold": "low", + "request": "0.03" } - }, - "providerDisplayName": "Nebius AI Studio", - "providerModelId": "mistralai/Mixtral-8x22B-Instruct-v0.1", - "providerGroup": "Nebius", - "quantization": "fp8", - "variant": "standard", - "isFree": false, - "canAbort": false, - "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", - "paidModels": { - "training": false, - "retainsPrompts": false - }, - "training": false, - "retainsPrompts": false - }, - "pricing": { - "prompt": "0.0000004", - "completion": "0.0000012", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], "isHidden": false, "isDeranked": false, "isDisabled": false, @@ -48178,134 +51288,62 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 164, - "newest": 268, - "throughputHighToLow": 78, - "latencyLowToHigh": 15, - "pricingLowToHigh": 176, - "pricingHighToLow": 143 + "topWeekly": 167, + "newest": 91, + "throughputHighToLow": 38, + "latencyLowToHigh": 294, + "pricingLowToHigh": 314, + "pricingHighToLow": 4 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 65536, - "maxCompletionTokens": null, - "providerModelId": "mistralai/Mixtral-8x22B-Instruct-v0.1", - "pricing": { - "prompt": "0.0000004", - "completion": "0.0000012", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" - ], - "inputCost": 0.4, - "outputCost": 1.2 - }, - { - "name": "Fireworks", - "slug": "fireworks", - "quantization": "unknown", - "context": 65536, - "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/mixtral-8x22b-instruct", - "pricing": { - "prompt": "0.0000009", - "completion": "0.0000009", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "response_format", - "structured_outputs", - "logit_bias", - "logprobs", - "top_logprobs" - ], - "inputCost": 0.9, - "outputCost": 0.9 - }, - { - "name": "Mistral", - "slug": "mistral", - "quantization": "unknown", - "context": 65536, - "maxCompletionTokens": null, - "providerModelId": "open-mixtral-8x22b", + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": null, + "context": 128000, + "maxCompletionTokens": 16384, + "providerModelId": "gpt-4o-search-preview-2025-03-11", "pricing": { - "prompt": "0.000002", - "completion": "0.000006", - "image": "0", - "request": "0", + "prompt": "0.0000025", + "completion": "0.00001", + "image": "0.003613", + "request": "0.035", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", + "web_search_options", "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", "response_format", - "structured_outputs", - "seed" + "structured_outputs" ], - "inputCost": 2, - "outputCost": 6 + "inputCost": 2.5, + "outputCost": 10, + "throughput": 145.686, + "latency": 4742 } ] }, { - "slug": "qwen/qwen-2-72b-instruct", - "hfSlug": "Qwen/Qwen2-72B-Instruct", + "slug": "perplexity/llama-3.1-sonar-small-128k-online", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-06-07T00:00:00+00:00", + "createdAt": "2024-08-01T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen 2 72B Instruct", - "shortName": "Qwen 2 72B Instruct", - "author": "qwen", - "description": "Qwen2 72B is a transformer-based model that excels in language understanding, multilingual capabilities, coding, mathematics, and reasoning.\n\nIt features SwiGLU activation, attention QKV bias, and group query attention. It is pretrained on extensive data with supervised finetuning and direct preference optimization.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2/) and [GitHub repo](https://github.com/QwenLM/Qwen2).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", - "modelVersionGroupId": null, - "contextLength": 32768, + "name": "Perplexity: Llama 3.1 Sonar 8B Online", + "shortName": "Llama 3.1 Sonar 8B Online", + "author": "perplexity", + "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3.1-sonar-small-128k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", + "modelVersionGroupId": "9e4cc81b-5f14-4987-9e40-74db1c86ecda", + "contextLength": 127072, "inputModalities": [ "text" ], @@ -48313,36 +51351,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", + "group": "Llama3", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-2-72b-instruct", + "permaslug": "perplexity/llama-3.1-sonar-small-128k-online", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "89dce4f1-1836-4539-88f3-ee86d594d7a5", - "name": "Together | qwen/qwen-2-72b-instruct", - "contextLength": 32768, + "id": "48b44b03-a839-42e9-863e-1b6c07aa42ae", + "name": "Perplexity | perplexity/llama-3.1-sonar-small-128k-online", + "contextLength": 127072, "model": { - "slug": "qwen/qwen-2-72b-instruct", - "hfSlug": "Qwen/Qwen2-72B-Instruct", + "slug": "perplexity/llama-3.1-sonar-small-128k-online", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-06-07T00:00:00+00:00", + "createdAt": "2024-08-01T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen 2 72B Instruct", - "shortName": "Qwen 2 72B Instruct", - "author": "qwen", - "description": "Qwen2 72B is a transformer-based model that excels in language understanding, multilingual capabilities, coding, mathematics, and reasoning.\n\nIt features SwiGLU activation, attention QKV bias, and group query attention. It is pretrained on extensive data with supervised finetuning and direct preference optimization.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2/) and [GitHub repo](https://github.com/QwenLM/Qwen2).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", - "modelVersionGroupId": null, - "contextLength": 32768, + "name": "Perplexity: Llama 3.1 Sonar 8B Online", + "shortName": "Llama 3.1 Sonar 8B Online", + "author": "perplexity", + "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3.1-sonar-small-128k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", + "modelVersionGroupId": "9e4cc81b-5f14-4987-9e40-74db1c86ecda", + "contextLength": 127072, "inputModalities": [ "text" ], @@ -48350,94 +51384,82 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", + "group": "Llama3", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-2-72b-instruct", + "permaslug": "perplexity/llama-3.1-sonar-small-128k-online", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "qwen/qwen-2-72b-instruct", - "modelVariantPermaslug": "qwen/qwen-2-72b-instruct", - "providerName": "Together", + "modelVariantSlug": "perplexity/llama-3.1-sonar-small-128k-online", + "modelVariantPermaslug": "perplexity/llama-3.1-sonar-small-128k-online", + "providerName": "Perplexity", "providerInfo": { - "name": "Together", - "displayName": "Together", - "slug": "together", + "name": "Perplexity", + "displayName": "Perplexity", + "slug": "perplexity", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", + "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": false + "training": false } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, + "hasCompletions": false, + "isAbortable": false, "moderationRequired": false, - "group": "Together", + "group": "Perplexity", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.perplexity.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" + "url": "/images/icons/Perplexity.svg" } - }, - "providerDisplayName": "Together", - "providerModelId": "Qwen/Qwen2-72B-Instruct", - "providerGroup": "Together", + }, + "providerDisplayName": "Perplexity", + "providerModelId": "llama-3.1-sonar-small-128k-online", + "providerGroup": "Perplexity", "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "frequency_penalty", + "presence_penalty" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", + "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": false + "training": false }, - "training": false, - "retainsPrompts": false + "training": false }, "pricing": { - "prompt": "0.0000009", - "completion": "0.0000009", + "prompt": "0.0000002", + "completion": "0.0000002", "image": "0", - "request": "0", + "request": "0.005", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -48451,7 +51473,7 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -48459,26 +51481,28 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 165, - "newest": 248, - "throughputHighToLow": 212, - "latencyLowToHigh": 95, - "pricingLowToHigh": 215, - "pricingHighToLow": 104 + "topWeekly": 168, + "newest": 227, + "throughputHighToLow": 8, + "latencyLowToHigh": 193, + "pricingLowToHigh": 286, + "pricingHighToLow": 32 }, + "authorIcon": "https://openrouter.ai/images/icons/Perplexity.svg", "providers": [ { - "name": "Together", - "slug": "together", + "name": "Perplexity", + "icon": "https://openrouter.ai/images/icons/Perplexity.svg", + "slug": "perplexity", "quantization": "unknown", - "context": 32768, - "maxCompletionTokens": 4096, - "providerModelId": "Qwen/Qwen2-72B-Instruct", + "context": 127072, + "maxCompletionTokens": null, + "providerModelId": "llama-3.1-sonar-small-128k-online", "pricing": { - "prompt": "0.0000009", - "completion": "0.0000009", + "prompt": "0.0000002", + "completion": "0.0000002", "image": "0", - "request": "0", + "request": "0.005", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -48487,17 +51511,14 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "frequency_penalty", + "presence_penalty" ], - "inputCost": 0.9, - "outputCost": 0.9 + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 240.3005, + "latency": 1296 } ] }, @@ -48663,16 +51684,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 135, - "newest": 279, - "throughputHighToLow": 121, + "topWeekly": 134, + "newest": 281, + "throughputHighToLow": 116, "latencyLowToHigh": 128, "pricingLowToHigh": 278, "pricingHighToLow": 50 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", "slug": "anthropic", "quantization": "unknown", "context": 200000, @@ -48699,10 +51722,13 @@ export default { "stop" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 66.9785, + "latency": 754 }, { "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", "slug": "vertex", "quantization": "unknown", "context": 200000, @@ -48729,176 +51755,23 @@ export default { "stop" ], "inputCost": 3, - "outputCost": 15 + "outputCost": 15, + "throughput": 68.034, + "latency": 1041 } ] }, { - "slug": "opengvlab/internvl3-14b", - "hfSlug": "OpenGVLab/InternVL3-14B", - "updatedAt": "2025-04-30T14:14:32.387955+00:00", - "createdAt": "2025-04-30T13:55:55.014183+00:00", - "hfUpdatedAt": null, - "name": "OpenGVLab: InternVL3 14B (free)", - "shortName": "InternVL3 14B (free)", - "author": "opengvlab", - "description": "The 14b version of the InternVL3 series. An advanced multimodal large language model (MLLM) series that demonstrates superior overall performance. Compared to InternVL 2.5, InternVL3 exhibits superior multimodal perception and reasoning capabilities, while further extending its multimodal capabilities to encompass tool usage, GUI agents, industrial image analysis, 3D vision perception, and more.", - "modelVersionGroupId": null, - "contextLength": 32000, - "inputModalities": [ - "image", - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "opengvlab/internvl3-14b", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "3be27401-83ab-4e46-9418-462abbc1c8af", - "name": "Nineteen | opengvlab/internvl3-14b:free", - "contextLength": 32000, - "model": { - "slug": "opengvlab/internvl3-14b", - "hfSlug": "OpenGVLab/InternVL3-14B", - "updatedAt": "2025-04-30T14:14:32.387955+00:00", - "createdAt": "2025-04-30T13:55:55.014183+00:00", - "hfUpdatedAt": null, - "name": "OpenGVLab: InternVL3 14B", - "shortName": "InternVL3 14B", - "author": "opengvlab", - "description": "The 14b version of the InternVL3 series. An advanced multimodal large language model (MLLM) series that demonstrates superior overall performance. Compared to InternVL 2.5, InternVL3 exhibits superior multimodal perception and reasoning capabilities, while further extending its multimodal capabilities to encompass tool usage, GUI agents, industrial image analysis, 3D vision perception, and more.", - "modelVersionGroupId": null, - "contextLength": 32000, - "inputModalities": [ - "image", - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "opengvlab/internvl3-14b", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "opengvlab/internvl3-14b:free", - "modelVariantPermaslug": "opengvlab/internvl3-14b:free", - "providerName": "Nineteen", - "providerInfo": { - "name": "Nineteen", - "displayName": "Nineteen", - "slug": "nineteen", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://nineteen.ai/tos", - "paidModels": { - "training": false - } - }, - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "Nineteen", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nineteen.ai/&size=256" - } - }, - "providerDisplayName": "Nineteen", - "providerModelId": "OpenGVLab/InternVL3-14B", - "providerGroup": "Nineteen", - "quantization": "bf16", - "variant": "free", - "isFree": true, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://nineteen.ai/tos", - "paidModels": { - "training": false - }, - "training": false - }, - "pricing": { - "prompt": "0", - "completion": "0", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 167, - "newest": 17, - "throughputHighToLow": 52, - "latencyLowToHigh": 189, - "pricingLowToHigh": 6, - "pricingHighToLow": 254 - }, - "providers": [] - }, - { - "slug": "nvidia/llama-3.3-nemotron-super-49b-v1", - "hfSlug": "nvidia/Llama-3_3-Nemotron-Super-49B-v1", - "updatedAt": "2025-04-08T14:53:49.407401+00:00", - "createdAt": "2025-04-08T13:38:14.544231+00:00", + "slug": "mistralai/mistral-large-2407", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-11-19T01:06:55.27469+00:00", "hfUpdatedAt": null, - "name": "NVIDIA: Llama 3.3 Nemotron Super 49B v1", - "shortName": "Llama 3.3 Nemotron Super 49B v1", - "author": "nvidia", - "description": "Llama-3.3-Nemotron-Super-49B-v1 is a large language model (LLM) optimized for advanced reasoning, conversational interactions, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta's Llama-3.3-70B-Instruct, it employs a Neural Architecture Search (NAS) approach, significantly enhancing efficiency and reducing memory requirements. This allows the model to support a context length of up to 128K tokens and fit efficiently on single high-performance GPUs, such as NVIDIA H200.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.", - "modelVersionGroupId": null, + "name": "Mistral Large 2407", + "shortName": "Mistral Large 2407", + "author": "mistralai", + "description": "This is Mistral AI's flagship model, Mistral Large 2 (version mistral-large-2407). It's a proprietary weights-available model and excels at reasoning, code, JSON, chat, and more. Read the launch announcement [here](https://mistral.ai/news/mistral-large-2407/).\n\nIt supports dozens of languages including French, German, Spanish, Italian, Portuguese, Arabic, Hindi, Russian, Chinese, Japanese, and Korean, along with 80+ coding languages including Python, Java, C, C++, JavaScript, and Bash. Its long context window allows precise information recall from large documents.\n", + "modelVersionGroupId": "83129748-0564-4485-982a-d7a37a1ef3ec", "contextLength": 131072, "inputModalities": [ "text" @@ -48907,32 +51780,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "Mistral", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "nvidia/llama-3.3-nemotron-super-49b-v1", + "permaslug": "mistralai/mistral-large-2407", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "1b7717f4-53d6-4a42-96ea-a58872ef08a0", - "name": "Nebius | nvidia/llama-3.3-nemotron-super-49b-v1", + "id": "4a128170-b056-42d7-8462-a5cea647f9ad", + "name": "Mistral | mistralai/mistral-large-2407", "contextLength": 131072, "model": { - "slug": "nvidia/llama-3.3-nemotron-super-49b-v1", - "hfSlug": "nvidia/Llama-3_3-Nemotron-Super-49B-v1", - "updatedAt": "2025-04-08T14:53:49.407401+00:00", - "createdAt": "2025-04-08T13:38:14.544231+00:00", + "slug": "mistralai/mistral-large-2407", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-11-19T01:06:55.27469+00:00", "hfUpdatedAt": null, - "name": "NVIDIA: Llama 3.3 Nemotron Super 49B v1", - "shortName": "Llama 3.3 Nemotron Super 49B v1", - "author": "nvidia", - "description": "Llama-3.3-Nemotron-Super-49B-v1 is a large language model (LLM) optimized for advanced reasoning, conversational interactions, retrieval-augmented generation (RAG), and tool-calling tasks. Derived from Meta's Llama-3.3-70B-Instruct, it employs a Neural Architecture Search (NAS) approach, significantly enhancing efficiency and reducing memory requirements. This allows the model to support a context length of up to 128K tokens and fit efficiently on single high-performance GPUs, such as NVIDIA H200.\n\nNote: you must include `detailed thinking on` in the system prompt to enable reasoning. Please see [Usage Recommendations](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1#quick-start-and-usage-recommendations) for more.", - "modelVersionGroupId": null, - "contextLength": 131072, + "name": "Mistral Large 2407", + "shortName": "Mistral Large 2407", + "author": "mistralai", + "description": "This is Mistral AI's flagship model, Mistral Large 2 (version mistral-large-2407). It's a proprietary weights-available model and excels at reasoning, code, JSON, chat, and more. Read the launch announcement [here](https://mistral.ai/news/mistral-large-2407/).\n\nIt supports dozens of languages including French, German, Spanish, Italian, Portuguese, Arabic, Hindi, Russian, Chinese, Japanese, and Korean, along with 80+ coding languages including Python, Java, C, C++, JavaScript, and Bash. Its long context window allows precise information recall from large documents.\n", + "modelVersionGroupId": "83129748-0564-4485-982a-d7a37a1ef3ec", + "contextLength": 128000, "inputModalities": [ "text" ], @@ -48940,39 +51813,40 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "Mistral", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "nvidia/llama-3.3-nemotron-super-49b-v1", + "permaslug": "mistralai/mistral-large-2407", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "nvidia/llama-3.3-nemotron-super-49b-v1", - "modelVariantPermaslug": "nvidia/llama-3.3-nemotron-super-49b-v1", - "providerName": "Nebius", + "modelVariantSlug": "mistralai/mistral-large-2407", + "modelVariantPermaslug": "mistralai/mistral-large-2407", + "providerName": "Mistral", "providerInfo": { - "name": "Nebius", - "displayName": "Nebius AI Studio", - "slug": "nebius", + "name": "Mistral", + "displayName": "Mistral", + "slug": "mistral", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 } }, - "headquarters": "DE", + "headquarters": "FR", "hasChatCompletions": true, - "hasCompletions": true, + "hasCompletions": false, "isAbortable": false, "moderationRequired": false, - "group": "Nebius", + "group": "Mistral", "editors": [], "owners": [], "isMultipartSupported": true, @@ -48980,14 +51854,13 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", - "invertRequired": true + "url": "/images/icons/Mistral.png" } }, - "providerDisplayName": "Nebius AI Studio", - "providerModelId": "nvidia/Llama-3_3-Nemotron-Super-49B-v1", - "providerGroup": "Nebius", - "quantization": "fp8", + "providerDisplayName": "Mistral", + "providerModelId": "mistral-large-2407", + "providerGroup": "Mistral", + "quantization": null, "variant": "standard", "isFree": false, "canAbort": false, @@ -48996,33 +51869,35 @@ export default { "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" + "response_format", + "structured_outputs", + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.00000013", - "completion": "0.0000004", + "prompt": "0.000002", + "completion": "0.000006", "image": "0", "request": "0", "webSearch": "0", @@ -49033,12 +51908,12 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { "supportedParameters": {}, @@ -49047,24 +51922,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 150, - "newest": 58, - "throughputHighToLow": 181, - "latencyLowToHigh": 21, - "pricingLowToHigh": 24, - "pricingHighToLow": 179 + "topWeekly": 170, + "newest": 168, + "throughputHighToLow": 207, + "latencyLowToHigh": 67, + "pricingLowToHigh": 244, + "pricingHighToLow": 73 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", + "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", + "slug": "mistral", + "quantization": null, "context": 131072, "maxCompletionTokens": null, - "providerModelId": "nvidia/Llama-3_3-Nemotron-Super-49B-v1", + "providerModelId": "mistral-large-2407", "pricing": { - "prompt": "0.00000013", - "completion": "0.0000004", + "prompt": "0.000002", + "completion": "0.000006", "image": "0", "request": "0", "webSearch": "0", @@ -49072,20 +51949,22 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" + "response_format", + "structured_outputs", + "seed" ], - "inputCost": 0.13, - "outputCost": 0.4 + "inputCost": 2, + "outputCost": 6, + "throughput": 40.9275, + "latency": 491 } ] }, @@ -49227,8 +52106,8 @@ export default { "retainsPrompts": true }, "pricing": { - "prompt": "0.00000009375", - "completion": "0.00000075", + "prompt": "0.00000015", + "completion": "0.0000009375", "image": "0", "request": "0", "webSearch": "0", @@ -49252,24 +52131,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 169, + "topWeekly": 171, "newest": 206, - "throughputHighToLow": 244, - "latencyLowToHigh": 81, - "pricingLowToHigh": 135, - "pricingHighToLow": 181 + "throughputHighToLow": 241, + "latencyLowToHigh": 84, + "pricingLowToHigh": 151, + "pricingHighToLow": 165 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "Mancer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", "slug": "mancer", "quantization": "fp16", "context": 32768, "maxCompletionTokens": 2048, "providerModelId": "lumi-8b-v2", "pricing": { - "prompt": "0.00000009375", - "completion": "0.00000075", + "prompt": "0.00000015", + "completion": "0.0000009375", "image": "0", "request": "0", "webSearch": "0", @@ -49290,19 +52171,22 @@ export default { "seed", "top_a" ], - "inputCost": 0.09, - "outputCost": 0.75 + "inputCost": 0.15, + "outputCost": 0.94, + "throughput": 31.739, + "latency": 568.5 }, { "name": "Mancer (private)", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", "slug": "mancer (private)", "quantization": "fp16", "context": 32768, "maxCompletionTokens": 2048, "providerModelId": "lumi-8b-v2", "pricing": { - "prompt": "0.000000125", - "completion": "0.000001", + "prompt": "0.0000002", + "completion": "0.00000125", "image": "0", "request": "0", "webSearch": "0", @@ -49323,11 +52207,14 @@ export default { "seed", "top_a" ], - "inputCost": 0.13, - "outputCost": 1 + "inputCost": 0.2, + "outputCost": 1.25, + "throughput": 31.676, + "latency": 369 }, { "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", "slug": "featherless", "quantization": "fp8", "context": 16384, @@ -49355,22 +52242,24 @@ export default { "seed" ], "inputCost": 0.8, - "outputCost": 1.2 + "outputCost": 1.2, + "throughput": 26.631, + "latency": 1179 } ] }, { - "slug": "mistralai/mistral-7b-instruct", - "hfSlug": "mistralai/Mistral-7B-Instruct-v0.3", + "slug": "google/gemma-2-27b-it", + "hfSlug": "google/gemma-2-27b-it", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-05-27T00:00:00+00:00", + "createdAt": "2024-07-13T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral 7B Instruct (free)", - "shortName": "Mistral 7B Instruct (free)", - "author": "mistralai", - "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*", - "modelVersionGroupId": "1d07cc56-c54d-4587-b785-5093496397a4", - "contextLength": 32768, + "name": "Google: Gemma 2 27B", + "shortName": "Gemma 2 27B", + "author": "google", + "description": "Gemma 2 27B by Google is an open model built from the same research and technology used to create the [Gemini models](/models?q=gemini).\n\nGemma models are well-suited for a variety of text generation tasks, including question answering, summarization, and reasoning.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).", + "modelVersionGroupId": null, + "contextLength": 8192, "inputModalities": [ "text" ], @@ -49378,35 +52267,36 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "mistral", + "group": "Gemini", + "instructType": "gemma", "defaultSystem": null, "defaultStops": [ - "[INST]", - "" + "", + "", + "" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-7b-instruct", + "permaslug": "google/gemma-2-27b-it", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "d09ade5a-2aaf-4a93-9881-3db42f1ba63a", - "name": "DeepInfra | mistralai/mistral-7b-instruct:free", - "contextLength": 32768, + "id": "2ddecefa-fddf-44bc-b8db-fa904f993eef", + "name": "Nebius | google/gemma-2-27b-it", + "contextLength": 8192, "model": { - "slug": "mistralai/mistral-7b-instruct", - "hfSlug": "mistralai/Mistral-7B-Instruct-v0.3", + "slug": "google/gemma-2-27b-it", + "hfSlug": "google/gemma-2-27b-it", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-05-27T00:00:00+00:00", + "createdAt": "2024-07-13T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral 7B Instruct", - "shortName": "Mistral 7B Instruct", - "author": "mistralai", - "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\n*Mistral 7B Instruct has multiple version variants, and this is intended to be the latest version.*", - "modelVersionGroupId": "1d07cc56-c54d-4587-b785-5093496397a4", - "contextLength": 32768, + "name": "Google: Gemma 2 27B", + "shortName": "Gemma 2 27B", + "author": "google", + "description": "Gemma 2 27B by Google is an open model built from the same research and technology used to create the [Gemini models](/models?q=gemini).\n\nGemma models are well-suited for a variety of text generation tasks, including question answering, summarization, and reasoning.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).", + "modelVersionGroupId": null, + "contextLength": 8192, "inputModalities": [ "text" ], @@ -49414,43 +52304,43 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "mistral", + "group": "Gemini", + "instructType": "gemma", "defaultSystem": null, "defaultStops": [ - "[INST]", - "" + "", + "", + "" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-7b-instruct", + "permaslug": "google/gemma-2-27b-it", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "mistralai/mistral-7b-instruct:free", - "modelVariantPermaslug": "mistralai/mistral-7b-instruct:free", - "providerName": "DeepInfra", + "modelVariantSlug": "google/gemma-2-27b-it", + "modelVariantPermaslug": "google/gemma-2-27b-it", + "providerName": "Nebius", "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", + "name": "Nebius", + "displayName": "Nebius AI Studio", + "slug": "nebius", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", "paidModels": { "training": false, "retainsPrompts": false } }, - "headquarters": "US", + "headquarters": "DE", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": true, + "isAbortable": false, "moderationRequired": false, - "group": "DeepInfra", + "group": "Nebius", "editors": [], "owners": [], "isMultipartSupported": true, @@ -49458,41 +52348,39 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/DeepInfra.webp" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "invertRequired": true } }, - "providerDisplayName": "DeepInfra", - "providerModelId": "mistralai/Mistral-7B-Instruct-v0.3", - "providerGroup": "DeepInfra", - "quantization": "bf16", - "variant": "free", - "isFree": true, - "canAbort": true, + "providerDisplayName": "Nebius AI Studio", + "providerModelId": "google/gemma-2-27b-it", + "providerGroup": "Nebius", + "quantization": "fp8", + "variant": "standard", + "isFree": false, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 16384, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", + "frequency_penalty", + "presence_penalty", "seed", - "min_p" + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", "paidModels": { "training": false, "retainsPrompts": false @@ -49501,8 +52389,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.0000001", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -49513,7 +52401,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -49527,116 +52415,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 97, - "newest": 251, - "throughputHighToLow": 66, - "latencyLowToHigh": 143, - "pricingLowToHigh": 70, - "pricingHighToLow": 235 + "topWeekly": 172, + "newest": 238, + "throughputHighToLow": 186, + "latencyLowToHigh": 57, + "pricingLowToHigh": 124, + "pricingHighToLow": 195 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [ { - "name": "Enfer", - "slug": "enfer", - "quantization": null, - "context": 32768, - "maxCompletionTokens": 16384, - "providerModelId": "mistralai/mistral-7b-instruct-v0-3", - "pricing": { - "prompt": "0.000000028", - "completion": "0.000000054", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "logit_bias", - "logprobs" - ], - "inputCost": 0.03, - "outputCost": 0.05 - }, - { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": 16384, - "providerModelId": "mistralai/Mistral-7B-Instruct-v0.3", - "pricing": { - "prompt": "0.000000029", - "completion": "0.000000055", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "inputCost": 0.03, - "outputCost": 0.06 - }, - { - "name": "NextBit", - "slug": "nextBit", - "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "mistral:7b", - "pricing": { - "prompt": "0.000000029", - "completion": "0.000000058", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "structured_outputs" - ], - "inputCost": 0.03, - "outputCost": 0.06 - }, - { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": "unknown", - "context": 32768, + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 8192, "maxCompletionTokens": null, - "providerModelId": "mistralai/mistral-7b-instruct", + "providerModelId": "google/gemma-2-27b-it", "pricing": { - "prompt": "0.000000029", - "completion": "0.000000059", + "prompt": "0.0000001", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -49652,51 +52450,26 @@ export default { "presence_penalty", "seed", "top_k", - "min_p", - "repetition_penalty", - "logit_bias" - ], - "inputCost": 0.03, - "outputCost": 0.06 - }, - { - "name": "Parasail", - "slug": "parasail", - "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": 32768, - "providerModelId": "parasail-mistral-7b-instruct-03", - "pricing": { - "prompt": "0.00000011", - "completion": "0.00000011", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "presence_penalty", - "frequency_penalty", - "repetition_penalty", - "top_k" + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 0.11, - "outputCost": 0.11 + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 45.5765, + "latency": 488 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "unknown", - "context": 32768, - "maxCompletionTokens": 4096, - "providerModelId": "mistralai/Mistral-7B-Instruct-v0.3", + "context": 8192, + "maxCompletionTokens": 2048, + "providerModelId": "google/gemma-2-27b-it", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", + "prompt": "0.0000008", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -49716,213 +52489,204 @@ export default { "min_p", "response_format" ], - "inputCost": 0.2, - "outputCost": 0.2 + "inputCost": 0.8, + "outputCost": 0.8, + "throughput": 70.401, + "latency": 496 } ] }, { - "slug": "qwen/qwen3-8b", - "hfSlug": "Qwen/Qwen3-8B", - "updatedAt": "2025-05-12T00:35:52.624106+00:00", - "createdAt": "2025-04-28T21:43:52.421936+00:00", + "slug": "perplexity/sonar-pro", + "hfSlug": "", + "updatedAt": "2025-05-05T14:37:56.196383+00:00", + "createdAt": "2025-03-07T01:53:43+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 8B (free)", - "shortName": "Qwen3 8B (free)", - "author": "qwen", - "description": "Qwen3-8B is a dense 8.2B parameter causal language model from the Qwen3 series, designed for both reasoning-heavy tasks and efficient dialogue. It supports seamless switching between \"thinking\" mode for math, coding, and logical inference, and \"non-thinking\" mode for general conversation. The model is fine-tuned for instruction-following, agent integration, creative writing, and multilingual use across 100+ languages and dialects. It natively supports a 32K token context window and can extend to 131K tokens with YaRN scaling.", + "name": "Perplexity: Sonar Pro", + "shortName": "Sonar Pro", + "author": "perplexity", + "description": "Note: Sonar Pro pricing includes Perplexity search pricing. See [details here](https://docs.perplexity.ai/guides/pricing#detailed-pricing-breakdown-for-sonar-reasoning-pro-and-sonar-pro)\n\nFor enterprises seeking more advanced capabilities, the Sonar Pro API can handle in-depth, multi-step queries with added extensibility, like double the number of citations per search as Sonar on average. Plus, with a larger context window, it can handle longer and more nuanced searches and follow-up questions. ", "modelVersionGroupId": null, - "contextLength": 40960, + "contextLength": 200000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>" - ], + "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "qwen/qwen3-8b-04-28", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "warningMessage": "", + "permaslug": "perplexity/sonar-pro", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "3ce04781-3def-489d-a591-a32f0d012971", - "name": "Chutes | qwen/qwen3-8b-04-28:free", - "contextLength": 40960, + "id": "e19b1036-fab9-4f97-9579-8ea67959cc9b", + "name": "Perplexity | perplexity/sonar-pro", + "contextLength": 200000, "model": { - "slug": "qwen/qwen3-8b", - "hfSlug": "Qwen/Qwen3-8B", - "updatedAt": "2025-05-12T00:35:52.624106+00:00", - "createdAt": "2025-04-28T21:43:52.421936+00:00", + "slug": "perplexity/sonar-pro", + "hfSlug": "", + "updatedAt": "2025-05-05T14:37:56.196383+00:00", + "createdAt": "2025-03-07T01:53:43+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 8B", - "shortName": "Qwen3 8B", - "author": "qwen", - "description": "Qwen3-8B is a dense 8.2B parameter causal language model from the Qwen3 series, designed for both reasoning-heavy tasks and efficient dialogue. It supports seamless switching between \"thinking\" mode for math, coding, and logical inference, and \"non-thinking\" mode for general conversation. The model is fine-tuned for instruction-following, agent integration, creative writing, and multilingual use across 100+ languages and dialects. It natively supports a 32K token context window and can extend to 131K tokens with YaRN scaling.", + "name": "Perplexity: Sonar Pro", + "shortName": "Sonar Pro", + "author": "perplexity", + "description": "Note: Sonar Pro pricing includes Perplexity search pricing. See [details here](https://docs.perplexity.ai/guides/pricing#detailed-pricing-breakdown-for-sonar-reasoning-pro-and-sonar-pro)\n\nFor enterprises seeking more advanced capabilities, the Sonar Pro API can handle in-depth, multi-step queries with added extensibility, like double the number of citations per search as Sonar on average. Plus, with a larger context window, it can handle longer and more nuanced searches and follow-up questions. ", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 200000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>" - ], + "defaultStops": [], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "qwen/qwen3-8b-04-28", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "warningMessage": "", + "permaslug": "perplexity/sonar-pro", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "qwen/qwen3-8b:free", - "modelVariantPermaslug": "qwen/qwen3-8b-04-28:free", - "providerName": "Chutes", + "modelVariantSlug": "perplexity/sonar-pro", + "modelVariantPermaslug": "perplexity/sonar-pro", + "providerName": "Perplexity", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "Perplexity", + "displayName": "Perplexity", + "slug": "perplexity", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", + "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false } }, + "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, + "hasCompletions": false, + "isAbortable": false, "moderationRequired": false, - "group": "Chutes", + "group": "Perplexity", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.perplexity.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "/images/icons/Perplexity.svg" } }, - "providerDisplayName": "Chutes", - "providerModelId": "Qwen/Qwen3-8B", - "providerGroup": "Chutes", + "providerDisplayName": "Perplexity", + "providerModelId": "sonar-pro", + "providerGroup": "Perplexity", "quantization": null, - "variant": "free", - "isFree": true, - "canAbort": true, + "variant": "standard", + "isFree": false, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 40960, + "maxCompletionTokens": 8000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", + "web_search_options", "top_k", - "min_p", - "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" + "frequency_penalty", + "presence_penalty" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", + "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false }, - "training": true, - "retainsPrompts": true + "training": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.000003", + "completion": "0.000015", "image": "0", "request": "0", - "webSearch": "0", + "webSearch": "0.005", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [], + "variablePricings": [ + { + "type": "search-threshold", + "threshold": "high", + "request": "0.014" + }, + { + "type": "search-threshold", + "threshold": "medium", + "request": "0.01" + }, + { + "type": "search-threshold", + "threshold": "low", + "request": "0.006" + } + ], "isHidden": false, "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 96, - "newest": 24, - "throughputHighToLow": 27, - "latencyLowToHigh": 116, - "pricingLowToHigh": 10, - "pricingHighToLow": 229 + "topWeekly": 173, + "newest": 99, + "throughputHighToLow": 150, + "latencyLowToHigh": 277, + "pricingLowToHigh": 268, + "pricingHighToLow": 40 }, + "authorIcon": "https://openrouter.ai/images/icons/Perplexity.svg", "providers": [ { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": "fp8", - "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "qwen/qwen3-8b-fp8", + "name": "Perplexity", + "icon": "https://openrouter.ai/images/icons/Perplexity.svg", + "slug": "perplexity", + "quantization": null, + "context": 200000, + "maxCompletionTokens": 8000, + "providerModelId": "sonar-pro", "pricing": { - "prompt": "0.000000035", - "completion": "0.000000138", + "prompt": "0.000003", + "completion": "0.000015", "image": "0", "request": "0", - "webSearch": "0", + "webSearch": "0.005", "internalReasoning": "0", "discount": 0 }, @@ -49930,34 +52694,30 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", + "web_search_options", "top_k", - "min_p", - "repetition_penalty", - "logit_bias" + "frequency_penalty", + "presence_penalty" ], - "inputCost": 0.04, - "outputCost": 0.14 + "inputCost": 3, + "outputCost": 15, + "throughput": 57.4775, + "latency": 2631.5 } ] }, { - "slug": "deepseek/deepseek-r1-distill-qwen-32b", - "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-01-29T23:53:50.865297+00:00", + "slug": "qwen/qwen3-8b", + "hfSlug": "Qwen/Qwen3-8B", + "updatedAt": "2025-05-12T00:35:52.624106+00:00", + "createdAt": "2025-04-28T21:43:52.421936+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: R1 Distill Qwen 32B (free)", - "shortName": "R1 Distill Qwen 32B (free)", - "author": "deepseek", - "description": "DeepSeek R1 Distill Qwen 32B is a distilled large language model based on [Qwen 2.5 32B](https://huggingface.co/Qwen/Qwen2.5-32B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 72.6\n- MATH-500 pass@1: 94.3\n- CodeForces Rating: 1691\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", - "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", - "contextLength": 16000, + "name": "Qwen: Qwen3 8B (free)", + "shortName": "Qwen3 8B (free)", + "author": "qwen", + "description": "Qwen3-8B is a dense 8.2B parameter causal language model from the Qwen3 series, designed for both reasoning-heavy tasks and efficient dialogue. It supports seamless switching between \"thinking\" mode for math, coding, and logical inference, and \"non-thinking\" mode for general conversation. The model is fine-tuned for instruction-following, agent integration, creative writing, and multilingual use across 100+ languages and dialects. It natively supports a 32K token context window and can extend to 131K tokens with YaRN scaling.", + "modelVersionGroupId": null, + "contextLength": 40960, "inputModalities": [ "text" ], @@ -49965,17 +52725,17 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "deepseek-r1", + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" + "<|im_start|>", + "<|im_end|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "deepseek/deepseek-r1-distill-qwen-32b", + "permaslug": "qwen/qwen3-8b-04-28", "reasoningConfig": { "startToken": "", "endToken": "" @@ -49987,21 +52747,21 @@ export default { } }, "endpoint": { - "id": "749d8a55-ab77-43a8-aa1a-4d0331c2e738", - "name": "Nineteen | deepseek/deepseek-r1-distill-qwen-32b:free", - "contextLength": 16000, + "id": "3ce04781-3def-489d-a591-a32f0d012971", + "name": "Chutes | qwen/qwen3-8b-04-28:free", + "contextLength": 40960, "model": { - "slug": "deepseek/deepseek-r1-distill-qwen-32b", - "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-01-29T23:53:50.865297+00:00", + "slug": "qwen/qwen3-8b", + "hfSlug": "Qwen/Qwen3-8B", + "updatedAt": "2025-05-12T00:35:52.624106+00:00", + "createdAt": "2025-04-28T21:43:52.421936+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: R1 Distill Qwen 32B", - "shortName": "R1 Distill Qwen 32B", - "author": "deepseek", - "description": "DeepSeek R1 Distill Qwen 32B is a distilled large language model based on [Qwen 2.5 32B](https://huggingface.co/Qwen/Qwen2.5-32B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 72.6\n- MATH-500 pass@1: 94.3\n- CodeForces Rating: 1691\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", - "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", - "contextLength": 128000, + "name": "Qwen: Qwen3 8B", + "shortName": "Qwen3 8B", + "author": "qwen", + "description": "Qwen3-8B is a dense 8.2B parameter causal language model from the Qwen3 series, designed for both reasoning-heavy tasks and efficient dialogue. It supports seamless switching between \"thinking\" mode for math, coding, and logical inference, and \"non-thinking\" mode for general conversation. The model is fine-tuned for instruction-following, agent integration, creative writing, and multilingual use across 100+ languages and dialects. It natively supports a 32K token context window and can extend to 131K tokens with YaRN scaling.", + "modelVersionGroupId": null, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -50009,17 +52769,17 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "deepseek-r1", + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" + "<|im_start|>", + "<|im_end|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "deepseek/deepseek-r1-distill-qwen-32b", + "permaslug": "qwen/qwen3-8b-04-28", "reasoningConfig": { "startToken": "", "endToken": "" @@ -50031,25 +52791,26 @@ export default { } } }, - "modelVariantSlug": "deepseek/deepseek-r1-distill-qwen-32b:free", - "modelVariantPermaslug": "deepseek/deepseek-r1-distill-qwen-32b:free", - "providerName": "Nineteen", + "modelVariantSlug": "qwen/qwen3-8b:free", + "modelVariantPermaslug": "qwen/qwen3-8b-04-28:free", + "providerName": "Chutes", "providerInfo": { - "name": "Nineteen", - "displayName": "Nineteen", - "slug": "nineteen", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://nineteen.ai/tos", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true } }, "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Nineteen", + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, @@ -50057,18 +52818,18 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nineteen.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "Nineteen", - "providerModelId": "casperhansen/deepseek-r1-distill-qwen-32b-awq", - "providerGroup": "Nineteen", - "quantization": "bf16", + "providerDisplayName": "Chutes", + "providerModelId": "Qwen/Qwen3-8B", + "providerGroup": "Chutes", + "quantization": null, "variant": "free", "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 16000, + "maxCompletionTokens": 40960, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -50076,16 +52837,28 @@ export default { "temperature", "top_p", "reasoning", - "include_reasoning" + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://nineteen.ai/tos", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true }, - "training": false + "training": true, + "retainsPrompts": true }, "pricing": { "prompt": "0", @@ -50108,63 +52881,32 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 114, - "newest": 133, - "throughputHighToLow": 49, - "latencyLowToHigh": 74, - "pricingLowToHigh": 50, - "pricingHighToLow": 192 + "topWeekly": 99, + "newest": 24, + "throughputHighToLow": 27, + "latencyLowToHigh": 132, + "pricingLowToHigh": 10, + "pricingHighToLow": 229 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ - { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": 16384, - "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", - "pricing": { - "prompt": "0.00000012", - "completion": "0.00000018", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "inputCost": 0.12, - "outputCost": 0.18 - }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", - "quantization": null, - "context": 64000, - "maxCompletionTokens": 64000, - "providerModelId": "deepseek/deepseek-r1-distill-qwen-32b", + "quantization": "fp8", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "qwen/qwen3-8b-fp8", "pricing": { - "prompt": "0.0000003", - "completion": "0.0000003", + "prompt": "0.000000035", + "completion": "0.000000138", "image": "0", "request": "0", "webSearch": "0", @@ -50186,128 +52928,99 @@ export default { "repetition_penalty", "logit_bias" ], - "inputCost": 0.3, - "outputCost": 0.3 - }, - { - "name": "Cloudflare", - "slug": "cloudflare", - "quantization": null, - "context": 80000, - "maxCompletionTokens": null, - "providerModelId": "@cf/deepseek-ai/deepseek-r1-distill-qwen-32b", - "pricing": { - "prompt": "0.0000005", - "completion": "0.00000488", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "top_k", - "seed", - "repetition_penalty", - "frequency_penalty", - "presence_penalty" - ], - "inputCost": 0.5, - "outputCost": 4.88 + "inputCost": 0.04, + "outputCost": 0.14, + "throughput": 89.547, + "latency": 751 } ] }, { - "slug": "qwen/qwen2.5-vl-32b-instruct", - "hfSlug": "Qwen/Qwen2.5-VL-32B-Instruct", + "slug": "amazon/nova-micro-v1", + "hfSlug": "", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-24T18:10:38.542849+00:00", + "createdAt": "2024-12-05T22:20:37.90344+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen2.5 VL 32B Instruct (free)", - "shortName": "Qwen2.5 VL 32B Instruct (free)", - "author": "qwen", - "description": "Qwen2.5-VL-32B is a multimodal vision-language model fine-tuned through reinforcement learning for enhanced mathematical reasoning, structured outputs, and visual problem-solving capabilities. It excels at visual analysis tasks, including object recognition, textual interpretation within images, and precise event localization in extended videos. Qwen2.5-VL-32B demonstrates state-of-the-art performance across multimodal benchmarks such as MMMU, MathVista, and VideoMME, while maintaining strong reasoning and clarity in text-based tasks like MMLU, mathematical problem-solving, and code generation.", + "name": "Amazon: Nova Micro 1.0", + "shortName": "Nova Micro 1.0", + "author": "amazon", + "description": "Amazon Nova Micro 1.0 is a text-only model that delivers the lowest latency responses in the Amazon Nova family of models at a very low cost. With a context length of 128K tokens and optimized for speed and cost, Amazon Nova Micro excels at tasks such as text summarization, translation, content classification, interactive chat, and brainstorming. It has simple mathematical reasoning and coding abilities.", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", + "group": "Nova", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen2.5-vl-32b-instruct", + "permaslug": "amazon/nova-micro-v1", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "dc909078-3013-470e-9410-258908dfab6a", - "name": "Alibaba | qwen/qwen2.5-vl-32b-instruct:free", - "contextLength": 8192, + "id": "474f0074-66f9-42f0-a866-81a2ffebb001", + "name": "Amazon Bedrock | amazon/nova-micro-v1", + "contextLength": 128000, "model": { - "slug": "qwen/qwen2.5-vl-32b-instruct", - "hfSlug": "Qwen/Qwen2.5-VL-32B-Instruct", + "slug": "amazon/nova-micro-v1", + "hfSlug": "", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-24T18:10:38.542849+00:00", + "createdAt": "2024-12-05T22:20:37.90344+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen2.5 VL 32B Instruct", - "shortName": "Qwen2.5 VL 32B Instruct", - "author": "qwen", - "description": "Qwen2.5-VL-32B is a multimodal vision-language model fine-tuned through reinforcement learning for enhanced mathematical reasoning, structured outputs, and visual problem-solving capabilities. It excels at visual analysis tasks, including object recognition, textual interpretation within images, and precise event localization in extended videos. Qwen2.5-VL-32B demonstrates state-of-the-art performance across multimodal benchmarks such as MMMU, MathVista, and VideoMME, while maintaining strong reasoning and clarity in text-based tasks like MMLU, mathematical problem-solving, and code generation.", + "name": "Amazon: Nova Micro 1.0", + "shortName": "Nova Micro 1.0", + "author": "amazon", + "description": "Amazon Nova Micro 1.0 is a text-only model that delivers the lowest latency responses in the Amazon Nova family of models at a very low cost. With a context length of 128K tokens and optimized for speed and cost, Amazon Nova Micro excels at tasks such as text summarization, translation, content classification, interactive chat, and brainstorming. It has simple mathematical reasoning and coding abilities.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", + "group": "Nova", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen2.5-vl-32b-instruct", + "permaslug": "amazon/nova-micro-v1", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "qwen/qwen2.5-vl-32b-instruct:free", - "modelVariantPermaslug": "qwen/qwen2.5-vl-32b-instruct:free", - "providerName": "Alibaba", + "modelVariantSlug": "amazon/nova-micro-v1", + "modelVariantPermaslug": "amazon/nova-micro-v1", + "providerName": "Amazon Bedrock", "providerInfo": { - "name": "Alibaba", - "displayName": "Alibaba", - "slug": "alibaba", + "name": "Amazon Bedrock", + "displayName": "Amazon Bedrock", + "slug": "amazon-bedrock", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", - "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", + "termsOfServiceUrl": "https://aws.amazon.com/service-terms/", + "privacyPolicyUrl": "https://aws.amazon.com/privacy", + "dataPolicyUrl": "https://docs.aws.amazon.com/bedrock/latest/userguide/data-protection.html", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false } }, - "headquarters": "CN", + "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, + "hasCompletions": false, "isAbortable": false, - "moderationRequired": false, - "group": "Alibaba", + "moderationRequired": true, + "group": "Amazon Bedrock", "editors": [], "owners": [], "isMultipartSupported": true, @@ -50315,41 +53028,44 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256" + "url": "/images/icons/Bedrock.svg" } }, - "providerDisplayName": "Alibaba", - "providerModelId": "qwen2.5-vl-32b-instruct", - "providerGroup": "Alibaba", - "quantization": "bf16", - "variant": "free", - "isFree": true, + "providerDisplayName": "Amazon Bedrock", + "providerModelId": "us.amazon.nova-micro-v1:0", + "providerGroup": "Amazon Bedrock", + "quantization": null, + "variant": "standard", + "isFree": false, "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 5120, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", "max_tokens", "temperature", "top_p", - "seed", - "response_format", - "presence_penalty" + "top_k", + "stop" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", - "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", + "termsOfServiceUrl": "https://aws.amazon.com/service-terms/", + "privacyPolicyUrl": "https://aws.amazon.com/privacy", + "dataPolicyUrl": "https://docs.aws.amazon.com/bedrock/latest/userguide/data-protection.html", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false }, - "training": false + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.000000035", + "completion": "0.00000014", "image": "0", "request": "0", "webSearch": "0", @@ -50360,12 +53076,12 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -50373,24 +53089,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 119, - "newest": 73, - "throughputHighToLow": 131, - "latencyLowToHigh": 118, - "pricingLowToHigh": 32, - "pricingHighToLow": 102 + "topWeekly": 175, + "newest": 160, + "throughputHighToLow": 6, + "latencyLowToHigh": 50, + "pricingLowToHigh": 90, + "pricingHighToLow": 228 }, + "authorIcon": "https://openrouter.ai/images/icons/Bedrock.svg", "providers": [ { - "name": "Fireworks", - "slug": "fireworks", + "name": "Amazon Bedrock", + "icon": "https://openrouter.ai/images/icons/Bedrock.svg", + "slug": "amazonBedrock", "quantization": null, "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "accounts/fireworks/models/qwen2p5-vl-32b-instruct", + "maxCompletionTokens": 5120, + "providerModelId": "us.amazon.nova-micro-v1:0", "pricing": { - "prompt": "0.0000009", - "completion": "0.0000009", + "prompt": "0.000000035", + "completion": "0.00000014", "image": "0", "request": "0", "webSearch": "0", @@ -50398,22 +53116,17 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", "top_k", - "repetition_penalty", - "response_format", - "structured_outputs", - "logit_bias", - "logprobs", - "top_logprobs" + "stop" ], - "inputCost": 0.9, - "outputCost": 0.9 + "inputCost": 0.04, + "outputCost": 0.14, + "throughput": 257.281, + "latency": 414 } ] }, @@ -50571,55 +53284,323 @@ export default { "features": { "supportsDocumentUrl": null }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 174, - "newest": 209, - "throughputHighToLow": 317, - "latencyLowToHigh": 296, - "pricingLowToHigh": 234, - "pricingHighToLow": 88 - }, - "providers": [ + "providerRegion": null + }, + "sorting": { + "topWeekly": 176, + "newest": 209, + "throughputHighToLow": 317, + "latencyLowToHigh": 289, + "pricingLowToHigh": 234, + "pricingHighToLow": 88 + }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", + "providers": [ + { + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": "unknown", + "context": 128000, + "maxCompletionTokens": 65536, + "providerModelId": "o1-mini", + "pricing": { + "prompt": "0.0000011", + "completion": "0.0000044", + "image": "0", + "request": "0", + "inputCacheRead": "0.00000055", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "seed", + "max_tokens" + ], + "inputCost": 1.1, + "outputCost": 4.4, + "throughput": 148.163, + "latency": 4658 + } + ] + }, + { + "slug": "qwen/qwq-32b-preview", + "hfSlug": "Qwen/QwQ-32B-Preview", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2024-11-28T00:42:21.381013+00:00", + "hfUpdatedAt": null, + "name": "Qwen: QwQ 32B Preview", + "shortName": "QwQ 32B Preview", + "author": "qwen", + "description": "QwQ-32B-Preview is an experimental research model focused on AI reasoning capabilities developed by the Qwen Team. As a preview release, it demonstrates promising analytical abilities while having several important limitations:\n\n1. **Language Mixing and Code-Switching**: The model may mix languages or switch between them unexpectedly, affecting response clarity.\n2. **Recursive Reasoning Loops**: The model may enter circular reasoning patterns, leading to lengthy responses without a conclusive answer.\n3. **Safety and Ethical Considerations**: The model requires enhanced safety measures to ensure reliable and secure performance, and users should exercise caution when deploying it.\n4. **Performance and Benchmark Limitations**: The model excels in math and coding but has room for improvement in other areas, such as common sense reasoning and nuanced language understanding.\n\n", + "modelVersionGroupId": null, + "contextLength": 32768, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Qwen", + "instructType": "deepseek-r1", + "defaultSystem": null, + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "qwen/qwq-32b-preview", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, + "endpoint": { + "id": "f7d63c65-9ad5-4f28-bb53-5fc9242efaea", + "name": "Nebius | qwen/qwq-32b-preview", + "contextLength": 32768, + "model": { + "slug": "qwen/qwq-32b-preview", + "hfSlug": "Qwen/QwQ-32B-Preview", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2024-11-28T00:42:21.381013+00:00", + "hfUpdatedAt": null, + "name": "Qwen: QwQ 32B Preview", + "shortName": "QwQ 32B Preview", + "author": "qwen", + "description": "QwQ-32B-Preview is an experimental research model focused on AI reasoning capabilities developed by the Qwen Team. As a preview release, it demonstrates promising analytical abilities while having several important limitations:\n\n1. **Language Mixing and Code-Switching**: The model may mix languages or switch between them unexpectedly, affecting response clarity.\n2. **Recursive Reasoning Loops**: The model may enter circular reasoning patterns, leading to lengthy responses without a conclusive answer.\n3. **Safety and Ethical Considerations**: The model requires enhanced safety measures to ensure reliable and secure performance, and users should exercise caution when deploying it.\n4. **Performance and Benchmark Limitations**: The model excels in math and coding but has room for improvement in other areas, such as common sense reasoning and nuanced language understanding.\n\n", + "modelVersionGroupId": null, + "contextLength": 32768, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Qwen", + "instructType": "deepseek-r1", + "defaultSystem": null, + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "qwen/qwq-32b-preview", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } + }, + "modelVariantSlug": "qwen/qwq-32b-preview", + "modelVariantPermaslug": "qwen/qwq-32b-preview", + "providerName": "Nebius", + "providerInfo": { + "name": "Nebius", + "displayName": "Nebius AI Studio", + "slug": "nebius", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "paidModels": { + "training": false, + "retainsPrompts": false + } + }, + "headquarters": "DE", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": false, + "moderationRequired": false, + "group": "Nebius", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "invertRequired": true + } + }, + "providerDisplayName": "Nebius AI Studio", + "providerModelId": "Qwen/QwQ-32B-Preview", + "providerGroup": "Nebius", + "quantization": "fp8", + "variant": "standard", + "isFree": false, + "canAbort": false, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "paidModels": { + "training": false, + "retainsPrompts": false + }, + "training": false, + "retainsPrompts": false + }, + "pricing": { + "prompt": "0.00000009", + "completion": "0.00000027", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 177, + "newest": 162, + "throughputHighToLow": 255, + "latencyLowToHigh": 284, + "pricingLowToHigh": 57, + "pricingHighToLow": 200 + }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", + "providers": [ + { + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "Qwen/QwQ-32B-Preview", + "pricing": { + "prompt": "0.00000009", + "completion": "0.00000027", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.09, + "outputCost": 0.27, + "throughput": 29.426, + "latency": 1264.5 + }, { - "name": "OpenAI", - "slug": "openAi", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 65536, - "providerModelId": "o1-mini", + "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", + "slug": "hyperbolic", + "quantization": "fp8", + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "Qwen/QwQ-32B-Preview", "pricing": { - "prompt": "0.0000011", - "completion": "0.0000044", + "prompt": "0.0000002", + "completion": "0.0000002", "image": "0", "request": "0", - "inputCacheRead": "0.00000055", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "logprobs", + "top_logprobs", "seed", - "max_tokens" + "logit_bias", + "top_k", + "min_p", + "repetition_penalty" ], - "inputCost": 1.1, - "outputCost": 4.4 + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 24.9545, + "latency": 21851.5 } ] }, { - "slug": "x-ai/grok-beta", + "slug": "perplexity/llama-3.1-sonar-large-128k-online", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-10-20T00:00:00+00:00", + "createdAt": "2024-08-01T00:00:00+00:00", "hfUpdatedAt": null, - "name": "xAI: Grok Beta", - "shortName": "Grok Beta", - "author": "x-ai", - "description": "Grok Beta is xAI's experimental language model with state-of-the-art reasoning capabilities, best for complex and multi-step use cases.\n\nIt is the successor of [Grok 2](https://x.ai/blog/grok-2) with enhanced context length.", - "modelVersionGroupId": null, - "contextLength": 131072, + "name": "Perplexity: Llama 3.1 Sonar 70B Online", + "shortName": "Llama 3.1 Sonar 70B Online", + "author": "perplexity", + "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3.1-sonar-large-128k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", + "modelVersionGroupId": "e0349fb1-b84a-4f4f-9023-e7f1c1e73b33", + "contextLength": 127072, "inputModalities": [ "text" ], @@ -50627,32 +53608,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Grok", + "group": "Llama3", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "x-ai/grok-beta", + "permaslug": "perplexity/llama-3.1-sonar-large-128k-online", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "7f10aaf2-5317-4238-b4be-212fe355d24c", - "name": "xAI | x-ai/grok-beta", - "contextLength": 131072, + "id": "aaad39f2-2e0d-46a6-8078-928b6eebe5d3", + "name": "Perplexity | perplexity/llama-3.1-sonar-large-128k-online", + "contextLength": 127072, "model": { - "slug": "x-ai/grok-beta", + "slug": "perplexity/llama-3.1-sonar-large-128k-online", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-10-20T00:00:00+00:00", + "createdAt": "2024-08-01T00:00:00+00:00", "hfUpdatedAt": null, - "name": "xAI: Grok Beta", - "shortName": "Grok Beta", - "author": "x-ai", - "description": "Grok Beta is xAI's experimental language model with state-of-the-art reasoning capabilities, best for complex and multi-step use cases.\n\nIt is the successor of [Grok 2](https://x.ai/blog/grok-2) with enhanced context length.", - "modelVersionGroupId": null, - "contextLength": 131072, + "name": "Perplexity: Llama 3.1 Sonar 70B Online", + "shortName": "Llama 3.1 Sonar 70B Online", + "author": "perplexity", + "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3.1-sonar-large-128k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", + "modelVersionGroupId": "e0349fb1-b84a-4f4f-9023-e7f1c1e73b33", + "contextLength": 127072, "inputModalities": [ "text" ], @@ -50660,94 +53641,82 @@ export default { "text" ], "hasTextOutput": true, - "group": "Grok", + "group": "Llama3", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "x-ai/grok-beta", + "permaslug": "perplexity/llama-3.1-sonar-large-128k-online", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "x-ai/grok-beta", - "modelVariantPermaslug": "x-ai/grok-beta", - "providerName": "xAI", + "modelVariantSlug": "perplexity/llama-3.1-sonar-large-128k-online", + "modelVariantPermaslug": "perplexity/llama-3.1-sonar-large-128k-online", + "providerName": "Perplexity", "providerInfo": { - "name": "xAI", - "displayName": "xAI", - "slug": "xai", + "name": "Perplexity", + "displayName": "Perplexity", + "slug": "perplexity", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://x.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://x.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", + "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, + "hasCompletions": false, + "isAbortable": false, "moderationRequired": false, - "group": "xAI", + "group": "Perplexity", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.x.ai/", + "statusPageUrl": "https://status.perplexity.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://x.ai/&size=256" + "url": "/images/icons/Perplexity.svg" } }, - "providerDisplayName": "xAI", - "providerModelId": "grok-beta", - "providerGroup": "xAI", + "providerDisplayName": "Perplexity", + "providerModelId": "llama-3.1-sonar-large-128k-online", + "providerGroup": "Perplexity", "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "stop", + "top_k", "frequency_penalty", - "presence_penalty", - "seed", - "logprobs", - "top_logprobs", - "response_format" + "presence_penalty" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://x.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://x.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", + "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, "pricing": { - "prompt": "0.000005", - "completion": "0.000015", + "prompt": "0.000001", + "completion": "0.000001", "image": "0", - "request": "0", + "request": "0.005", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -50756,12 +53725,12 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -50769,61 +53738,59 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 175, - "newest": 185, - "throughputHighToLow": 139, - "latencyLowToHigh": 25, - "pricingLowToHigh": 291, - "pricingHighToLow": 27 + "topWeekly": 178, + "newest": 228, + "throughputHighToLow": 97, + "latencyLowToHigh": 218, + "pricingLowToHigh": 288, + "pricingHighToLow": 31 }, + "authorIcon": "https://openrouter.ai/images/icons/Perplexity.svg", "providers": [ { - "name": "xAI", - "slug": "xAi", + "name": "Perplexity", + "icon": "https://openrouter.ai/images/icons/Perplexity.svg", + "slug": "perplexity", "quantization": "unknown", - "context": 131072, + "context": 127072, "maxCompletionTokens": null, - "providerModelId": "grok-beta", + "providerModelId": "llama-3.1-sonar-large-128k-online", "pricing": { - "prompt": "0.000005", - "completion": "0.000015", + "prompt": "0.000001", + "completion": "0.000001", "image": "0", - "request": "0", + "request": "0.005", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "stop", + "top_k", "frequency_penalty", - "presence_penalty", - "seed", - "logprobs", - "top_logprobs", - "response_format" + "presence_penalty" ], - "inputCost": 5, - "outputCost": 15 + "inputCost": 1, + "outputCost": 1, + "throughput": 77.599, + "latency": 1730 } ] }, { - "slug": "deepseek/deepseek-r1-distill-qwen-14b", - "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B", + "slug": "deepseek/deepseek-r1-distill-qwen-32b", + "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-01-29T23:39:00.13687+00:00", + "createdAt": "2025-01-29T23:53:50.865297+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: R1 Distill Qwen 14B", - "shortName": "R1 Distill Qwen 14B", + "name": "DeepSeek: R1 Distill Qwen 32B (free)", + "shortName": "R1 Distill Qwen 32B (free)", "author": "deepseek", - "description": "DeepSeek R1 Distill Qwen 14B is a distilled large language model based on [Qwen 2.5 14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 69.7\n- MATH-500 pass@1: 93.9\n- CodeForces Rating: 1481\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", + "description": "DeepSeek R1 Distill Qwen 32B is a distilled large language model based on [Qwen 2.5 32B](https://huggingface.co/Qwen/Qwen2.5-32B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 72.6\n- MATH-500 pass@1: 94.3\n- CodeForces Rating: 1691\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", - "contextLength": 64000, + "contextLength": 16000, "inputModalities": [ "text" ], @@ -50840,8 +53807,8 @@ export default { ], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "deepseek/deepseek-r1-distill-qwen-14b", + "warningMessage": null, + "permaslug": "deepseek/deepseek-r1-distill-qwen-32b", "reasoningConfig": { "startToken": "", "endToken": "" @@ -50853,21 +53820,21 @@ export default { } }, "endpoint": { - "id": "a18c0e49-6efb-439a-9287-fb2f398f3c5a", - "name": "Novita | deepseek/deepseek-r1-distill-qwen-14b", - "contextLength": 64000, + "id": "749d8a55-ab77-43a8-aa1a-4d0331c2e738", + "name": "Nineteen | deepseek/deepseek-r1-distill-qwen-32b:free", + "contextLength": 16000, "model": { - "slug": "deepseek/deepseek-r1-distill-qwen-14b", - "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B", + "slug": "deepseek/deepseek-r1-distill-qwen-32b", + "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-01-29T23:39:00.13687+00:00", + "createdAt": "2025-01-29T23:53:50.865297+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: R1 Distill Qwen 14B", - "shortName": "R1 Distill Qwen 14B", + "name": "DeepSeek: R1 Distill Qwen 32B", + "shortName": "R1 Distill Qwen 32B", "author": "deepseek", - "description": "DeepSeek R1 Distill Qwen 14B is a distilled large language model based on [Qwen 2.5 14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 69.7\n- MATH-500 pass@1: 93.9\n- CodeForces Rating: 1481\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", + "description": "DeepSeek R1 Distill Qwen 32B is a distilled large language model based on [Qwen 2.5 32B](https://huggingface.co/Qwen/Qwen2.5-32B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 72.6\n- MATH-500 pass@1: 94.3\n- CodeForces Rating: 1691\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", - "contextLength": 131072, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -50884,8 +53851,8 @@ export default { ], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "deepseek/deepseek-r1-distill-qwen-14b", + "warningMessage": null, + "permaslug": "deepseek/deepseek-r1-distill-qwen-32b", "reasoningConfig": { "startToken": "", "endToken": "" @@ -50897,47 +53864,44 @@ export default { } } }, - "modelVariantSlug": "deepseek/deepseek-r1-distill-qwen-14b", - "modelVariantPermaslug": "deepseek/deepseek-r1-distill-qwen-14b", - "providerName": "Novita", + "modelVariantSlug": "deepseek/deepseek-r1-distill-qwen-32b:free", + "modelVariantPermaslug": "deepseek/deepseek-r1-distill-qwen-32b:free", + "providerName": "Nineteen", "providerInfo": { - "name": "Novita", - "displayName": "NovitaAI", - "slug": "novita", + "name": "Nineteen", + "displayName": "Nineteen", + "slug": "nineteen", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://nineteen.ai/tos", "paidModels": { "training": false } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Novita", + "group": "Nineteen", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.novita.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nineteen.ai/&size=256" } }, - "providerDisplayName": "NovitaAI", - "providerModelId": "deepseek/deepseek-r1-distill-qwen-14b", - "providerGroup": "Novita", - "quantization": null, - "variant": "standard", - "isFree": false, + "providerDisplayName": "Nineteen", + "providerModelId": "casperhansen/deepseek-r1-distill-qwen-32b-awq", + "providerGroup": "Nineteen", + "quantization": "bf16", + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 64000, + "maxCompletionTokens": 16000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -50945,29 +53909,20 @@ export default { "temperature", "top_p", "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" + "include_reasoning" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://nineteen.ai/tos", "paidModels": { "training": false }, "training": false }, "pricing": { - "prompt": "0.00000015", - "completion": "0.00000015", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -50991,24 +53946,63 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 176, - "newest": 135, - "throughputHighToLow": 125, - "latencyLowToHigh": 229, - "pricingLowToHigh": 51, - "pricingHighToLow": 184 + "topWeekly": 117, + "newest": 133, + "throughputHighToLow": 53, + "latencyLowToHigh": 118, + "pricingLowToHigh": 50, + "pricingHighToLow": 192 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [ + { + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": 16384, + "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", + "pricing": { + "prompt": "0.00000012", + "completion": "0.00000018", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" + ], + "inputCost": 0.12, + "outputCost": 0.18, + "throughput": 45.5815, + "latency": 668 + }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": null, "context": 64000, "maxCompletionTokens": 64000, - "providerModelId": "deepseek/deepseek-r1-distill-qwen-14b", + "providerModelId": "deepseek/deepseek-r1-distill-qwen-32b", "pricing": { - "prompt": "0.00000015", - "completion": "0.00000015", + "prompt": "0.0000003", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -51030,19 +54024,22 @@ export default { "repetition_penalty", "logit_bias" ], - "inputCost": 0.15, - "outputCost": 0.15 + "inputCost": 0.3, + "outputCost": 0.3, + "throughput": 19.1495, + "latency": 2299.5 }, { - "name": "Together", - "slug": "together", + "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", + "slug": "cloudflare", "quantization": null, - "context": 131072, - "maxCompletionTokens": 32768, - "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B", + "context": 80000, + "maxCompletionTokens": null, + "providerModelId": "@cf/deepseek-ai/deepseek-r1-distill-qwen-32b", "pricing": { - "prompt": "0.0000016", - "completion": "0.0000016", + "prompt": "0.0000005", + "completion": "0.00000488", "image": "0", "request": "0", "webSearch": "0", @@ -51055,17 +54052,16 @@ export default { "top_p", "reasoning", "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", "top_k", + "seed", "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "frequency_penalty", + "presence_penalty" ], - "inputCost": 1.6, - "outputCost": 1.6 + "inputCost": 0.5, + "outputCost": 4.88, + "throughput": 32.7715, + "latency": 867 } ] }, @@ -51226,27 +54222,28 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 177, + "topWeekly": 180, "newest": 46, - "throughputHighToLow": 254, - "latencyLowToHigh": 221, + "throughputHighToLow": 252, + "latencyLowToHigh": 245, "pricingLowToHigh": 20, "pricingHighToLow": 268 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { - "slug": "perplexity/sonar-pro", - "hfSlug": "", - "updatedAt": "2025-05-05T14:37:56.196383+00:00", - "createdAt": "2025-03-07T01:53:43+00:00", + "slug": "qwen/qwen2.5-vl-32b-instruct", + "hfSlug": "Qwen/Qwen2.5-VL-32B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-03-24T18:10:38.542849+00:00", "hfUpdatedAt": null, - "name": "Perplexity: Sonar Pro", - "shortName": "Sonar Pro", - "author": "perplexity", - "description": "Note: Sonar Pro pricing includes Perplexity search pricing. See [details here](https://docs.perplexity.ai/guides/pricing#detailed-pricing-breakdown-for-sonar-reasoning-pro-and-sonar-pro)\n\nFor enterprises seeking more advanced capabilities, the Sonar Pro API can handle in-depth, multi-step queries with added extensibility, like double the number of citations per search as Sonar on average. Plus, with a larger context window, it can handle longer and more nuanced searches and follow-up questions. ", + "name": "Qwen: Qwen2.5 VL 32B Instruct (free)", + "shortName": "Qwen2.5 VL 32B Instruct (free)", + "author": "qwen", + "description": "Qwen2.5-VL-32B is a multimodal vision-language model fine-tuned through reinforcement learning for enhanced mathematical reasoning, structured outputs, and visual problem-solving capabilities. It excels at visual analysis tasks, including object recognition, textual interpretation within images, and precise event localization in extended videos. Qwen2.5-VL-32B demonstrates state-of-the-art performance across multimodal benchmarks such as MMMU, MathVista, and VideoMME, while maintaining strong reasoning and clarity in text-based tasks like MMLU, mathematical problem-solving, and code generation.", "modelVersionGroupId": null, - "contextLength": 200000, + "contextLength": 8192, "inputModalities": [ "text", "image" @@ -51255,32 +54252,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "Qwen", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "perplexity/sonar-pro", + "warningMessage": null, + "permaslug": "qwen/qwen2.5-vl-32b-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "e19b1036-fab9-4f97-9579-8ea67959cc9b", - "name": "Perplexity | perplexity/sonar-pro", - "contextLength": 200000, + "id": "dc909078-3013-470e-9410-258908dfab6a", + "name": "Alibaba | qwen/qwen2.5-vl-32b-instruct:free", + "contextLength": 8192, "model": { - "slug": "perplexity/sonar-pro", - "hfSlug": "", - "updatedAt": "2025-05-05T14:37:56.196383+00:00", - "createdAt": "2025-03-07T01:53:43+00:00", + "slug": "qwen/qwen2.5-vl-32b-instruct", + "hfSlug": "Qwen/Qwen2.5-VL-32B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-03-24T18:10:38.542849+00:00", "hfUpdatedAt": null, - "name": "Perplexity: Sonar Pro", - "shortName": "Sonar Pro", - "author": "perplexity", - "description": "Note: Sonar Pro pricing includes Perplexity search pricing. See [details here](https://docs.perplexity.ai/guides/pricing#detailed-pricing-breakdown-for-sonar-reasoning-pro-and-sonar-pro)\n\nFor enterprises seeking more advanced capabilities, the Sonar Pro API can handle in-depth, multi-step queries with added extensibility, like double the number of citations per search as Sonar on average. Plus, with a larger context window, it can handle longer and more nuanced searches and follow-up questions. ", + "name": "Qwen: Qwen2.5 VL 32B Instruct", + "shortName": "Qwen2.5 VL 32B Instruct", + "author": "qwen", + "description": "Qwen2.5-VL-32B is a multimodal vision-language model fine-tuned through reinforcement learning for enhanced mathematical reasoning, structured outputs, and visual problem-solving capabilities. It excels at visual analysis tasks, including object recognition, textual interpretation within images, and precise event localization in extended videos. Qwen2.5-VL-32B demonstrates state-of-the-art performance across multimodal benchmarks such as MMMU, MathVista, and VideoMME, while maintaining strong reasoning and clarity in text-based tasks like MMLU, mathematical problem-solving, and code generation.", "modelVersionGroupId": null, - "contextLength": 200000, + "contextLength": 32768, "inputModalities": [ "text", "image" @@ -51289,104 +54286,87 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "Qwen", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "perplexity/sonar-pro", + "warningMessage": null, + "permaslug": "qwen/qwen2.5-vl-32b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "perplexity/sonar-pro", - "modelVariantPermaslug": "perplexity/sonar-pro", - "providerName": "Perplexity", + "modelVariantSlug": "qwen/qwen2.5-vl-32b-instruct:free", + "modelVariantPermaslug": "qwen/qwen2.5-vl-32b-instruct:free", + "providerName": "Alibaba", "providerInfo": { - "name": "Perplexity", - "displayName": "Perplexity", - "slug": "perplexity", + "name": "Alibaba", + "displayName": "Alibaba", + "slug": "alibaba", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", - "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", + "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", + "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", "paidModels": { "training": false } }, - "headquarters": "US", + "headquarters": "CN", "hasChatCompletions": true, - "hasCompletions": false, + "hasCompletions": true, "isAbortable": false, "moderationRequired": false, - "group": "Perplexity", + "group": "Alibaba", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.perplexity.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Perplexity.svg" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256" } }, - "providerDisplayName": "Perplexity", - "providerModelId": "sonar-pro", - "providerGroup": "Perplexity", - "quantization": null, - "variant": "standard", - "isFree": false, + "providerDisplayName": "Alibaba", + "providerModelId": "qwen2.5-vl-32b-instruct", + "providerGroup": "Alibaba", + "quantization": "bf16", + "variant": "free", + "isFree": true, "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 8000, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "web_search_options", - "top_k", - "frequency_penalty", + "seed", + "response_format", "presence_penalty" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", - "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", + "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", + "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", "paidModels": { "training": false }, "training": false }, "pricing": { - "prompt": "0.000003", - "completion": "0.000015", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", - "webSearch": "0.005", + "webSearch": "0", "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [ - { - "type": "search-threshold", - "threshold": "high", - "request": "0.014" - }, - { - "type": "search-threshold", - "threshold": "medium", - "request": "0.01" - }, - { - "type": "search-threshold", - "threshold": "low", - "request": "0.006" - } - ], + "discount": 0 + }, + "variablePricings": [], "isHidden": false, "isDeranked": false, "isDisabled": false, @@ -51395,7 +54375,7 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -51403,27 +54383,29 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 178, - "newest": 99, - "throughputHighToLow": 146, - "latencyLowToHigh": 266, - "pricingLowToHigh": 268, - "pricingHighToLow": 40 + "topWeekly": 114, + "newest": 73, + "throughputHighToLow": 129, + "latencyLowToHigh": 161, + "pricingLowToHigh": 32, + "pricingHighToLow": 103 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { - "name": "Perplexity", - "slug": "perplexity", + "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", + "slug": "fireworks", "quantization": null, - "context": 200000, - "maxCompletionTokens": 8000, - "providerModelId": "sonar-pro", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "accounts/fireworks/models/qwen2p5-vl-32b-instruct", "pricing": { - "prompt": "0.000003", - "completion": "0.000015", + "prompt": "0.0000009", + "completion": "0.0000009", "image": "0", "request": "0", - "webSearch": "0.005", + "webSearch": "0", "internalReasoning": "0", "discount": 0 }, @@ -51431,159 +54413,191 @@ export default { "max_tokens", "temperature", "top_p", - "web_search_options", + "stop", + "frequency_penalty", + "presence_penalty", "top_k", + "repetition_penalty", + "response_format", + "structured_outputs", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.9, + "outputCost": 0.9, + "throughput": 61.0595, + "latency": 937 + }, + { + "name": "InoCloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inocloud.com/&size=256", + "slug": "inoCloud", + "quantization": "bf16", + "context": 128000, + "maxCompletionTokens": 128000, + "providerModelId": "qwen/qwen2.5-vl-32b-instruct", + "pricing": { + "prompt": "0.0000011", + "completion": "0.0000011", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", "frequency_penalty", "presence_penalty" ], - "inputCost": 3, - "outputCost": 15 + "inputCost": 1.1, + "outputCost": 1.1, + "throughput": 26.49, + "latency": 3097 } ] }, { - "slug": "google/gemma-3-4b-it", - "hfSlug": "google/gemma-3-4b-it", + "slug": "cohere/command-r-plus-08-2024", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-13T22:38:30.653142+00:00", + "createdAt": "2024-08-30T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Google: Gemma 3 4B (free)", - "shortName": "Gemma 3 4B (free)", - "author": "google", - "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling.", - "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", - "contextLength": 131072, + "name": "Cohere: Command R+ (08-2024)", + "shortName": "Command R+ (08-2024)", + "author": "cohere", + "description": "command-r-plus-08-2024 is an update of the [Command R+](/models/cohere/command-r-plus) with roughly 50% higher throughput and 25% lower latencies as compared to the previous Command R+ version, while keeping the hardware footprint the same.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", + "modelVersionGroupId": null, + "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Gemini", - "instructType": "gemma", + "group": "Cohere", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "", - "", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "google/gemma-3-4b-it", + "permaslug": "cohere/command-r-plus-08-2024", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "1bbec5f0-58e0-4af2-b8ca-99c202b4ceca", - "name": "Chutes | google/gemma-3-4b-it:free", - "contextLength": 131072, + "id": "cd63714a-d459-4806-bdf2-0dfea4f6614c", + "name": "Cohere | cohere/command-r-plus-08-2024", + "contextLength": 128000, "model": { - "slug": "google/gemma-3-4b-it", - "hfSlug": "google/gemma-3-4b-it", + "slug": "cohere/command-r-plus-08-2024", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-13T22:38:30.653142+00:00", + "createdAt": "2024-08-30T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Google: Gemma 3 4B", - "shortName": "Gemma 3 4B", - "author": "google", - "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling.", - "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", - "contextLength": 131072, + "name": "Cohere: Command R+ (08-2024)", + "shortName": "Command R+ (08-2024)", + "author": "cohere", + "description": "command-r-plus-08-2024 is an update of the [Command R+](/models/cohere/command-r-plus) with roughly 50% higher throughput and 25% lower latencies as compared to the previous Command R+ version, while keeping the hardware footprint the same.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", + "modelVersionGroupId": null, + "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Gemini", - "instructType": "gemma", + "group": "Cohere", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "", - "", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "google/gemma-3-4b-it", + "permaslug": "cohere/command-r-plus-08-2024", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "google/gemma-3-4b-it:free", - "modelVariantPermaslug": "google/gemma-3-4b-it:free", - "providerName": "Chutes", + "modelVariantSlug": "cohere/command-r-plus-08-2024", + "modelVariantPermaslug": "cohere/command-r-plus-08-2024", + "providerName": "Cohere", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "Cohere", + "displayName": "Cohere", + "slug": "cohere", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "privacyPolicyUrl": "https://cohere.com/privacy", + "termsOfServiceUrl": "https://cohere.com/terms-of-use", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": true, + "retentionDays": 30 } }, + "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, + "hasCompletions": false, "isAbortable": true, "moderationRequired": false, - "group": "Chutes", + "group": "Cohere", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.cohere.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "/images/icons/Cohere.png" } }, - "providerDisplayName": "Chutes", - "providerModelId": "unsloth/gemma-3-4b-it", - "providerGroup": "Chutes", - "quantization": "bf16", - "variant": "free", - "isFree": true, + "providerDisplayName": "Cohere", + "providerModelId": "command-r-plus-08-2024", + "providerGroup": "Cohere", + "quantization": "unknown", + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 8192, + "maxCompletionTokens": 4000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "seed", "top_k", - "min_p", - "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" + "seed", + "response_format", + "structured_outputs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "privacyPolicyUrl": "https://cohere.com/privacy", + "termsOfServiceUrl": "https://cohere.com/terms-of-use", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.0000025", + "completion": "0.00001", "image": "0", "request": "0", "webSearch": "0", @@ -51594,37 +54608,40 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 63, - "newest": 83, - "throughputHighToLow": 62, - "latencyLowToHigh": 34, - "pricingLowToHigh": 38, - "pricingHighToLow": 241 + "topWeekly": 182, + "newest": 212, + "throughputHighToLow": 209, + "latencyLowToHigh": 46, + "pricingLowToHigh": 263, + "pricingHighToLow": 57 }, + "authorIcon": "https://openrouter.ai/images/icons/Cohere.png", "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "google/gemma-3-4b-it", + "name": "Cohere", + "icon": "https://openrouter.ai/images/icons/Cohere.png", + "slug": "cohere", + "quantization": "unknown", + "context": 128000, + "maxCompletionTokens": 4000, + "providerModelId": "command-r-plus-08-2024", "pricing": { - "prompt": "0.00000002", - "completion": "0.00000004", + "prompt": "0.0000025", + "completion": "0.00001", "image": "0", "request": "0", "webSearch": "0", @@ -51632,130 +54649,113 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", "top_k", "seed", - "min_p" + "response_format", + "structured_outputs" ], - "inputCost": 0.02, - "outputCost": 0.04 + "inputCost": 2.5, + "outputCost": 10, + "throughput": 40.121, + "latency": 420 } ] }, { - "slug": "qwen/qwq-32b-preview", - "hfSlug": "Qwen/QwQ-32B-Preview", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2024-11-28T00:42:21.381013+00:00", + "slug": "meta-llama/llama-guard-4-12b", + "hfSlug": "meta-llama/Llama-Guard-4-12B", + "updatedAt": "2025-04-30T14:10:44.640461+00:00", + "createdAt": "2025-04-30T01:06:33.531556+00:00", "hfUpdatedAt": null, - "name": "Qwen: QwQ 32B Preview", - "shortName": "QwQ 32B Preview", - "author": "qwen", - "description": "QwQ-32B-Preview is an experimental research model focused on AI reasoning capabilities developed by the Qwen Team. As a preview release, it demonstrates promising analytical abilities while having several important limitations:\n\n1. **Language Mixing and Code-Switching**: The model may mix languages or switch between them unexpectedly, affecting response clarity.\n2. **Recursive Reasoning Loops**: The model may enter circular reasoning patterns, leading to lengthy responses without a conclusive answer.\n3. **Safety and Ethical Considerations**: The model requires enhanced safety measures to ensure reliable and secure performance, and users should exercise caution when deploying it.\n4. **Performance and Benchmark Limitations**: The model excels in math and coding but has room for improvement in other areas, such as common sense reasoning and nuanced language understanding.\n\n", + "name": "Meta: Llama Guard 4 12B", + "shortName": "Llama Guard 4 12B", + "author": "meta-llama", + "description": "Llama Guard 4 is a Llama 4 Scout-derived multimodal pretrained model, fine-tuned for content safety classification. Similar to previous versions, it can be used to classify content in both LLM inputs (prompt classification) and in LLM responses (response classification). It acts as an LLM—generating text in its output that indicates whether a given prompt or response is safe or unsafe, and if unsafe, it also lists the content categories violated.\n\nLlama Guard 4 was aligned to safeguard against the standardized MLCommons hazards taxonomy and designed to support multimodal Llama 4 capabilities. Specifically, it combines features from previous Llama Guard models, providing content moderation for English and multiple supported languages, along with enhanced capabilities to handle mixed text-and-image prompts, including multiple images. Additionally, Llama Guard 4 is integrated into the Llama Moderations API, extending robust safety classification to text and images.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 163840, "inputModalities": [ + "image", "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "deepseek-r1", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwq-32b-preview", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "permaslug": "meta-llama/llama-guard-4-12b", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "f7d63c65-9ad5-4f28-bb53-5fc9242efaea", - "name": "Nebius | qwen/qwq-32b-preview", - "contextLength": 32768, + "id": "850b84c3-42a7-4cec-99c0-b5582d0da66b", + "name": "DeepInfra | meta-llama/llama-guard-4-12b", + "contextLength": 163840, "model": { - "slug": "qwen/qwq-32b-preview", - "hfSlug": "Qwen/QwQ-32B-Preview", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2024-11-28T00:42:21.381013+00:00", + "slug": "meta-llama/llama-guard-4-12b", + "hfSlug": "meta-llama/Llama-Guard-4-12B", + "updatedAt": "2025-04-30T14:10:44.640461+00:00", + "createdAt": "2025-04-30T01:06:33.531556+00:00", "hfUpdatedAt": null, - "name": "Qwen: QwQ 32B Preview", - "shortName": "QwQ 32B Preview", - "author": "qwen", - "description": "QwQ-32B-Preview is an experimental research model focused on AI reasoning capabilities developed by the Qwen Team. As a preview release, it demonstrates promising analytical abilities while having several important limitations:\n\n1. **Language Mixing and Code-Switching**: The model may mix languages or switch between them unexpectedly, affecting response clarity.\n2. **Recursive Reasoning Loops**: The model may enter circular reasoning patterns, leading to lengthy responses without a conclusive answer.\n3. **Safety and Ethical Considerations**: The model requires enhanced safety measures to ensure reliable and secure performance, and users should exercise caution when deploying it.\n4. **Performance and Benchmark Limitations**: The model excels in math and coding but has room for improvement in other areas, such as common sense reasoning and nuanced language understanding.\n\n", + "name": "Meta: Llama Guard 4 12B", + "shortName": "Llama Guard 4 12B", + "author": "meta-llama", + "description": "Llama Guard 4 is a Llama 4 Scout-derived multimodal pretrained model, fine-tuned for content safety classification. Similar to previous versions, it can be used to classify content in both LLM inputs (prompt classification) and in LLM responses (response classification). It acts as an LLM—generating text in its output that indicates whether a given prompt or response is safe or unsafe, and if unsafe, it also lists the content categories violated.\n\nLlama Guard 4 was aligned to safeguard against the standardized MLCommons hazards taxonomy and designed to support multimodal Llama 4 capabilities. Specifically, it combines features from previous Llama Guard models, providing content moderation for English and multiple supported languages, along with enhanced capabilities to handle mixed text-and-image prompts, including multiple images. Additionally, Llama Guard 4 is integrated into the Llama Moderations API, extending robust safety classification to text and images.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 163840, "inputModalities": [ + "image", "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "deepseek-r1", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwq-32b-preview", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "permaslug": "meta-llama/llama-guard-4-12b", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "qwen/qwq-32b-preview", - "modelVariantPermaslug": "qwen/qwq-32b-preview", - "providerName": "Nebius", + "modelVariantSlug": "meta-llama/llama-guard-4-12b", + "modelVariantPermaslug": "meta-llama/llama-guard-4-12b", + "providerName": "DeepInfra", "providerInfo": { - "name": "Nebius", - "displayName": "Nebius AI Studio", - "slug": "nebius", + "name": "DeepInfra", + "displayName": "DeepInfra", + "slug": "deepinfra", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { "training": false, "retainsPrompts": false } }, - "headquarters": "DE", + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, + "isAbortable": true, "moderationRequired": false, - "group": "Nebius", + "group": "DeepInfra", "editors": [], "owners": [], "isMultipartSupported": true, @@ -51763,17 +54763,16 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", - "invertRequired": true + "url": "/images/icons/DeepInfra.webp" } }, - "providerDisplayName": "Nebius AI Studio", - "providerModelId": "Qwen/QwQ-32B-Preview", - "providerGroup": "Nebius", - "quantization": "fp8", + "providerDisplayName": "DeepInfra", + "providerModelId": "meta-llama/Llama-Guard-4-12B", + "providerGroup": "DeepInfra", + "quantization": "bf16", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, "maxPromptImages": null, @@ -51785,17 +54784,18 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", + "repetition_penalty", + "response_format", "top_k", - "logit_bias", - "logprobs", - "top_logprobs" + "seed", + "min_p" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "privacyPolicyUrl": "https://deepinfra.com/privacy", + "termsOfServiceUrl": "https://deepinfra.com/terms", + "dataPolicyUrl": "https://deepinfra.com/docs/data", "paidModels": { "training": false, "retainsPrompts": false @@ -51804,8 +54804,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000009", - "completion": "0.00000027", + "prompt": "0.00000005", + "completion": "0.00000005", "image": "0", "request": "0", "webSearch": "0", @@ -51830,24 +54830,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 180, - "newest": 162, - "throughputHighToLow": 246, - "latencyLowToHigh": 46, - "pricingLowToHigh": 57, - "pricingHighToLow": 200 + "topWeekly": 183, + "newest": 21, + "throughputHighToLow": 89, + "latencyLowToHigh": 63, + "pricingLowToHigh": 96, + "pricingHighToLow": 222 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 32768, + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 163840, "maxCompletionTokens": null, - "providerModelId": "Qwen/QwQ-32B-Preview", + "providerModelId": "meta-llama/Llama-Guard-4-12B", "pricing": { - "prompt": "0.00000009", - "completion": "0.00000027", + "prompt": "0.00000005", + "completion": "0.00000005", "image": "0", "request": "0", "webSearch": "0", @@ -51861,22 +54863,25 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", + "repetition_penalty", + "response_format", "top_k", - "logit_bias", - "logprobs", - "top_logprobs" + "seed", + "min_p" ], - "inputCost": 0.09, - "outputCost": 0.27 + "inputCost": 0.05, + "outputCost": 0.05, + "throughput": 87.8985, + "latency": 476 }, { - "name": "Hyperbolic", - "slug": "hyperbolic", - "quantization": "fp8", - "context": 32768, + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": null, + "context": 1048576, "maxCompletionTokens": null, - "providerModelId": "Qwen/QwQ-32B-Preview", + "providerModelId": "meta-llama/Llama-Guard-4-12B", "pricing": { "prompt": "0.0000002", "completion": "0.0000002", @@ -51893,16 +54898,49 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "logprobs", - "top_logprobs", - "seed", - "logit_bias", "top_k", + "repetition_penalty", + "logit_bias", "min_p", - "repetition_penalty" + "response_format" ], "inputCost": 0.2, "outputCost": 0.2 + }, + { + "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", + "slug": "groq", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 1024, + "providerModelId": "meta-llama/llama-guard-4-12b", + "pricing": { + "prompt": "0.0000002", + "completion": "0.0000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "top_logprobs", + "logprobs", + "logit_bias", + "seed" + ], + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 370.611, + "latency": 757 } ] }, @@ -52095,16 +55133,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 181, + "topWeekly": 184, "newest": 261, - "throughputHighToLow": 92, - "latencyLowToHigh": 88, + "throughputHighToLow": 75, + "latencyLowToHigh": 87, "pricingLowToHigh": 293, "pricingHighToLow": 29 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", "slug": "openAi", "quantization": "unknown", "context": 128000, @@ -52137,10 +55177,13 @@ export default { "structured_outputs" ], "inputCost": 5, - "outputCost": 15 + "outputCost": 15, + "throughput": 89.162, + "latency": 604 }, { "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", "slug": "azure", "quantization": "unknown", "context": 128000, @@ -52173,22 +55216,24 @@ export default { "structured_outputs" ], "inputCost": 5, - "outputCost": 15 + "outputCost": 15, + "throughput": 146.878, + "latency": 1550 } ] }, { - "slug": "perplexity/llama-3.1-sonar-large-128k-online", + "slug": "x-ai/grok-beta", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-01T00:00:00+00:00", + "createdAt": "2024-10-20T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Perplexity: Llama 3.1 Sonar 70B Online", - "shortName": "Llama 3.1 Sonar 70B Online", - "author": "perplexity", - "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3.1-sonar-large-128k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", - "modelVersionGroupId": "e0349fb1-b84a-4f4f-9023-e7f1c1e73b33", - "contextLength": 127072, + "name": "xAI: Grok Beta", + "shortName": "Grok Beta", + "author": "x-ai", + "description": "Grok Beta is xAI's experimental language model with state-of-the-art reasoning capabilities, best for complex and multi-step use cases.\n\nIt is the successor of [Grok 2](https://x.ai/blog/grok-2) with enhanced context length.", + "modelVersionGroupId": null, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -52196,32 +55241,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", + "group": "Grok", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "perplexity/llama-3.1-sonar-large-128k-online", + "permaslug": "x-ai/grok-beta", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "aaad39f2-2e0d-46a6-8078-928b6eebe5d3", - "name": "Perplexity | perplexity/llama-3.1-sonar-large-128k-online", - "contextLength": 127072, + "id": "7f10aaf2-5317-4238-b4be-212fe355d24c", + "name": "xAI | x-ai/grok-beta", + "contextLength": 131072, "model": { - "slug": "perplexity/llama-3.1-sonar-large-128k-online", + "slug": "x-ai/grok-beta", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-01T00:00:00+00:00", + "createdAt": "2024-10-20T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Perplexity: Llama 3.1 Sonar 70B Online", - "shortName": "Llama 3.1 Sonar 70B Online", - "author": "perplexity", - "description": "Llama 3.1 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3.1-sonar-large-128k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", - "modelVersionGroupId": "e0349fb1-b84a-4f4f-9023-e7f1c1e73b33", - "contextLength": 127072, + "name": "xAI: Grok Beta", + "shortName": "Grok Beta", + "author": "x-ai", + "description": "Grok Beta is xAI's experimental language model with state-of-the-art reasoning capabilities, best for complex and multi-step use cases.\n\nIt is the successor of [Grok 2](https://x.ai/blog/grok-2) with enhanced context length.", + "modelVersionGroupId": null, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -52229,82 +55274,94 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", + "group": "Grok", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "perplexity/llama-3.1-sonar-large-128k-online", + "permaslug": "x-ai/grok-beta", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "perplexity/llama-3.1-sonar-large-128k-online", - "modelVariantPermaslug": "perplexity/llama-3.1-sonar-large-128k-online", - "providerName": "Perplexity", + "modelVariantSlug": "x-ai/grok-beta", + "modelVariantPermaslug": "x-ai/grok-beta", + "providerName": "xAI", "providerInfo": { - "name": "Perplexity", - "displayName": "Perplexity", - "slug": "perplexity", + "name": "xAI", + "displayName": "xAI", + "slug": "xai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", - "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", + "termsOfServiceUrl": "https://x.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://x.ai/legal/privacy-policy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": true, + "retentionDays": 30 } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, + "hasCompletions": true, + "isAbortable": true, "moderationRequired": false, - "group": "Perplexity", + "group": "xAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.perplexity.ai/", + "statusPageUrl": "https://status.x.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Perplexity.svg" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://x.ai/&size=256" } }, - "providerDisplayName": "Perplexity", - "providerModelId": "llama-3.1-sonar-large-128k-online", - "providerGroup": "Perplexity", + "providerDisplayName": "xAI", + "providerModelId": "grok-beta", + "providerGroup": "xAI", "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "top_k", + "stop", "frequency_penalty", - "presence_penalty" + "presence_penalty", + "seed", + "logprobs", + "top_logprobs", + "response_format" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", - "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", + "termsOfServiceUrl": "https://x.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://x.ai/legal/privacy-policy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "training": false + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.000001", - "completion": "0.000001", + "prompt": "0.000005", + "completion": "0.000015", "image": "0", - "request": "0.005", + "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -52313,12 +55370,12 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -52326,201 +55383,203 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 182, - "newest": 227, - "throughputHighToLow": 100, - "latencyLowToHigh": 234, - "pricingLowToHigh": 288, - "pricingHighToLow": 31 + "topWeekly": 185, + "newest": 185, + "throughputHighToLow": 130, + "latencyLowToHigh": 23, + "pricingLowToHigh": 291, + "pricingHighToLow": 27 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://x.ai/\\u0026size=256", "providers": [ { - "name": "Perplexity", - "slug": "perplexity", + "name": "xAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://x.ai/&size=256", + "slug": "xAi", "quantization": "unknown", - "context": 127072, + "context": 131072, "maxCompletionTokens": null, - "providerModelId": "llama-3.1-sonar-large-128k-online", + "providerModelId": "grok-beta", "pricing": { - "prompt": "0.000001", - "completion": "0.000001", + "prompt": "0.000005", + "completion": "0.000015", "image": "0", - "request": "0.005", + "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "top_k", + "stop", "frequency_penalty", - "presence_penalty" + "presence_penalty", + "seed", + "logprobs", + "top_logprobs", + "response_format" ], - "inputCost": 1, - "outputCost": 1 + "inputCost": 5, + "outputCost": 15, + "throughput": 60.0075, + "latency": 282 } ] }, { - "slug": "openai/gpt-4o-search-preview", - "hfSlug": "", - "updatedAt": "2025-04-22T22:33:59.221907+00:00", - "createdAt": "2025-03-12T22:19:09.996816+00:00", + "slug": "google/gemma-3-4b-it", + "hfSlug": "google/gemma-3-4b-it", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-03-13T22:38:30.653142+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4o Search Preview", - "shortName": "GPT-4o Search Preview", - "author": "openai", - "description": "GPT-4o Search Previewis a specialized model for web search in Chat Completions. It is trained to understand and execute web search queries.", - "modelVersionGroupId": null, - "contextLength": 128000, + "name": "Google: Gemma 3 4B (free)", + "shortName": "Gemma 3 4B (free)", + "author": "google", + "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling.", + "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", + "contextLength": 131072, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, + "group": "Gemini", + "instructType": "gemma", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "", + "", + "" + ], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "openai/gpt-4o-search-preview-2025-03-11", + "warningMessage": null, + "permaslug": "google/gemma-3-4b-it", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "f37536d3-fa09-47a3-b63c-831a1965253e", - "name": "OpenAI | openai/gpt-4o-search-preview-2025-03-11", - "contextLength": 128000, + "id": "1bbec5f0-58e0-4af2-b8ca-99c202b4ceca", + "name": "Chutes | google/gemma-3-4b-it:free", + "contextLength": 131072, "model": { - "slug": "openai/gpt-4o-search-preview", - "hfSlug": "", - "updatedAt": "2025-04-22T22:33:59.221907+00:00", - "createdAt": "2025-03-12T22:19:09.996816+00:00", + "slug": "google/gemma-3-4b-it", + "hfSlug": "google/gemma-3-4b-it", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-03-13T22:38:30.653142+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4o Search Preview", - "shortName": "GPT-4o Search Preview", - "author": "openai", - "description": "GPT-4o Search Previewis a specialized model for web search in Chat Completions. It is trained to understand and execute web search queries.", - "modelVersionGroupId": null, - "contextLength": 128000, + "name": "Google: Gemma 3 4B", + "shortName": "Gemma 3 4B", + "author": "google", + "description": "Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling.", + "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", + "contextLength": 131072, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, + "group": "Gemini", + "instructType": "gemma", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "", + "", + "" + ], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "openai/gpt-4o-search-preview-2025-03-11", + "warningMessage": null, + "permaslug": "google/gemma-3-4b-it", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/gpt-4o-search-preview", - "modelVariantPermaslug": "openai/gpt-4o-search-preview-2025-03-11", - "providerName": "OpenAI", + "modelVariantSlug": "google/gemma-3-4b-it:free", + "modelVariantPermaslug": "google/gemma-3-4b-it:free", + "providerName": "Chutes", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "training": true, + "retainsPrompts": true + } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", + "moderationRequired": false, + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "OpenAI", - "providerModelId": "gpt-4o-search-preview-2025-03-11", - "providerGroup": "OpenAI", - "quantization": null, - "variant": "standard", - "isFree": false, + "providerDisplayName": "Chutes", + "providerModelId": "unsloth/gemma-3-4b-it", + "providerGroup": "Chutes", + "quantization": "bf16", + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 16384, + "maxCompletionTokens": 8192, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "web_search_options", "max_tokens", - "response_format", - "structured_outputs" + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.0000025", - "completion": "0.00001", - "image": "0.003613", - "request": "0.035", + "prompt": "0", + "completion": "0", + "image": "0", + "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [ - { - "type": "search-threshold", - "threshold": "high", - "request": "0.05" - }, - { - "type": "search-threshold", - "threshold": "medium", - "request": "0.035" - }, - { - "type": "search-threshold", - "threshold": "low", - "request": "0.03" - } - ], + "variablePricings": [], "isHidden": false, "isDeranked": false, "isDisabled": false, @@ -52537,121 +55596,150 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 183, - "newest": 91, - "throughputHighToLow": 32, - "latencyLowToHigh": 286, - "pricingLowToHigh": 314, - "pricingHighToLow": 4 + "topWeekly": 59, + "newest": 83, + "throughputHighToLow": 70, + "latencyLowToHigh": 26, + "pricingLowToHigh": 38, + "pricingHighToLow": 241 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [ { - "name": "OpenAI", - "slug": "openAi", - "quantization": null, - "context": 128000, - "maxCompletionTokens": 16384, - "providerModelId": "gpt-4o-search-preview-2025-03-11", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "google/gemma-3-4b-it", "pricing": { - "prompt": "0.0000025", - "completion": "0.00001", - "image": "0.003613", - "request": "0.035", + "prompt": "0.00000002", + "completion": "0.00000004", + "image": "0", + "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "web_search_options", "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", "response_format", - "structured_outputs" + "top_k", + "seed", + "min_p" ], - "inputCost": 2.5, - "outputCost": 10 + "inputCost": 0.02, + "outputCost": 0.04, + "throughput": 80.2565, + "latency": 299 } ] }, { - "slug": "microsoft/phi-4-multimodal-instruct", - "hfSlug": "microsoft/Phi-4-multimodal-instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-08T01:11:24.652063+00:00", + "slug": "deepseek/deepseek-r1-distill-qwen-14b", + "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-01-29T23:39:00.13687+00:00", "hfUpdatedAt": null, - "name": "Microsoft: Phi 4 Multimodal Instruct", - "shortName": "Phi 4 Multimodal Instruct", - "author": "microsoft", - "description": "Phi-4 Multimodal Instruct is a versatile 5.6B parameter foundation model that combines advanced reasoning and instruction-following capabilities across both text and visual inputs, providing accurate text outputs. The unified architecture enables efficient, low-latency inference, suitable for edge and mobile deployments. Phi-4 Multimodal Instruct supports text inputs in multiple languages including Arabic, Chinese, English, French, German, Japanese, Spanish, and more, with visual input optimized primarily for English. It delivers impressive performance on multimodal tasks involving mathematical, scientific, and document reasoning, providing developers and enterprises a powerful yet compact model for sophisticated interactive applications. For more information, see the [Phi-4 Multimodal blog post](https://azure.microsoft.com/en-us/blog/empowering-innovation-the-next-generation-of-the-phi-family/).\n", - "modelVersionGroupId": null, - "contextLength": 131072, + "name": "DeepSeek: R1 Distill Qwen 14B", + "shortName": "R1 Distill Qwen 14B", + "author": "deepseek", + "description": "DeepSeek R1 Distill Qwen 14B is a distilled large language model based on [Qwen 2.5 14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 69.7\n- MATH-500 pass@1: 93.9\n- CodeForces Rating: 1481\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", + "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", + "contextLength": 64000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Qwen", + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "microsoft/phi-4-multimodal-instruct", - "reasoningConfig": null, - "features": {}, + "warningMessage": "", + "permaslug": "deepseek/deepseek-r1-distill-qwen-14b", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "345f416d-3f33-4530-8033-614ac900a8fd", - "name": "DeepInfra | microsoft/phi-4-multimodal-instruct", - "contextLength": 131072, + "id": "a18c0e49-6efb-439a-9287-fb2f398f3c5a", + "name": "Novita | deepseek/deepseek-r1-distill-qwen-14b", + "contextLength": 64000, "model": { - "slug": "microsoft/phi-4-multimodal-instruct", - "hfSlug": "microsoft/Phi-4-multimodal-instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-08T01:11:24.652063+00:00", + "slug": "deepseek/deepseek-r1-distill-qwen-14b", + "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-01-29T23:39:00.13687+00:00", "hfUpdatedAt": null, - "name": "Microsoft: Phi 4 Multimodal Instruct", - "shortName": "Phi 4 Multimodal Instruct", - "author": "microsoft", - "description": "Phi-4 Multimodal Instruct is a versatile 5.6B parameter foundation model that combines advanced reasoning and instruction-following capabilities across both text and visual inputs, providing accurate text outputs. The unified architecture enables efficient, low-latency inference, suitable for edge and mobile deployments. Phi-4 Multimodal Instruct supports text inputs in multiple languages including Arabic, Chinese, English, French, German, Japanese, Spanish, and more, with visual input optimized primarily for English. It delivers impressive performance on multimodal tasks involving mathematical, scientific, and document reasoning, providing developers and enterprises a powerful yet compact model for sophisticated interactive applications. For more information, see the [Phi-4 Multimodal blog post](https://azure.microsoft.com/en-us/blog/empowering-innovation-the-next-generation-of-the-phi-family/).\n", - "modelVersionGroupId": null, + "name": "DeepSeek: R1 Distill Qwen 14B", + "shortName": "R1 Distill Qwen 14B", + "author": "deepseek", + "description": "DeepSeek R1 Distill Qwen 14B is a distilled large language model based on [Qwen 2.5 14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 69.7\n- MATH-500 pass@1: 93.9\n- CodeForces Rating: 1481\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", + "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Qwen", + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "microsoft/phi-4-multimodal-instruct", - "reasoningConfig": null, - "features": {} + "warningMessage": "", + "permaslug": "deepseek/deepseek-r1-distill-qwen-14b", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "microsoft/phi-4-multimodal-instruct", - "modelVariantPermaslug": "microsoft/phi-4-multimodal-instruct", - "providerName": "DeepInfra", + "modelVariantSlug": "deepseek/deepseek-r1-distill-qwen-14b", + "modelVariantPermaslug": "deepseek/deepseek-r1-distill-qwen-14b", + "providerName": "Novita", "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", + "name": "Novita", + "displayName": "NovitaAI", + "slug": "novita", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": false + "training": false } }, "headquarters": "US", @@ -52659,58 +55747,58 @@ export default { "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "DeepInfra", + "group": "Novita", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.novita.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/DeepInfra.webp" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "invertRequired": true } }, - "providerDisplayName": "DeepInfra", - "providerModelId": "microsoft/Phi-4-multimodal-instruct", - "providerGroup": "DeepInfra", - "quantization": "bf16", + "providerDisplayName": "NovitaAI", + "providerModelId": "deepseek/deepseek-r1-distill-qwen-14b", + "providerGroup": "Novita", + "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": 1, - "maxTokensPerImage": 3537, + "maxCompletionTokens": 64000, + "maxPromptImages": null, + "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", "seed", - "min_p" + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": false + "training": false }, - "training": false, - "retainsPrompts": false + "training": false }, "pricing": { - "prompt": "0.00000005", - "completion": "0.0000001", - "image": "0.00017685", + "prompt": "0.00000015", + "completion": "0.00000015", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -52721,7 +55809,7 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, @@ -52733,25 +55821,27 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 184, - "newest": 97, - "throughputHighToLow": 112, - "latencyLowToHigh": 277, - "pricingLowToHigh": 97, - "pricingHighToLow": 221 + "topWeekly": 187, + "newest": 135, + "throughputHighToLow": 124, + "latencyLowToHigh": 192, + "pricingLowToHigh": 51, + "pricingHighToLow": 184 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "microsoft/Phi-4-multimodal-instruct", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": null, + "context": 64000, + "maxCompletionTokens": 64000, + "providerModelId": "deepseek/deepseek-r1-distill-qwen-14b", "pricing": { - "prompt": "0.00000005", - "completion": "0.0000001", - "image": "0.00017685", + "prompt": "0.00000015", + "completion": "0.00000015", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -52761,17 +55851,58 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", + "seed", + "top_k", + "min_p", "repetition_penalty", - "response_format", + "logit_bias" + ], + "inputCost": 0.15, + "outputCost": 0.15, + "throughput": 45.768, + "latency": 1328 + }, + { + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 32768, + "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B", + "pricing": { + "prompt": "0.0000016", + "completion": "0.0000016", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", "top_k", - "seed", - "min_p" + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" ], - "inputCost": 0.05, - "outputCost": 0.1 + "inputCost": 1.6, + "outputCost": 1.6, + "throughput": 170.562, + "latency": 362 } ] }, @@ -52938,16 +56069,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 185, + "topWeekly": 188, "newest": 247, "throughputHighToLow": 306, - "latencyLowToHigh": 263, - "pricingLowToHigh": 214, - "pricingHighToLow": 103 + "latencyLowToHigh": 268, + "pricingLowToHigh": 213, + "pricingHighToLow": 104 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "unknown", "context": 16000, @@ -52976,20 +56109,22 @@ export default { "logit_bias" ], "inputCost": 0.9, - "outputCost": 0.9 + "outputCost": 0.9, + "throughput": 11.4455, + "latency": 2470 } ] }, { - "slug": "amazon/nova-micro-v1", - "hfSlug": "", + "slug": "openai/gpt-4-1106-preview", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-05T22:20:37.90344+00:00", + "createdAt": "2023-11-06T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Amazon: Nova Micro 1.0", - "shortName": "Nova Micro 1.0", - "author": "amazon", - "description": "Amazon Nova Micro 1.0 is a text-only model that delivers the lowest latency responses in the Amazon Nova family of models at a very low cost. With a context length of 128K tokens and optimized for speed and cost, Amazon Nova Micro excels at tasks such as text summarization, translation, content classification, interactive chat, and brainstorming. It has simple mathematical reasoning and coding abilities.", + "name": "OpenAI: GPT-4 Turbo (older v1106)", + "shortName": "GPT-4 Turbo (older v1106)", + "author": "openai", + "description": "The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.\n\nTraining data: up to April 2023.", "modelVersionGroupId": null, "contextLength": 128000, "inputModalities": [ @@ -52999,30 +56134,30 @@ export default { "text" ], "hasTextOutput": true, - "group": "Nova", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "amazon/nova-micro-v1", + "permaslug": "openai/gpt-4-1106-preview", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "474f0074-66f9-42f0-a866-81a2ffebb001", - "name": "Amazon Bedrock | amazon/nova-micro-v1", + "id": "8744de26-f64f-41cf-bd0e-950a83d1a923", + "name": "OpenAI | openai/gpt-4-1106-preview", "contextLength": 128000, "model": { - "slug": "amazon/nova-micro-v1", - "hfSlug": "", + "slug": "openai/gpt-4-1106-preview", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-05T22:20:37.90344+00:00", + "createdAt": "2023-11-06T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Amazon: Nova Micro 1.0", - "shortName": "Nova Micro 1.0", - "author": "amazon", - "description": "Amazon Nova Micro 1.0 is a text-only model that delivers the lowest latency responses in the Amazon Nova family of models at a very low cost. With a context length of 128K tokens and optimized for speed and cost, Amazon Nova Micro excels at tasks such as text summarization, translation, content classification, interactive chat, and brainstorming. It has simple mathematical reasoning and coding abilities.", + "name": "OpenAI: GPT-4 Turbo (older v1106)", + "shortName": "GPT-4 Turbo (older v1106)", + "author": "openai", + "description": "The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.\n\nTraining data: up to April 2023.", "modelVersionGroupId": null, "contextLength": 128000, "inputModalities": [ @@ -53032,85 +56167,99 @@ export default { "text" ], "hasTextOutput": true, - "group": "Nova", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "amazon/nova-micro-v1", + "permaslug": "openai/gpt-4-1106-preview", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "amazon/nova-micro-v1", - "modelVariantPermaslug": "amazon/nova-micro-v1", - "providerName": "Amazon Bedrock", + "modelVariantSlug": "openai/gpt-4-1106-preview", + "modelVariantPermaslug": "openai/gpt-4-1106-preview", + "providerName": "OpenAI", "providerInfo": { - "name": "Amazon Bedrock", - "displayName": "Amazon Bedrock", - "slug": "amazon-bedrock", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://aws.amazon.com/service-terms/", - "privacyPolicyUrl": "https://aws.amazon.com/privacy", - "dataPolicyUrl": "https://docs.aws.amazon.com/bedrock/latest/userguide/data-protection.html", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false - } + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, + "hasCompletions": true, + "isAbortable": true, "moderationRequired": true, - "group": "Amazon Bedrock", + "group": "OpenAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Bedrock.svg" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Amazon Bedrock", - "providerModelId": "us.amazon.nova-micro-v1:0", - "providerGroup": "Amazon Bedrock", - "quantization": null, + "providerDisplayName": "OpenAI", + "providerModelId": "gpt-4-1106-preview", + "providerGroup": "OpenAI", + "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 5120, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "top_k", - "stop" + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" ], "isByok": false, "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://aws.amazon.com/service-terms/", - "privacyPolicyUrl": "https://aws.amazon.com/privacy", - "dataPolicyUrl": "https://docs.aws.amazon.com/bedrock/latest/userguide/data-protection.html", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, + "requiresUserIds": true, "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.000000035", - "completion": "0.00000014", + "prompt": "0.00001", + "completion": "0.00003", "image": "0", "request": "0", "webSearch": "0", @@ -53126,7 +56275,7 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -53134,24 +56283,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 186, - "newest": 160, - "throughputHighToLow": 8, - "latencyLowToHigh": 51, - "pricingLowToHigh": 90, - "pricingHighToLow": 228 + "topWeekly": 189, + "newest": 301, + "throughputHighToLow": 222, + "latencyLowToHigh": 105, + "pricingLowToHigh": 304, + "pricingHighToLow": 16 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "Amazon Bedrock", - "slug": "amazonBedrock", - "quantization": null, + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": "unknown", "context": 128000, - "maxCompletionTokens": 5120, - "providerModelId": "us.amazon.nova-micro-v1:0", + "maxCompletionTokens": 4096, + "providerModelId": "gpt-4-1106-preview", "pricing": { - "prompt": "0.000000035", - "completion": "0.00000014", + "prompt": "0.00001", + "completion": "0.00003", "image": "0", "request": "0", "webSearch": "0", @@ -53160,29 +56311,39 @@ export default { }, "supportedParameters": [ "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "top_k", - "stop" + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" ], - "inputCost": 0.04, - "outputCost": 0.14 + "inputCost": 10, + "outputCost": 30, + "throughput": 36.796, + "latency": 699 } ] }, { - "slug": "cohere/command-r-plus-08-2024", - "hfSlug": null, + "slug": "cohere/command-a", + "hfSlug": "CohereForAI/c4ai-command-a-03-2025", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-30T00:00:00+00:00", + "createdAt": "2025-03-13T19:32:22.069645+00:00", "hfUpdatedAt": null, - "name": "Cohere: Command R+ (08-2024)", - "shortName": "Command R+ (08-2024)", + "name": "Cohere: Command A", + "shortName": "Command A", "author": "cohere", - "description": "command-r-plus-08-2024 is an update of the [Command R+](/models/cohere/command-r-plus) with roughly 50% higher throughput and 25% lower latencies as compared to the previous Command R+ version, while keeping the hardware footprint the same.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", + "description": "Command A is an open-weights 111B parameter model with a 256k context window focused on delivering great performance across agentic, multilingual, and coding use cases.\nCompared to other leading proprietary and open-weights models Command A delivers maximum performance with minimum hardware costs, excelling on business-critical agentic and multilingual tasks.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 256000, "inputModalities": [ "text" ], @@ -53190,32 +56351,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Cohere", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cohere/command-r-plus-08-2024", + "permaslug": "cohere/command-a-03-2025", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "cd63714a-d459-4806-bdf2-0dfea4f6614c", - "name": "Cohere | cohere/command-r-plus-08-2024", - "contextLength": 128000, + "id": "2e152629-dac1-41b0-8f4f-f8c449d2d504", + "name": "Cohere | cohere/command-a-03-2025", + "contextLength": 256000, "model": { - "slug": "cohere/command-r-plus-08-2024", - "hfSlug": null, + "slug": "cohere/command-a", + "hfSlug": "CohereForAI/c4ai-command-a-03-2025", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-30T00:00:00+00:00", + "createdAt": "2025-03-13T19:32:22.069645+00:00", "hfUpdatedAt": null, - "name": "Cohere: Command R+ (08-2024)", - "shortName": "Command R+ (08-2024)", + "name": "Cohere: Command A", + "shortName": "Command A", "author": "cohere", - "description": "command-r-plus-08-2024 is an update of the [Command R+](/models/cohere/command-r-plus) with roughly 50% higher throughput and 25% lower latencies as compared to the previous Command R+ version, while keeping the hardware footprint the same.\n\nRead the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", + "description": "Command A is an open-weights 111B parameter model with a 256k context window focused on delivering great performance across agentic, multilingual, and coding use cases.\nCompared to other leading proprietary and open-weights models Command A delivers maximum performance with minimum hardware costs, excelling on business-critical agentic and multilingual tasks.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 256000, "inputModalities": [ "text" ], @@ -53223,19 +56384,19 @@ export default { "text" ], "hasTextOutput": true, - "group": "Cohere", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cohere/command-r-plus-08-2024", + "permaslug": "cohere/command-a-03-2025", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "cohere/command-r-plus-08-2024", - "modelVariantPermaslug": "cohere/command-r-plus-08-2024", + "modelVariantSlug": "cohere/command-a", + "modelVariantPermaslug": "cohere/command-a-03-2025", "providerName": "Cohere", "providerInfo": { "name": "Cohere", @@ -53268,18 +56429,17 @@ export default { } }, "providerDisplayName": "Cohere", - "providerModelId": "command-r-plus-08-2024", + "providerModelId": "command-a-03-2025", "providerGroup": "Cohere", - "quantization": "unknown", + "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4000, + "maxCompletionTokens": 8192, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", "max_tokens", "temperature", "top_p", @@ -53318,7 +56478,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -53326,27 +56486,28 @@ export default { "hasCompletions": false, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 187, - "newest": 213, - "throughputHighToLow": 207, - "latencyLowToHigh": 56, - "pricingLowToHigh": 263, - "pricingHighToLow": 57 + "topWeekly": 190, + "newest": 89, + "throughputHighToLow": 82, + "latencyLowToHigh": 21, + "pricingLowToHigh": 259, + "pricingHighToLow": 53 }, + "authorIcon": "https://openrouter.ai/images/icons/Cohere.png", "providers": [ { "name": "Cohere", + "icon": "https://openrouter.ai/images/icons/Cohere.png", "slug": "cohere", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 4000, - "providerModelId": "command-r-plus-08-2024", + "quantization": null, + "context": 256000, + "maxCompletionTokens": 8192, + "providerModelId": "command-a-03-2025", "pricing": { "prompt": "0.0000025", "completion": "0.00001", @@ -53357,7 +56518,6 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", "max_tokens", "temperature", "top_p", @@ -53370,22 +56530,24 @@ export default { "structured_outputs" ], "inputCost": 2.5, - "outputCost": 10 + "outputCost": 10, + "throughput": 92.2645, + "latency": 214 } ] }, { - "slug": "mistralai/mistral-large-2407", - "hfSlug": "", + "slug": "openai/gpt-4", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-19T01:06:55.27469+00:00", + "createdAt": "2023-05-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Mistral Large 2407", - "shortName": "Mistral Large 2407", - "author": "mistralai", - "description": "This is Mistral AI's flagship model, Mistral Large 2 (version mistral-large-2407). It's a proprietary weights-available model and excels at reasoning, code, JSON, chat, and more. Read the launch announcement [here](https://mistral.ai/news/mistral-large-2407/).\n\nIt supports dozens of languages including French, German, Spanish, Italian, Portuguese, Arabic, Hindi, Russian, Chinese, Japanese, and Korean, along with 80+ coding languages including Python, Java, C, C++, JavaScript, and Bash. Its long context window allows precise information recall from large documents.\n", - "modelVersionGroupId": "83129748-0564-4485-982a-d7a37a1ef3ec", - "contextLength": 131072, + "name": "OpenAI: GPT-4", + "shortName": "GPT-4", + "author": "openai", + "description": "OpenAI's flagship model, GPT-4 is a large-scale multimodal language model capable of solving difficult problems with greater accuracy than previous models due to its broader general knowledge and advanced reasoning capabilities. Training data: up to Sep 2021.", + "modelVersionGroupId": null, + "contextLength": 8191, "inputModalities": [ "text" ], @@ -53393,32 +56555,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-large-2407", + "permaslug": "openai/gpt-4", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "4a128170-b056-42d7-8462-a5cea647f9ad", - "name": "Mistral | mistralai/mistral-large-2407", - "contextLength": 131072, + "id": "355b1df4-06c8-4c36-a091-3d50477095fb", + "name": "OpenAI | openai/gpt-4", + "contextLength": 8191, "model": { - "slug": "mistralai/mistral-large-2407", - "hfSlug": "", + "slug": "openai/gpt-4", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-19T01:06:55.27469+00:00", + "createdAt": "2023-05-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Mistral Large 2407", - "shortName": "Mistral Large 2407", - "author": "mistralai", - "description": "This is Mistral AI's flagship model, Mistral Large 2 (version mistral-large-2407). It's a proprietary weights-available model and excels at reasoning, code, JSON, chat, and more. Read the launch announcement [here](https://mistral.ai/news/mistral-large-2407/).\n\nIt supports dozens of languages including French, German, Spanish, Italian, Portuguese, Arabic, Hindi, Russian, Chinese, Japanese, and Korean, along with 80+ coding languages including Python, Java, C, C++, JavaScript, and Bash. Its long context window allows precise information recall from large documents.\n", - "modelVersionGroupId": "83129748-0564-4485-982a-d7a37a1ef3ec", - "contextLength": 128000, + "name": "OpenAI: GPT-4", + "shortName": "GPT-4", + "author": "openai", + "description": "OpenAI's flagship model, GPT-4 is a large-scale multimodal language model capable of solving difficult problems with greater accuracy than previous models due to its broader general knowledge and advanced reasoning capabilities. Training data: up to Sep 2021.", + "modelVersionGroupId": null, + "contextLength": 8191, "inputModalities": [ "text" ], @@ -53426,59 +56588,62 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-large-2407", + "permaslug": "openai/gpt-4", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "mistralai/mistral-large-2407", - "modelVariantPermaslug": "mistralai/mistral-large-2407", - "providerName": "Mistral", + "modelVariantSlug": "openai/gpt-4", + "modelVariantPermaslug": "openai/gpt-4", + "providerName": "OpenAI", "providerInfo": { - "name": "Mistral", - "displayName": "Mistral", - "slug": "mistral", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 - } + }, + "requiresUserIds": true }, - "headquarters": "FR", + "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, - "moderationRequired": false, - "group": "Mistral", + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Mistral.png" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Mistral", - "providerModelId": "mistral-large-2407", - "providerGroup": "Mistral", - "quantization": null, + "providerDisplayName": "OpenAI", + "providerModelId": "gpt-4", + "providerGroup": "OpenAI", + "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -53490,27 +56655,31 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "response_format", - "structured_outputs", - "seed" + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 }, + "requiresUserIds": true, "training": false, "retainsPrompts": true, "retentionDays": 30 }, "pricing": { - "prompt": "0.000002", - "completion": "0.000006", + "prompt": "0.00003", + "completion": "0.00006", "image": "0", "request": "0", "webSearch": "0", @@ -53526,33 +56695,34 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 188, - "newest": 168, - "throughputHighToLow": 202, - "latencyLowToHigh": 63, - "pricingLowToHigh": 244, - "pricingHighToLow": 73 + "topWeekly": 191, + "newest": 317, + "throughputHighToLow": 260, + "latencyLowToHigh": 113, + "pricingLowToHigh": 312, + "pricingHighToLow": 5 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "Mistral", - "slug": "mistral", - "quantization": null, - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "mistral-large-2407", + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": "unknown", + "context": 8191, + "maxCompletionTokens": 4096, + "providerModelId": "gpt-4", "pricing": { - "prompt": "0.000002", - "completion": "0.000006", + "prompt": "0.00003", + "completion": "0.00006", "image": "0", "request": "0", "webSearch": "0", @@ -53568,27 +56738,68 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "response_format", - "structured_outputs", - "seed" + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" ], - "inputCost": 2, - "outputCost": 6 + "inputCost": 30, + "outputCost": 60, + "throughput": 24.3795, + "latency": 717 + }, + { + "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", + "slug": "azure", + "quantization": "unknown", + "context": 8191, + "maxCompletionTokens": 4096, + "providerModelId": "gpt-4", + "pricing": { + "prompt": "0.00003", + "completion": "0.00006", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" + ], + "inputCost": 30, + "outputCost": 60, + "throughput": 34.887, + "latency": 7744.5 } ] }, { - "slug": "google/gemma-2-9b-it", - "hfSlug": "google/gemma-2-9b-it", + "slug": "mistralai/mistral-7b-instruct-v0.2", + "hfSlug": "mistralai/Mistral-7B-Instruct-v0.2", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-06-28T00:00:00+00:00", + "createdAt": "2023-12-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Google: Gemma 2 9B (free)", - "shortName": "Gemma 2 9B (free)", - "author": "google", - "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).", - "modelVersionGroupId": null, - "contextLength": 8192, + "name": "Mistral: Mistral 7B Instruct v0.2", + "shortName": "Mistral 7B Instruct v0.2", + "author": "mistralai", + "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\nAn improved version of [Mistral 7B Instruct](/modelsmistralai/mistral-7b-instruct-v0.1), with the following changes:\n\n- 32k context window (vs 8k context in v0.1)\n- Rope-theta = 1e6\n- No Sliding-Window Attention", + "modelVersionGroupId": "1d07cc56-c54d-4587-b785-5093496397a4", + "contextLength": 32768, "inputModalities": [ "text" ], @@ -53596,36 +56807,35 @@ export default { "text" ], "hasTextOutput": true, - "group": "Gemini", - "instructType": "gemma", + "group": "Mistral", + "instructType": "mistral", "defaultSystem": null, "defaultStops": [ - "", - "", - "" + "[INST]", + "" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "google/gemma-2-9b-it", + "permaslug": "mistralai/mistral-7b-instruct-v0.2", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "c21a8adc-4c1d-4605-af09-eb65c192db01", - "name": "Chutes | google/gemma-2-9b-it:free", - "contextLength": 8192, + "id": "ae921682-8673-4287-b3bd-459304097932", + "name": "Together | mistralai/mistral-7b-instruct-v0.2", + "contextLength": 32768, "model": { - "slug": "google/gemma-2-9b-it", - "hfSlug": "google/gemma-2-9b-it", + "slug": "mistralai/mistral-7b-instruct-v0.2", + "hfSlug": "mistralai/Mistral-7B-Instruct-v0.2", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-06-28T00:00:00+00:00", + "createdAt": "2023-12-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Google: Gemma 2 9B", - "shortName": "Gemma 2 9B", - "author": "google", - "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).", - "modelVersionGroupId": null, - "contextLength": 8192, + "name": "Mistral: Mistral 7B Instruct v0.2", + "shortName": "Mistral 7B Instruct v0.2", + "author": "mistralai", + "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\nAn improved version of [Mistral 7B Instruct](/modelsmistralai/mistral-7b-instruct-v0.1), with the following changes:\n\n- 32k context window (vs 8k context in v0.1)\n- Rope-theta = 1e6\n- No Sliding-Window Attention", + "modelVersionGroupId": "1d07cc56-c54d-4587-b785-5093496397a4", + "contextLength": 32768, "inputModalities": [ "text" ], @@ -53633,41 +56843,42 @@ export default { "text" ], "hasTextOutput": true, - "group": "Gemini", - "instructType": "gemma", + "group": "Mistral", + "instructType": "mistral", "defaultSystem": null, "defaultStops": [ - "", - "", - "" + "[INST]", + "" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "google/gemma-2-9b-it", + "permaslug": "mistralai/mistral-7b-instruct-v0.2", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "google/gemma-2-9b-it:free", - "modelVariantPermaslug": "google/gemma-2-9b-it:free", - "providerName": "Chutes", + "modelVariantSlug": "mistralai/mistral-7b-instruct-v0.2", + "modelVariantPermaslug": "mistralai/mistral-7b-instruct-v0.2", + "providerName": "Together", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "Together", + "displayName": "Together", + "slug": "together", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false } }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Chutes", + "group": "Together", "editors": [], "owners": [], "isMultipartSupported": true, @@ -53675,18 +56886,18 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" } }, - "providerDisplayName": "Chutes", - "providerModelId": "unsloth/gemma-2-9b-it", - "providerGroup": "Chutes", - "quantization": null, - "variant": "free", - "isFree": true, + "providerDisplayName": "Together", + "providerModelId": "mistralai/Mistral-7B-Instruct-v0.2", + "providerGroup": "Together", + "quantization": "unknown", + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 8192, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -53696,28 +56907,27 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", "top_k", - "min_p", "repetition_penalty", - "logprobs", "logit_bias", - "top_logprobs" + "min_p", + "response_format" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.0000002", + "completion": "0.0000002", "image": "0", "request": "0", "webSearch": "0", @@ -53741,85 +56951,23 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 101, - "newest": 240, - "throughputHighToLow": 42, - "latencyLowToHigh": 8, - "pricingLowToHigh": 69, - "pricingHighToLow": 239 + "topWeekly": 192, + "newest": 291, + "throughputHighToLow": 11, + "latencyLowToHigh": 1, + "pricingLowToHigh": 149, + "pricingHighToLow": 173 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 8192, - "maxCompletionTokens": null, - "providerModelId": "google/gemma-2-9b-it", - "pricing": { - "prompt": "0.00000002", - "completion": "0.00000006", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" - ], - "inputCost": 0.02, - "outputCost": 0.06 - }, - { - "name": "NovitaAI", - "slug": "novitaAi", + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", "quantization": "unknown", - "context": 8192, + "context": 32768, "maxCompletionTokens": null, - "providerModelId": "google/gemma-2-9b-it", - "pricing": { - "prompt": "0.00000008", - "completion": "0.00000008", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" - ], - "inputCost": 0.08, - "outputCost": 0.08 - }, - { - "name": "Groq", - "slug": "groq", - "quantization": null, - "context": 8192, - "maxCompletionTokens": 8192, - "providerModelId": "gemma2-9b-it", + "providerModelId": "mistralai/Mistral-7B-Instruct-v0.2", "pricing": { "prompt": "0.0000002", "completion": "0.0000002", @@ -53829,38 +56977,6 @@ export default { "internalReasoning": "0", "discount": 0 }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "top_logprobs", - "logprobs", - "logit_bias", - "seed" - ], - "inputCost": 0.2, - "outputCost": 0.2 - }, - { - "name": "Together", - "slug": "together", - "quantization": null, - "context": 8192, - "maxCompletionTokens": 8192, - "providerModelId": "google/gemma-2-9b-it", - "pricing": { - "prompt": "0.0000003", - "completion": "0.0000003", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, "supportedParameters": [ "max_tokens", "temperature", @@ -53874,159 +56990,194 @@ export default { "min_p", "response_format" ], - "inputCost": 0.3, - "outputCost": 0.3 + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 217.773, + "latency": 124 } ] }, { - "slug": "cohere/command-a", - "hfSlug": "CohereForAI/c4ai-command-a-03-2025", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-13T19:32:22.069645+00:00", + "slug": "perplexity/sonar-reasoning-pro", + "hfSlug": "", + "updatedAt": "2025-05-05T14:36:52.712881+00:00", + "createdAt": "2025-03-07T02:08:28.125446+00:00", "hfUpdatedAt": null, - "name": "Cohere: Command A", - "shortName": "Command A", - "author": "cohere", - "description": "Command A is an open-weights 111B parameter model with a 256k context window focused on delivering great performance across agentic, multilingual, and coding use cases.\nCompared to other leading proprietary and open-weights models Command A delivers maximum performance with minimum hardware costs, excelling on business-critical agentic and multilingual tasks.", + "name": "Perplexity: Sonar Reasoning Pro", + "shortName": "Sonar Reasoning Pro", + "author": "perplexity", + "description": "Note: Sonar Pro pricing includes Perplexity search pricing. See [details here](https://docs.perplexity.ai/guides/pricing#detailed-pricing-breakdown-for-sonar-reasoning-pro-and-sonar-pro)\n\nSonar Reasoning Pro is a premier reasoning model powered by DeepSeek R1 with Chain of Thought (CoT). Designed for advanced use cases, it supports in-depth, multi-step queries with a larger context window and can surface more citations per search, enabling more comprehensive and extensible responses.", "modelVersionGroupId": null, - "contextLength": 256000, + "contextLength": 128000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, "group": "Other", - "instructType": null, + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cohere/command-a-03-2025", - "reasoningConfig": null, - "features": {}, + "permaslug": "perplexity/sonar-reasoning-pro", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "2e152629-dac1-41b0-8f4f-f8c449d2d504", - "name": "Cohere | cohere/command-a-03-2025", - "contextLength": 256000, + "id": "0d28660e-435a-4853-a2c6-9d916df28fc7", + "name": "Perplexity | perplexity/sonar-reasoning-pro", + "contextLength": 128000, "model": { - "slug": "cohere/command-a", - "hfSlug": "CohereForAI/c4ai-command-a-03-2025", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-13T19:32:22.069645+00:00", + "slug": "perplexity/sonar-reasoning-pro", + "hfSlug": "", + "updatedAt": "2025-05-05T14:36:52.712881+00:00", + "createdAt": "2025-03-07T02:08:28.125446+00:00", "hfUpdatedAt": null, - "name": "Cohere: Command A", - "shortName": "Command A", - "author": "cohere", - "description": "Command A is an open-weights 111B parameter model with a 256k context window focused on delivering great performance across agentic, multilingual, and coding use cases.\nCompared to other leading proprietary and open-weights models Command A delivers maximum performance with minimum hardware costs, excelling on business-critical agentic and multilingual tasks.", + "name": "Perplexity: Sonar Reasoning Pro", + "shortName": "Sonar Reasoning Pro", + "author": "perplexity", + "description": "Note: Sonar Pro pricing includes Perplexity search pricing. See [details here](https://docs.perplexity.ai/guides/pricing#detailed-pricing-breakdown-for-sonar-reasoning-pro-and-sonar-pro)\n\nSonar Reasoning Pro is a premier reasoning model powered by DeepSeek R1 with Chain of Thought (CoT). Designed for advanced use cases, it supports in-depth, multi-step queries with a larger context window and can surface more citations per search, enabling more comprehensive and extensible responses.", "modelVersionGroupId": null, - "contextLength": 256000, + "contextLength": 128000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, "group": "Other", - "instructType": null, + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cohere/command-a-03-2025", - "reasoningConfig": null, - "features": {} + "permaslug": "perplexity/sonar-reasoning-pro", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "cohere/command-a", - "modelVariantPermaslug": "cohere/command-a-03-2025", - "providerName": "Cohere", + "modelVariantSlug": "perplexity/sonar-reasoning-pro", + "modelVariantPermaslug": "perplexity/sonar-reasoning-pro", + "providerName": "Perplexity", "providerInfo": { - "name": "Cohere", - "displayName": "Cohere", - "slug": "cohere", + "name": "Perplexity", + "displayName": "Perplexity", + "slug": "perplexity", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://cohere.com/privacy", - "termsOfServiceUrl": "https://cohere.com/terms-of-use", + "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", + "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false } }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": false, - "isAbortable": true, + "isAbortable": false, "moderationRequired": false, - "group": "Cohere", + "group": "Perplexity", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.cohere.ai/", + "statusPageUrl": "https://status.perplexity.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Cohere.png" + "url": "/images/icons/Perplexity.svg" } }, - "providerDisplayName": "Cohere", - "providerModelId": "command-a-03-2025", - "providerGroup": "Cohere", + "providerDisplayName": "Perplexity", + "providerModelId": "sonar-reasoning-pro", + "providerGroup": "Perplexity", "quantization": null, "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 8192, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", + "reasoning", + "include_reasoning", + "web_search_options", "top_k", - "seed", - "response_format", - "structured_outputs" + "frequency_penalty", + "presence_penalty" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://cohere.com/privacy", - "termsOfServiceUrl": "https://cohere.com/terms-of-use", + "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", + "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, "pricing": { - "prompt": "0.0000025", - "completion": "0.00001", + "prompt": "0.000002", + "completion": "0.000008", "image": "0", "request": "0", - "webSearch": "0", + "webSearch": "0.005", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [], + "variablePricings": [ + { + "type": "search-threshold", + "threshold": "high", + "request": "0.014" + }, + { + "type": "search-threshold", + "threshold": "medium", + "request": "0.01" + }, + { + "type": "search-threshold", + "threshold": "low", + "request": "0.006" + } + ], "isHidden": false, "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, @@ -54038,27 +57189,29 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 190, - "newest": 89, - "throughputHighToLow": 81, - "latencyLowToHigh": 14, - "pricingLowToHigh": 259, - "pricingHighToLow": 53 + "topWeekly": 193, + "newest": 98, + "throughputHighToLow": 157, + "latencyLowToHigh": 261, + "pricingLowToHigh": 249, + "pricingHighToLow": 69 }, + "authorIcon": "https://openrouter.ai/images/icons/Perplexity.svg", "providers": [ { - "name": "Cohere", - "slug": "cohere", + "name": "Perplexity", + "icon": "https://openrouter.ai/images/icons/Perplexity.svg", + "slug": "perplexity", "quantization": null, - "context": 256000, - "maxCompletionTokens": 8192, - "providerModelId": "command-a-03-2025", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "sonar-reasoning-pro", "pricing": { - "prompt": "0.0000025", - "completion": "0.00001", + "prompt": "0.000002", + "completion": "0.000008", "image": "0", "request": "0", - "webSearch": "0", + "webSearch": "0.005", "internalReasoning": "0", "discount": 0 }, @@ -54066,31 +57219,32 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", + "reasoning", + "include_reasoning", + "web_search_options", "top_k", - "seed", - "response_format", - "structured_outputs" + "frequency_penalty", + "presence_penalty" ], - "inputCost": 2.5, - "outputCost": 10 + "inputCost": 2, + "outputCost": 8, + "throughput": 53.579, + "latency": 2171 } ] }, { - "slug": "openai/gpt-4", - "hfSlug": null, + "slug": "google/gemma-2-9b-it", + "hfSlug": "google/gemma-2-9b-it", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-05-28T00:00:00+00:00", + "createdAt": "2024-06-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4", - "shortName": "GPT-4", - "author": "openai", - "description": "OpenAI's flagship model, GPT-4 is a large-scale multimodal language model capable of solving difficult problems with greater accuracy than previous models due to its broader general knowledge and advanced reasoning capabilities. Training data: up to Sep 2021.", + "name": "Google: Gemma 2 9B (free)", + "shortName": "Gemma 2 9B (free)", + "author": "google", + "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).", "modelVersionGroupId": null, - "contextLength": 8191, + "contextLength": 8192, "inputModalities": [ "text" ], @@ -54098,32 +57252,36 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, + "group": "Gemini", + "instructType": "gemma", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "", + "", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-4", + "permaslug": "google/gemma-2-9b-it", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "355b1df4-06c8-4c36-a091-3d50477095fb", - "name": "OpenAI | openai/gpt-4", - "contextLength": 8191, + "id": "c21a8adc-4c1d-4605-af09-eb65c192db01", + "name": "Chutes | google/gemma-2-9b-it:free", + "contextLength": 8192, "model": { - "slug": "openai/gpt-4", - "hfSlug": null, + "slug": "google/gemma-2-9b-it", + "hfSlug": "google/gemma-2-9b-it", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-05-28T00:00:00+00:00", + "createdAt": "2024-06-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4", - "shortName": "GPT-4", - "author": "openai", - "description": "OpenAI's flagship model, GPT-4 is a large-scale multimodal language model capable of solving difficult problems with greater accuracy than previous models due to its broader general knowledge and advanced reasoning capabilities. Training data: up to Sep 2021.", + "name": "Google: Gemma 2 9B", + "shortName": "Gemma 2 9B", + "author": "google", + "description": "Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.\n\nDesigned for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.\n\nSee the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).", "modelVersionGroupId": null, - "contextLength": 8191, + "contextLength": 8192, "inputModalities": [ "text" ], @@ -54131,67 +57289,63 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, + "group": "Gemini", + "instructType": "gemma", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "", + "", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-4", + "permaslug": "google/gemma-2-9b-it", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/gpt-4", - "modelVariantPermaslug": "openai/gpt-4", - "providerName": "OpenAI", + "modelVariantSlug": "google/gemma-2-9b-it:free", + "modelVariantPermaslug": "google/gemma-2-9b-it:free", + "providerName": "Chutes", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "training": true, + "retainsPrompts": true + } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", + "moderationRequired": false, + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "OpenAI", - "providerModelId": "gpt-4", - "providerGroup": "OpenAI", - "quantization": "unknown", - "variant": "standard", - "isFree": false, + "providerDisplayName": "Chutes", + "providerModelId": "unsloth/gemma-2-9b-it", + "providerGroup": "Chutes", + "quantization": null, + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": 8192, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -54199,30 +57353,27 @@ export default { "frequency_penalty", "presence_penalty", "seed", - "logit_bias", + "top_k", + "min_p", + "repetition_penalty", "logprobs", - "top_logprobs", - "response_format" + "logit_bias", + "top_logprobs" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.00003", - "completion": "0.00006", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -54233,7 +57384,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -54246,24 +57397,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 191, - "newest": 315, - "throughputHighToLow": 263, - "latencyLowToHigh": 114, - "pricingLowToHigh": 312, - "pricingHighToLow": 5 + "topWeekly": 101, + "newest": 240, + "throughputHighToLow": 40, + "latencyLowToHigh": 6, + "pricingLowToHigh": 69, + "pricingHighToLow": 239 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [ { - "name": "OpenAI", - "slug": "openAi", - "quantization": "unknown", - "context": 8191, - "maxCompletionTokens": 4096, - "providerModelId": "gpt-4", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 8192, + "maxCompletionTokens": null, + "providerModelId": "google/gemma-2-9b-it", "pricing": { - "prompt": "0.00003", - "completion": "0.00006", + "prompt": "0.00000002", + "completion": "0.00000006", "image": "0", "request": "0", "webSearch": "0", @@ -54271,8 +57424,6 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -54280,24 +57431,27 @@ export default { "frequency_penalty", "presence_penalty", "seed", + "top_k", "logit_bias", "logprobs", - "top_logprobs", - "response_format" + "top_logprobs" ], - "inputCost": 30, - "outputCost": 60 + "inputCost": 0.02, + "outputCost": 0.06, + "throughput": 142.233, + "latency": 181 }, { - "name": "Azure", - "slug": "azure", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", "quantization": "unknown", - "context": 8191, - "maxCompletionTokens": 4096, - "providerModelId": "gpt-4", + "context": 8192, + "maxCompletionTokens": null, + "providerModelId": "google/gemma-2-9b-it", "pricing": { - "prompt": "0.00003", - "completion": "0.00006", + "prompt": "0.00000008", + "completion": "0.00000008", "image": "0", "request": "0", "webSearch": "0", @@ -54305,8 +57459,6 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -54314,28 +57466,100 @@ export default { "frequency_penalty", "presence_penalty", "seed", - "logit_bias", - "logprobs", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" + ], + "inputCost": 0.08, + "outputCost": 0.08, + "throughput": 32.0005, + "latency": 661 + }, + { + "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", + "slug": "groq", + "quantization": null, + "context": 8192, + "maxCompletionTokens": 8192, + "providerModelId": "gemma2-9b-it", + "pricing": { + "prompt": "0.0000002", + "completion": "0.0000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", "top_logprobs", + "logprobs", + "logit_bias", + "seed" + ], + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 849.364, + "latency": 330 + }, + { + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": null, + "context": 8192, + "maxCompletionTokens": 8192, + "providerModelId": "google/gemma-2-9b-it", + "pricing": { + "prompt": "0.0000003", + "completion": "0.0000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "min_p", "response_format" ], - "inputCost": 30, - "outputCost": 60 + "inputCost": 0.3, + "outputCost": 0.3, + "throughput": 116.9945, + "latency": 345 } ] }, { - "slug": "mistralai/mistral-small", + "slug": "cohere/command-r-plus", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-01-10T00:00:00+00:00", + "createdAt": "2024-04-04T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Mistral Small", - "shortName": "Mistral Small", - "author": "mistralai", - "description": "With 22 billion parameters, Mistral Small v24.09 offers a convenient mid-point between (Mistral NeMo 12B)[/mistralai/mistral-nemo] and (Mistral Large 2)[/mistralai/mistral-large], providing a cost-effective solution that can be deployed across various platforms and environments. It has better reasoning, exhibits more capabilities, can produce and reason about code, and is multiligual, supporting English, French, German, Italian, and Spanish.", + "name": "Cohere: Command R+", + "shortName": "Command R+", + "author": "cohere", + "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -54343,32 +57567,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "Cohere", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-small", + "permaslug": "cohere/command-r-plus", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "e9fb7fa9-5720-4fd1-ac88-06fba8fbca50", - "name": "Mistral | mistralai/mistral-small", - "contextLength": 32768, + "id": "7bfb1bda-c33b-4f0f-b8c5-bb3216a31b6b", + "name": "Cohere | cohere/command-r-plus", + "contextLength": 128000, "model": { - "slug": "mistralai/mistral-small", + "slug": "cohere/command-r-plus", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-01-10T00:00:00+00:00", + "createdAt": "2024-04-04T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Mistral Small", - "shortName": "Mistral Small", - "author": "mistralai", - "description": "With 22 billion parameters, Mistral Small v24.09 offers a convenient mid-point between (Mistral NeMo 12B)[/mistralai/mistral-nemo] and (Mistral Large 2)[/mistralai/mistral-large], providing a cost-effective solution that can be deployed across various platforms and environments. It has better reasoning, exhibits more capabilities, can produce and reason about code, and is multiligual, supporting English, French, German, Italian, and Spanish.", + "name": "Cohere: Command R+", + "shortName": "Command R+", + "author": "cohere", + "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", "modelVersionGroupId": null, - "contextLength": 32000, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -54376,79 +57600,79 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "Cohere", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-small", + "permaslug": "cohere/command-r-plus", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "mistralai/mistral-small", - "modelVariantPermaslug": "mistralai/mistral-small", - "providerName": "Mistral", + "modelVariantSlug": "cohere/command-r-plus", + "modelVariantPermaslug": "cohere/command-r-plus", + "providerName": "Cohere", "providerInfo": { - "name": "Mistral", - "displayName": "Mistral", - "slug": "mistral", + "name": "Cohere", + "displayName": "Cohere", + "slug": "cohere", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "privacyPolicyUrl": "https://cohere.com/privacy", + "termsOfServiceUrl": "https://cohere.com/terms-of-use", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 } }, - "headquarters": "FR", + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": false, - "isAbortable": false, + "isAbortable": true, "moderationRequired": false, - "group": "Mistral", + "group": "Cohere", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.cohere.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Mistral.png" + "url": "/images/icons/Cohere.png" } }, - "providerDisplayName": "Mistral", - "providerModelId": "mistral-small-2409", - "providerGroup": "Mistral", + "providerDisplayName": "Cohere", + "providerModelId": "command-r-plus", + "providerGroup": "Cohere", "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", + "top_k", + "seed", "response_format", - "structured_outputs", - "seed" + "structured_outputs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "privacyPolicyUrl": "https://cohere.com/privacy", + "termsOfServiceUrl": "https://cohere.com/terms-of-use", "paidModels": { "training": false, "retainsPrompts": true, @@ -54459,8 +57683,8 @@ export default { "retentionDays": 30 }, "pricing": { - "prompt": "0.0000002", - "completion": "0.0000006", + "prompt": "0.000003", + "completion": "0.000015", "image": "0", "request": "0", "webSearch": "0", @@ -54485,24 +57709,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 192, - "newest": 288, - "throughputHighToLow": 64, - "latencyLowToHigh": 38, - "pricingLowToHigh": 155, - "pricingHighToLow": 164 + "topWeekly": 195, + "newest": 272, + "throughputHighToLow": 156, + "latencyLowToHigh": 48, + "pricingLowToHigh": 276, + "pricingHighToLow": 48 }, + "authorIcon": "https://openrouter.ai/images/icons/Cohere.png", "providers": [ { - "name": "Mistral", - "slug": "mistral", + "name": "Cohere", + "icon": "https://openrouter.ai/images/icons/Cohere.png", + "slug": "cohere", "quantization": "unknown", - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "mistral-small-2409", + "context": 128000, + "maxCompletionTokens": 4000, + "providerModelId": "command-r-plus", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000006", + "prompt": "0.000003", + "completion": "0.000015", "image": "0", "request": "0", "webSearch": "0", @@ -54511,203 +57737,35 @@ export default { }, "supportedParameters": [ "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", + "top_k", + "seed", "response_format", - "structured_outputs", - "seed" + "structured_outputs" ], - "inputCost": 0.2, - "outputCost": 0.6 + "inputCost": 3, + "outputCost": 15, + "throughput": 56.4855, + "latency": 364 } ] }, { - "slug": "moonshotai/kimi-vl-a3b-thinking", - "hfSlug": "moonshotai/Kimi-VL-A3B-Thinking", - "updatedAt": "2025-04-10T17:10:17.814287+00:00", - "createdAt": "2025-04-10T17:07:21.175402+00:00", - "hfUpdatedAt": null, - "name": "Moonshot AI: Kimi VL A3B Thinking (free)", - "shortName": "Kimi VL A3B Thinking (free)", - "author": "moonshotai", - "description": "Kimi-VL is a lightweight Mixture-of-Experts vision-language model that activates only 2.8B parameters per step while delivering strong performance on multimodal reasoning and long-context tasks. The Kimi-VL-A3B-Thinking variant, fine-tuned with chain-of-thought and reinforcement learning, excels in math and visual reasoning benchmarks like MathVision, MMMU, and MathVista, rivaling much larger models such as Qwen2.5-VL-7B and Gemma-3-12B. It supports 128K context and high-resolution input via its MoonViT encoder.", - "modelVersionGroupId": null, - "contextLength": 131072, - "inputModalities": [ - "image", - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "moonshotai/kimi-vl-a3b-thinking", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "ba332390-9210-4e73-a2c4-0fab667c7fb1", - "name": "Chutes | moonshotai/kimi-vl-a3b-thinking:free", - "contextLength": 131072, - "model": { - "slug": "moonshotai/kimi-vl-a3b-thinking", - "hfSlug": "moonshotai/Kimi-VL-A3B-Thinking", - "updatedAt": "2025-04-10T17:10:17.814287+00:00", - "createdAt": "2025-04-10T17:07:21.175402+00:00", - "hfUpdatedAt": null, - "name": "Moonshot AI: Kimi VL A3B Thinking", - "shortName": "Kimi VL A3B Thinking", - "author": "moonshotai", - "description": "Kimi-VL is a lightweight Mixture-of-Experts vision-language model that activates only 2.8B parameters per step while delivering strong performance on multimodal reasoning and long-context tasks. The Kimi-VL-A3B-Thinking variant, fine-tuned with chain-of-thought and reinforcement learning, excels in math and visual reasoning benchmarks like MathVision, MMMU, and MathVista, rivaling much larger models such as Qwen2.5-VL-7B and Gemma-3-12B. It supports 128K context and high-resolution input via its MoonViT encoder.", - "modelVersionGroupId": null, - "contextLength": 131072, - "inputModalities": [ - "image", - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "moonshotai/kimi-vl-a3b-thinking", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "moonshotai/kimi-vl-a3b-thinking:free", - "modelVariantPermaslug": "moonshotai/kimi-vl-a3b-thinking:free", - "providerName": "Chutes", - "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", - "paidModels": { - "training": true, - "retainsPrompts": true - } - }, - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "Chutes", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" - } - }, - "providerDisplayName": "Chutes", - "providerModelId": "moonshotai/Kimi-VL-A3B-Thinking", - "providerGroup": "Chutes", - "quantization": null, - "variant": "free", - "isFree": true, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": 8, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", - "paidModels": { - "training": true, - "retainsPrompts": true - }, - "training": true, - "retainsPrompts": true - }, - "pricing": { - "prompt": "0", - "completion": "0", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": true, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 193, - "newest": 55, - "throughputHighToLow": 216, - "latencyLowToHigh": 294, - "pricingLowToHigh": 23, - "pricingHighToLow": 271 - }, - "providers": [] - }, - { - "slug": "mistralai/mistral-7b-instruct-v0.2", - "hfSlug": "mistralai/Mistral-7B-Instruct-v0.2", + "slug": "mistralai/mistral-small", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-12-28T00:00:00+00:00", + "createdAt": "2024-01-10T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral 7B Instruct v0.2", - "shortName": "Mistral 7B Instruct v0.2", + "name": "Mistral Small", + "shortName": "Mistral Small", "author": "mistralai", - "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\nAn improved version of [Mistral 7B Instruct](/modelsmistralai/mistral-7b-instruct-v0.1), with the following changes:\n\n- 32k context window (vs 8k context in v0.1)\n- Rope-theta = 1e6\n- No Sliding-Window Attention", - "modelVersionGroupId": "1d07cc56-c54d-4587-b785-5093496397a4", + "description": "With 22 billion parameters, Mistral Small v24.09 offers a convenient mid-point between (Mistral NeMo 12B)[/mistralai/mistral-nemo] and (Mistral Large 2)[/mistralai/mistral-large], providing a cost-effective solution that can be deployed across various platforms and environments. It has better reasoning, exhibits more capabilities, can produce and reason about code, and is multiligual, supporting English, French, German, Italian, and Spanish.", + "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ "text" @@ -54717,34 +57775,31 @@ export default { ], "hasTextOutput": true, "group": "Mistral", - "instructType": "mistral", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "[INST]", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-7b-instruct-v0.2", + "permaslug": "mistralai/mistral-small", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "ae921682-8673-4287-b3bd-459304097932", - "name": "Together | mistralai/mistral-7b-instruct-v0.2", + "id": "e9fb7fa9-5720-4fd1-ac88-06fba8fbca50", + "name": "Mistral | mistralai/mistral-small", "contextLength": 32768, "model": { - "slug": "mistralai/mistral-7b-instruct-v0.2", - "hfSlug": "mistralai/Mistral-7B-Instruct-v0.2", + "slug": "mistralai/mistral-small", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-12-28T00:00:00+00:00", + "createdAt": "2024-01-10T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral 7B Instruct v0.2", - "shortName": "Mistral 7B Instruct v0.2", + "name": "Mistral Small", + "shortName": "Mistral Small", "author": "mistralai", - "description": "A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.\n\nAn improved version of [Mistral 7B Instruct](/modelsmistralai/mistral-7b-instruct-v0.1), with the following changes:\n\n- 32k context window (vs 8k context in v0.1)\n- Rope-theta = 1e6\n- No Sliding-Window Attention", - "modelVersionGroupId": "1d07cc56-c54d-4587-b785-5093496397a4", - "contextLength": 32768, + "description": "With 22 billion parameters, Mistral Small v24.09 offers a convenient mid-point between (Mistral NeMo 12B)[/mistralai/mistral-nemo] and (Mistral Large 2)[/mistralai/mistral-large], providing a cost-effective solution that can be deployed across various platforms and environments. It has better reasoning, exhibits more capabilities, can produce and reason about code, and is multiligual, supporting English, French, German, Italian, and Spanish.", + "modelVersionGroupId": null, + "contextLength": 32000, "inputModalities": [ "text" ], @@ -54753,41 +57808,39 @@ export default { ], "hasTextOutput": true, "group": "Mistral", - "instructType": "mistral", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "[INST]", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-7b-instruct-v0.2", + "permaslug": "mistralai/mistral-small", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "mistralai/mistral-7b-instruct-v0.2", - "modelVariantPermaslug": "mistralai/mistral-7b-instruct-v0.2", - "providerName": "Together", + "modelVariantSlug": "mistralai/mistral-small", + "modelVariantPermaslug": "mistralai/mistral-small", + "providerName": "Mistral", "providerInfo": { - "name": "Together", - "displayName": "Together", - "slug": "together", + "name": "Mistral", + "displayName": "Mistral", + "slug": "mistral", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 } }, - "headquarters": "US", + "headquarters": "FR", "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, + "hasCompletions": false, + "isAbortable": false, "moderationRequired": false, - "group": "Together", + "group": "Mistral", "editors": [], "owners": [], "isMultipartSupported": true, @@ -54795,48 +57848,50 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" + "url": "/images/icons/Mistral.png" } }, - "providerDisplayName": "Together", - "providerModelId": "mistralai/Mistral-7B-Instruct-v0.2", - "providerGroup": "Together", + "providerDisplayName": "Mistral", + "providerModelId": "mistral-small-2409", + "providerGroup": "Mistral", "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "response_format", + "structured_outputs", + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { "prompt": "0.0000002", - "completion": "0.0000002", + "completion": "0.0000006", "image": "0", "request": "0", "webSearch": "0", @@ -54847,37 +57902,40 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 194, - "newest": 291, - "throughputHighToLow": 15, - "latencyLowToHigh": 4, - "pricingLowToHigh": 152, - "pricingHighToLow": 170 + "topWeekly": 196, + "newest": 289, + "throughputHighToLow": 68, + "latencyLowToHigh": 36, + "pricingLowToHigh": 155, + "pricingHighToLow": 164 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { - "name": "Together", - "slug": "together", + "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", + "slug": "mistral", "quantization": "unknown", "context": 32768, "maxCompletionTokens": null, - "providerModelId": "mistralai/Mistral-7B-Instruct-v0.2", + "providerModelId": "mistral-small-2409", "pricing": { "prompt": "0.0000002", - "completion": "0.0000002", + "completion": "0.0000006", "image": "0", "request": "0", "webSearch": "0", @@ -54885,164 +57943,166 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "response_format", + "structured_outputs", + "seed" ], "inputCost": 0.2, - "outputCost": 0.2 + "outputCost": 0.6, + "throughput": 108.867, + "latency": 343 } ] }, { - "slug": "anthropic/claude-3.5-haiku", - "hfSlug": null, + "slug": "infermatic/mn-inferor-12b", + "hfSlug": "Infermatic/MN-12B-Inferor-v0.0", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-04T00:00:00+00:00", + "createdAt": "2024-11-13T02:20:28.884947+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3.5 Haiku (self-moderated)", - "shortName": "Claude 3.5 Haiku (self-moderated)", - "author": "anthropic", - "description": "Claude 3.5 Haiku features offers enhanced capabilities in speed, coding accuracy, and tool use. Engineered to excel in real-time applications, it delivers quick response times that are essential for dynamic tasks such as chat interactions and immediate coding suggestions.\n\nThis makes it highly suitable for environments that demand both speed and precision, such as software development, customer service bots, and data management systems.\n\nThis model is currently pointing to [Claude 3.5 Haiku (2024-10-22)](/anthropic/claude-3-5-haiku-20241022).", - "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", - "contextLength": 200000, + "name": "Infermatic: Mistral Nemo Inferor 12B", + "shortName": "Mistral Nemo Inferor 12B", + "author": "infermatic", + "description": "Inferor 12B is a merge of top roleplay models, expert on immersive narratives and storytelling.\n\nThis model was merged using the [Model Stock](https://arxiv.org/abs/2403.19522) merge method using [anthracite-org/magnum-v4-12b](https://openrouter.ai/anthracite-org/magnum-v4-72b) as a base.\n", + "modelVersionGroupId": null, + "contextLength": 16384, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", - "instructType": null, + "group": "Mistral", + "instructType": "mistral", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "[INST]", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-5-haiku", + "permaslug": "infermatic/mn-inferor-12b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "ced809d4-9715-4634-8b4d-ea0e09d6bb85", - "name": "Anthropic | anthropic/claude-3-5-haiku:beta", - "contextLength": 200000, + "id": "cc0c2222-fb27-4c56-8ce6-1f6a6be16a93", + "name": "Featherless | infermatic/mn-inferor-12b", + "contextLength": 16384, "model": { - "slug": "anthropic/claude-3.5-haiku", - "hfSlug": null, + "slug": "infermatic/mn-inferor-12b", + "hfSlug": "Infermatic/MN-12B-Inferor-v0.0", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-04T00:00:00+00:00", + "createdAt": "2024-11-13T02:20:28.884947+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3.5 Haiku", - "shortName": "Claude 3.5 Haiku", - "author": "anthropic", - "description": "Claude 3.5 Haiku features offers enhanced capabilities in speed, coding accuracy, and tool use. Engineered to excel in real-time applications, it delivers quick response times that are essential for dynamic tasks such as chat interactions and immediate coding suggestions.\n\nThis makes it highly suitable for environments that demand both speed and precision, such as software development, customer service bots, and data management systems.\n\nThis model is currently pointing to [Claude 3.5 Haiku (2024-10-22)](/anthropic/claude-3-5-haiku-20241022).", - "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", - "contextLength": 200000, + "name": "Infermatic: Mistral Nemo Inferor 12B", + "shortName": "Mistral Nemo Inferor 12B", + "author": "infermatic", + "description": "Inferor 12B is a merge of top roleplay models, expert on immersive narratives and storytelling.\n\nThis model was merged using the [Model Stock](https://arxiv.org/abs/2403.19522) merge method using [anthracite-org/magnum-v4-12b](https://openrouter.ai/anthracite-org/magnum-v4-72b) as a base.\n", + "modelVersionGroupId": null, + "contextLength": 32000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", - "instructType": null, + "group": "Mistral", + "instructType": "mistral", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "[INST]", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-5-haiku", + "permaslug": "infermatic/mn-inferor-12b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "anthropic/claude-3.5-haiku:beta", - "modelVariantPermaslug": "anthropic/claude-3-5-haiku:beta", - "providerName": "Anthropic", + "modelVariantSlug": "infermatic/mn-inferor-12b", + "modelVariantPermaslug": "infermatic/mn-inferor-12b", + "providerName": "Featherless", "providerInfo": { - "name": "Anthropic", - "displayName": "Anthropic", - "slug": "anthropic", + "name": "Featherless", + "displayName": "Featherless", + "slug": "featherless", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "retainsPrompts": false + } }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": true, - "moderationRequired": true, - "group": "Anthropic", + "isAbortable": false, + "moderationRequired": false, + "group": "Featherless", "editors": [], "owners": [], - "isMultipartSupported": true, - "statusPageUrl": "https://status.anthropic.com/", + "isMultipartSupported": false, + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Anthropic.svg" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" } }, - "providerDisplayName": "Anthropic", - "providerModelId": "claude-3-5-haiku-20241022", - "providerGroup": "Anthropic", - "quantization": "unknown", - "variant": "beta", + "providerDisplayName": "Featherless", + "providerModelId": "Infermatic/MN-12B-Inferor-v0.0", + "providerGroup": "Featherless", + "quantization": null, + "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 8192, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", "top_k", - "stop" + "min_p", + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, - "requiresUserIds": true, "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, "pricing": { "prompt": "0.0000008", - "completion": "0.000004", + "completion": "0.0000012", "image": "0", "request": "0", - "inputCacheRead": "0.00000008", - "inputCacheWrite": "0.000001", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -55051,9 +58111,9 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, - "supportsMultipart": true, + "supportsMultipart": false, "limitRpm": null, "limitRpd": null, "hasCompletions": true, @@ -55064,146 +58124,63 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 48, - "newest": 179, - "throughputHighToLow": 129, - "latencyLowToHigh": 164, - "pricingLowToHigh": 221, - "pricingHighToLow": 93 + "topWeekly": 197, + "newest": 171, + "throughputHighToLow": 282, + "latencyLowToHigh": 266, + "pricingLowToHigh": 206, + "pricingHighToLow": 110 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://infermatic.ai/\\u0026size=256", "providers": [ { - "name": "Anthropic", - "slug": "anthropic", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 8192, - "providerModelId": "claude-3-5-haiku-20241022", - "pricing": { - "prompt": "0.0000008", - "completion": "0.000004", - "image": "0", - "request": "0", - "inputCacheRead": "0.00000008", - "inputCacheWrite": "0.000001", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "top_k", - "stop" - ], - "inputCost": 0.8, - "outputCost": 4 - }, - { - "name": "Google Vertex", - "slug": "vertex", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 8192, - "providerModelId": "claude-3-5-haiku@20241022", - "pricing": { - "prompt": "0.0000008", - "completion": "0.000004", - "image": "0", - "request": "0", - "inputCacheRead": "0.00000008", - "inputCacheWrite": "0.000001", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "top_k", - "stop" - ], - "inputCost": 0.8, - "outputCost": 4 - }, - { - "name": "Amazon Bedrock", - "slug": "amazonBedrock", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 8192, - "providerModelId": "us.anthropic.claude-3-5-haiku-20241022-v1:0", - "pricing": { - "prompt": "0.0000008", - "completion": "0.000004", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "top_k", - "stop" - ], - "inputCost": 0.8, - "outputCost": 4 - }, - { - "name": "Amazon Bedrock (US-WEST)", - "slug": "amazonBedrock (usWest)", + "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", + "slug": "featherless", "quantization": null, - "context": 200000, - "maxCompletionTokens": 8192, - "providerModelId": "us.anthropic.claude-3-5-haiku-20241022-v1:0", + "context": 16384, + "maxCompletionTokens": 4096, + "providerModelId": "Infermatic/MN-12B-Inferor-v0.0", "pricing": { "prompt": "0.0000008", - "completion": "0.000004", + "completion": "0.0000012", "image": "0", "request": "0", - "inputCacheRead": "0.00000008", - "inputCacheWrite": "0.000001", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", "top_k", - "stop" + "min_p", + "seed" ], "inputCost": 0.8, - "outputCost": 4 + "outputCost": 1.2, + "throughput": 18.723, + "latency": 1936.5 } ] }, { - "slug": "cohere/command-r-plus", - "hfSlug": null, + "slug": "sao10k/l3-euryale-70b", + "hfSlug": "Sao10K/L3-70B-Euryale-v2.1", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-04-04T00:00:00+00:00", + "createdAt": "2024-06-18T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Cohere: Command R+", - "shortName": "Command R+", - "author": "cohere", - "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", + "name": "Sao10k: Llama 3 Euryale 70B v2.1", + "shortName": "Llama 3 Euryale 70B v2.1", + "author": "sao10k", + "description": "Euryale 70B v2.1 is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k).\n\n- Better prompt adherence.\n- Better anatomy / spatial awareness.\n- Adapts much better to unique and custom formatting / reply formats.\n- Very creative, lots of unique swipes.\n- Is not restrictive during roleplays.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 8192, "inputModalities": [ "text" ], @@ -55211,32 +58188,35 @@ export default { "text" ], "hasTextOutput": true, - "group": "Cohere", - "instructType": null, + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cohere/command-r-plus", + "permaslug": "sao10k/l3-euryale-70b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "7bfb1bda-c33b-4f0f-b8c5-bb3216a31b6b", - "name": "Cohere | cohere/command-r-plus", - "contextLength": 128000, + "id": "2ab67dc9-421b-408e-b61a-b4b86ea1df70", + "name": "Novita | sao10k/l3-euryale-70b", + "contextLength": 8192, "model": { - "slug": "cohere/command-r-plus", - "hfSlug": null, + "slug": "sao10k/l3-euryale-70b", + "hfSlug": "Sao10K/L3-70B-Euryale-v2.1", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-04-04T00:00:00+00:00", + "createdAt": "2024-06-18T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Cohere: Command R+", - "shortName": "Command R+", - "author": "cohere", - "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", + "name": "Sao10k: Llama 3 Euryale 70B v2.1", + "shortName": "Llama 3 Euryale 70B v2.1", + "author": "sao10k", + "description": "Euryale 70B v2.1 is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k).\n\n- Better prompt adherence.\n- Better anatomy / spatial awareness.\n- Adapts much better to unique and custom formatting / reply formats.\n- Very creative, lots of unique swipes.\n- Is not restrictive during roleplays.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 8192, "inputModalities": [ "text" ], @@ -55244,91 +58224,89 @@ export default { "text" ], "hasTextOutput": true, - "group": "Cohere", - "instructType": null, + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cohere/command-r-plus", + "permaslug": "sao10k/l3-euryale-70b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "cohere/command-r-plus", - "modelVariantPermaslug": "cohere/command-r-plus", - "providerName": "Cohere", + "modelVariantSlug": "sao10k/l3-euryale-70b", + "modelVariantPermaslug": "sao10k/l3-euryale-70b", + "providerName": "Novita", "providerInfo": { - "name": "Cohere", - "displayName": "Cohere", - "slug": "cohere", + "name": "Novita", + "displayName": "NovitaAI", + "slug": "novita", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://cohere.com/privacy", - "termsOfServiceUrl": "https://cohere.com/terms-of-use", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, + "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Cohere", + "group": "Novita", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.cohere.ai/", + "statusPageUrl": "https://status.novita.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Cohere.png" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "invertRequired": true } }, - "providerDisplayName": "Cohere", - "providerModelId": "command-r-plus", - "providerGroup": "Cohere", - "quantization": "unknown", + "providerDisplayName": "NovitaAI", + "providerModelId": "sao10k/l3-70b-euryale-v2.1", + "providerGroup": "Novita", + "quantization": "fp8", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4000, + "maxCompletionTokens": 8192, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "top_k", "seed", - "response_format", - "structured_outputs" + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://cohere.com/privacy", - "termsOfServiceUrl": "https://cohere.com/terms-of-use", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, "pricing": { - "prompt": "0.000003", - "completion": "0.000015", + "prompt": "0.00000148", + "completion": "0.00000148", "image": "0", "request": "0", "webSearch": "0", @@ -55339,38 +58317,39 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 196, - "newest": 272, - "throughputHighToLow": 138, - "latencyLowToHigh": 36, - "pricingLowToHigh": 276, - "pricingHighToLow": 48 + "topWeekly": 198, + "newest": 246, + "throughputHighToLow": 228, + "latencyLowToHigh": 148, + "pricingLowToHigh": 235, + "pricingHighToLow": 83 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { - "name": "Cohere", - "slug": "cohere", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 4000, - "providerModelId": "command-r-plus", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": "fp8", + "context": 8192, + "maxCompletionTokens": 8192, + "providerModelId": "sao10k/l3-70b-euryale-v2.1", "pricing": { - "prompt": "0.000003", - "completion": "0.000015", + "prompt": "0.00000148", + "completion": "0.00000148", "image": "0", "request": "0", "webSearch": "0", @@ -55378,35 +58357,37 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "top_k", "seed", - "response_format", - "structured_outputs" + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" ], - "inputCost": 3, - "outputCost": 15 + "inputCost": 1.48, + "outputCost": 1.48, + "throughput": 34.967, + "latency": 885 } ] }, { - "slug": "deepseek/deepseek-coder", - "hfSlug": "deepseek-ai/DeepSeek-Coder-V2-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-05-14T00:00:00+00:00", + "slug": "inception/mercury-coder-small-beta", + "hfSlug": "", + "updatedAt": "2025-04-30T17:36:42.979824+00:00", + "createdAt": "2025-04-30T17:24:40+00:00", "hfUpdatedAt": null, - "name": "DeepSeek-Coder-V2", - "shortName": "DeepSeek-Coder-V2", - "author": "deepseek", - "description": "DeepSeek-Coder-V2, an open-source Mixture-of-Experts (MoE) code language model. It is further pre-trained from an intermediate checkpoint of DeepSeek-V2 with additional 6 trillion tokens.\n\nThe original V1 model was trained from scratch on 2T tokens, with a composition of 87% code and 13% natural language in both English and Chinese. It was pre-trained on project-level code corpus by employing a extra fill-in-the-blank task.", + "name": "Inception: Mercury Coder Small Beta", + "shortName": "Mercury Coder Small Beta", + "author": "inception", + "description": "Mercury Coder Small is the first diffusion large language model (dLLM). Applying a breakthrough discrete diffusion approach, the model runs 5-10x faster than even speed optimized models like Claude 3.5 Haiku and GPT-4o Mini while matching their performance. Mercury Coder Small's speed means that developers can stay in the flow while coding, enjoying rapid chat-based iteration and responsive code completion suggestions. On Copilot Arena, Mercury Coder ranks 1st in speed and ties for 2nd in quality. Read more in the [blog post here](https://www.inceptionlabs.ai/introducing-mercury).", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 32000, "inputModalities": [ "text" ], @@ -55421,25 +58402,25 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "deepseek/deepseek-coder", + "permaslug": "inception/mercury-coder-small-beta", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "91fc52a2-a00f-4b45-b3a6-0d74a352bc07", - "name": "Nebius | deepseek/deepseek-coder", - "contextLength": 128000, + "id": "eb51fb37-11b0-42d1-924a-c25fa2375569", + "name": "Inception | inception/mercury-coder-small-beta", + "contextLength": 32000, "model": { - "slug": "deepseek/deepseek-coder", - "hfSlug": "deepseek-ai/DeepSeek-Coder-V2-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-05-14T00:00:00+00:00", + "slug": "inception/mercury-coder-small-beta", + "hfSlug": "", + "updatedAt": "2025-04-30T17:36:42.979824+00:00", + "createdAt": "2025-04-30T17:24:40+00:00", "hfUpdatedAt": null, - "name": "DeepSeek-Coder-V2", - "shortName": "DeepSeek-Coder-V2", - "author": "deepseek", - "description": "DeepSeek-Coder-V2, an open-source Mixture-of-Experts (MoE) code language model. It is further pre-trained from an intermediate checkpoint of DeepSeek-V2 with additional 6 trillion tokens.\n\nThe original V1 model was trained from scratch on 2T tokens, with a composition of 87% code and 13% natural language in both English and Chinese. It was pre-trained on project-level code corpus by employing a extra fill-in-the-blank task.", + "name": "Inception: Mercury Coder Small Beta", + "shortName": "Mercury Coder Small Beta", + "author": "inception", + "description": "Mercury Coder Small is the first diffusion large language model (dLLM). Applying a breakthrough discrete diffusion approach, the model runs 5-10x faster than even speed optimized models like Claude 3.5 Haiku and GPT-4o Mini while matching their performance. Mercury Coder Small's speed means that developers can stay in the flow while coding, enjoying rapid chat-based iteration and responsive code completion suggestions. On Copilot Arena, Mercury Coder ranks 1st in speed and ties for 2nd in quality. Read more in the [blog post here](https://www.inceptionlabs.ai/introducing-mercury).", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 32000, "inputModalities": [ "text" ], @@ -55454,32 +58435,31 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "deepseek/deepseek-coder", + "permaslug": "inception/mercury-coder-small-beta", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "deepseek/deepseek-coder", - "modelVariantPermaslug": "deepseek/deepseek-coder", - "providerName": "Nebius", + "modelVariantSlug": "inception/mercury-coder-small-beta", + "modelVariantPermaslug": "inception/mercury-coder-small-beta", + "providerName": "Inception", "providerInfo": { - "name": "Nebius", - "displayName": "Nebius AI Studio", - "slug": "nebius", + "name": "Inception", + "displayName": "Inception", + "slug": "inception", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "termsOfServiceUrl": "https://www.inceptionlabs.ai/terms", + "privacyPolicyUrl": "https://www.inceptionlabs.ai/terms#privacy-policy", "paidModels": { "training": false, "retainsPrompts": false } }, - "headquarters": "DE", "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": false, + "hasCompletions": false, + "isAbortable": true, "moderationRequired": false, - "group": "Nebius", + "group": "Inception", "editors": [], "owners": [], "isMultipartSupported": true, @@ -55487,39 +58467,32 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.inceptionlabs.ai/&size=256", "invertRequired": true } }, - "providerDisplayName": "Nebius AI Studio", - "providerModelId": "deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct", - "providerGroup": "Nebius", - "quantization": "fp8", + "providerDisplayName": "Inception", + "providerModelId": "mercury-coder-small", + "providerGroup": "Inception", + "quantization": null, "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", - "temperature", - "top_p", - "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" + "stop" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "termsOfServiceUrl": "https://www.inceptionlabs.ai/terms", + "privacyPolicyUrl": "https://www.inceptionlabs.ai/terms#privacy-policy", "paidModels": { "training": false, "retainsPrompts": false @@ -55528,8 +58501,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000004", - "completion": "0.00000012", + "prompt": "0.00000025", + "completion": "0.000001", "image": "0", "request": "0", "webSearch": "0", @@ -55545,7 +58518,7 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { "supportedParameters": {}, @@ -55554,24 +58527,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 197, - "newest": 256, - "throughputHighToLow": 104, - "latencyLowToHigh": 16, - "pricingLowToHigh": 91, - "pricingHighToLow": 227 + "topWeekly": 199, + "newest": 14, + "throughputHighToLow": 0, + "latencyLowToHigh": 65, + "pricingLowToHigh": 166, + "pricingHighToLow": 151 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://www.inceptionlabs.ai/\\u0026size=256", "providers": [ { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 128000, + "name": "Inception", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.inceptionlabs.ai/&size=256", + "slug": "inception", + "quantization": null, + "context": 32000, "maxCompletionTokens": null, - "providerModelId": "deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct", + "providerModelId": "mercury-coder-small", "pricing": { - "prompt": "0.00000004", - "completion": "0.00000012", + "prompt": "0.00000025", + "completion": "0.000001", "image": "0", "request": "0", "webSearch": "0", @@ -55580,34 +58555,29 @@ export default { }, "supportedParameters": [ "max_tokens", - "temperature", - "top_p", - "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" + "stop" ], - "inputCost": 0.04, - "outputCost": 0.12 + "inputCost": 0.25, + "outputCost": 1, + "throughput": 699.068, + "latency": 530 } ] }, { - "slug": "openai/gpt-4-1106-preview", - "hfSlug": null, + "slug": "cognitivecomputations/dolphin3.0-mistral-24b", + "hfSlug": "cognitivecomputations/Dolphin3.0-Mistral-24B", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-11-06T00:00:00+00:00", + "createdAt": "2025-02-13T15:53:39.941479+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4 Turbo (older v1106)", - "shortName": "GPT-4 Turbo (older v1106)", - "author": "openai", - "description": "The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.\n\nTraining data: up to April 2023.", + "name": "Dolphin3.0 Mistral 24B (free)", + "shortName": "Dolphin3.0 Mistral 24B (free)", + "author": "cognitivecomputations", + "description": "Dolphin 3.0 is the next generation of the Dolphin series of instruct-tuned models. Designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.\n\nDolphin aims to be a general purpose instruct model, similar to the models behind ChatGPT, Claude, Gemini. \n\nPart of the [Dolphin 3.0 Collection](https://huggingface.co/collections/cognitivecomputations/dolphin-30-677ab47f73d7ff66743979a3) Curated and trained by [Eric Hartford](https://huggingface.co/ehartford), [Ben Gitter](https://huggingface.co/bigstorm), [BlouseJury](https://huggingface.co/BlouseJury) and [Cognitive Computations](https://huggingface.co/cognitivecomputations)", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -55615,32 +58585,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-4-1106-preview", + "permaslug": "cognitivecomputations/dolphin3.0-mistral-24b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "8744de26-f64f-41cf-bd0e-950a83d1a923", - "name": "OpenAI | openai/gpt-4-1106-preview", - "contextLength": 128000, + "id": "4c7841c6-0d75-49bc-be70-478c2ae4f12b", + "name": "Chutes | cognitivecomputations/dolphin3.0-mistral-24b:free", + "contextLength": 32768, "model": { - "slug": "openai/gpt-4-1106-preview", - "hfSlug": null, + "slug": "cognitivecomputations/dolphin3.0-mistral-24b", + "hfSlug": "cognitivecomputations/Dolphin3.0-Mistral-24B", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-11-06T00:00:00+00:00", + "createdAt": "2025-02-13T15:53:39.941479+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4 Turbo (older v1106)", - "shortName": "GPT-4 Turbo (older v1106)", - "author": "openai", - "description": "The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.\n\nTraining data: up to April 2023.", + "name": "Dolphin3.0 Mistral 24B", + "shortName": "Dolphin3.0 Mistral 24B", + "author": "cognitivecomputations", + "description": "Dolphin 3.0 is the next generation of the Dolphin series of instruct-tuned models. Designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.\n\nDolphin aims to be a general purpose instruct model, similar to the models behind ChatGPT, Claude, Gemini. \n\nPart of the [Dolphin 3.0 Collection](https://huggingface.co/collections/cognitivecomputations/dolphin-30-677ab47f73d7ff66743979a3) Curated and trained by [Eric Hartford](https://huggingface.co/ehartford), [Ben Gitter](https://huggingface.co/bigstorm), [BlouseJury](https://huggingface.co/BlouseJury) and [Cognitive Computations](https://huggingface.co/cognitivecomputations)", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -55648,67 +58618,59 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-4-1106-preview", + "permaslug": "cognitivecomputations/dolphin3.0-mistral-24b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/gpt-4-1106-preview", - "modelVariantPermaslug": "openai/gpt-4-1106-preview", - "providerName": "OpenAI", + "modelVariantSlug": "cognitivecomputations/dolphin3.0-mistral-24b:free", + "modelVariantPermaslug": "cognitivecomputations/dolphin3.0-mistral-24b:free", + "providerName": "Chutes", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "training": true, + "retainsPrompts": true + } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", + "moderationRequired": false, + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "OpenAI", - "providerModelId": "gpt-4-1106-preview", - "providerGroup": "OpenAI", - "quantization": "unknown", - "variant": "standard", - "isFree": false, + "providerDisplayName": "Chutes", + "providerModelId": "cognitivecomputations/Dolphin3.0-Mistral-24B", + "providerGroup": "Chutes", + "quantization": null, + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -55716,31 +58678,27 @@ export default { "frequency_penalty", "presence_penalty", "seed", - "logit_bias", + "top_k", + "min_p", + "repetition_penalty", "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" + "logit_bias", + "top_logprobs" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.00001", - "completion": "0.00003", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -55751,7 +58709,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -55764,63 +58722,28 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 198, - "newest": 300, - "throughputHighToLow": 221, - "latencyLowToHigh": 125, - "pricingLowToHigh": 304, - "pricingHighToLow": 16 + "topWeekly": 200, + "newest": 114, + "throughputHighToLow": 111, + "latencyLowToHigh": 226, + "pricingLowToHigh": 47, + "pricingHighToLow": 295 }, - "providers": [ - { - "name": "OpenAI", - "slug": "openAi", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 4096, - "providerModelId": "gpt-4-1106-preview", - "pricing": { - "prompt": "0.00001", - "completion": "0.00003", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" - ], - "inputCost": 10, - "outputCost": 30 - } - ] + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [] }, { - "slug": "sao10k/l3-euryale-70b", - "hfSlug": "Sao10K/L3-70B-Euryale-v2.1", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-06-18T00:00:00+00:00", + "slug": "perplexity/sonar-reasoning", + "hfSlug": "", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-01-29T06:11:47.38089+00:00", "hfUpdatedAt": null, - "name": "Sao10k: Llama 3 Euryale 70B v2.1", - "shortName": "Llama 3 Euryale 70B v2.1", - "author": "sao10k", - "description": "Euryale 70B v2.1 is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k).\n\n- Better prompt adherence.\n- Better anatomy / spatial awareness.\n- Adapts much better to unique and custom formatting / reply formats.\n- Very creative, lots of unique swipes.\n- Is not restrictive during roleplays.", + "name": "Perplexity: Sonar Reasoning", + "shortName": "Sonar Reasoning", + "author": "perplexity", + "description": "Sonar Reasoning is a reasoning model provided by Perplexity based on [DeepSeek R1](/deepseek/deepseek-r1).\n\nIt allows developers to utilize long chain of thought with built-in web search. Sonar Reasoning is uncensored and hosted in US datacenters. ", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 127000, "inputModalities": [ "text" ], @@ -55828,35 +58751,43 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Other", + "instructType": "deepseek-r1", "defaultSystem": null, "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" + "<｜User｜>", + "<｜end▁of▁sentence｜>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "sao10k/l3-euryale-70b", - "reasoningConfig": null, - "features": {}, + "permaslug": "perplexity/sonar-reasoning", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "2ab67dc9-421b-408e-b61a-b4b86ea1df70", - "name": "Novita | sao10k/l3-euryale-70b", - "contextLength": 8192, + "id": "dbca9058-f3e6-4487-9651-f2f4c9f44190", + "name": "Perplexity | perplexity/sonar-reasoning", + "contextLength": 127000, "model": { - "slug": "sao10k/l3-euryale-70b", - "hfSlug": "Sao10K/L3-70B-Euryale-v2.1", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-06-18T00:00:00+00:00", + "slug": "perplexity/sonar-reasoning", + "hfSlug": "", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-01-29T06:11:47.38089+00:00", "hfUpdatedAt": null, - "name": "Sao10k: Llama 3 Euryale 70B v2.1", - "shortName": "Llama 3 Euryale 70B v2.1", - "author": "sao10k", - "description": "Euryale 70B v2.1 is a model focused on creative roleplay from [Sao10k](https://ko-fi.com/sao10k).\n\n- Better prompt adherence.\n- Better anatomy / spatial awareness.\n- Adapts much better to unique and custom formatting / reply formats.\n- Very creative, lots of unique swipes.\n- Is not restrictive during roleplays.", + "name": "Perplexity: Sonar Reasoning", + "shortName": "Sonar Reasoning", + "author": "perplexity", + "description": "Sonar Reasoning is a reasoning model provided by Perplexity based on [DeepSeek R1](/deepseek/deepseek-r1).\n\nIt allows developers to utilize long chain of thought with built-in web search. Sonar Reasoning is uncensored and hosted in US datacenters. ", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 127000, "inputModalities": [ "text" ], @@ -55864,105 +58795,126 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Other", + "instructType": "deepseek-r1", "defaultSystem": null, "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" + "<｜User｜>", + "<｜end▁of▁sentence｜>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "sao10k/l3-euryale-70b", - "reasoningConfig": null, - "features": {} + "permaslug": "perplexity/sonar-reasoning", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "sao10k/l3-euryale-70b", - "modelVariantPermaslug": "sao10k/l3-euryale-70b", - "providerName": "Novita", + "modelVariantSlug": "perplexity/sonar-reasoning", + "modelVariantPermaslug": "perplexity/sonar-reasoning", + "providerName": "Perplexity", "providerInfo": { - "name": "Novita", - "displayName": "NovitaAI", - "slug": "novita", + "name": "Perplexity", + "displayName": "Perplexity", + "slug": "perplexity", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", + "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", "paidModels": { "training": false } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, + "hasCompletions": false, + "isAbortable": false, "moderationRequired": false, - "group": "Novita", + "group": "Perplexity", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.novita.ai/", + "statusPageUrl": "https://status.perplexity.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", - "invertRequired": true + "url": "/images/icons/Perplexity.svg" } }, - "providerDisplayName": "NovitaAI", - "providerModelId": "sao10k/l3-70b-euryale-v2.1", - "providerGroup": "Novita", - "quantization": "fp8", + "providerDisplayName": "Perplexity", + "providerModelId": "sonar-reasoning", + "providerGroup": "Perplexity", + "quantization": null, "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 8192, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", + "reasoning", + "include_reasoning", + "web_search_options", "top_k", - "min_p", - "repetition_penalty", - "logit_bias" + "frequency_penalty", + "presence_penalty" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", + "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", "paidModels": { "training": false }, "training": false }, "pricing": { - "prompt": "0.00000148", - "completion": "0.00000148", + "prompt": "0.000001", + "completion": "0.000005", "image": "0", - "request": "0", + "request": "0.005", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [], + "variablePricings": [ + { + "type": "search-threshold", + "threshold": "high", + "request": "0.012" + }, + { + "type": "search-threshold", + "threshold": "medium", + "request": "0.008" + }, + { + "type": "search-threshold", + "threshold": "low", + "request": "0.005" + } + ], "isHidden": false, "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -55970,26 +58922,28 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 199, - "newest": 246, - "throughputHighToLow": 229, - "latencyLowToHigh": 141, - "pricingLowToHigh": 235, - "pricingHighToLow": 83 + "topWeekly": 201, + "newest": 137, + "throughputHighToLow": 90, + "latencyLowToHigh": 242, + "pricingLowToHigh": 289, + "pricingHighToLow": 25 }, + "authorIcon": "https://openrouter.ai/images/icons/Perplexity.svg", "providers": [ { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": "fp8", - "context": 8192, - "maxCompletionTokens": 8192, - "providerModelId": "sao10k/l3-70b-euryale-v2.1", + "name": "Perplexity", + "icon": "https://openrouter.ai/images/icons/Perplexity.svg", + "slug": "perplexity", + "quantization": null, + "context": 127000, + "maxCompletionTokens": null, + "providerModelId": "sonar-reasoning", "pricing": { - "prompt": "0.00000148", - "completion": "0.00000148", + "prompt": "0.000001", + "completion": "0.000005", "image": "0", - "request": "0", + "request": "0.005", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -55998,95 +58952,96 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", + "reasoning", + "include_reasoning", + "web_search_options", "top_k", - "min_p", - "repetition_penalty", - "logit_bias" + "frequency_penalty", + "presence_penalty" ], - "inputCost": 1.48, - "outputCost": 1.48 + "inputCost": 1, + "outputCost": 5, + "throughput": 88.3535, + "latency": 1746.5 } ] }, { - "slug": "openai/gpt-3.5-turbo-0125", + "slug": "anthropic/claude-3.5-haiku", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-05-28T00:00:00+00:00", + "createdAt": "2024-11-04T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-3.5 Turbo 16k", - "shortName": "GPT-3.5 Turbo 16k", - "author": "openai", - "description": "The latest GPT-3.5 Turbo model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more. Training data: up to Sep 2021.\n\nThis version has a higher accuracy at responding in requested formats and a fix for a bug which caused a text encoding issue for non-English language function calls.", - "modelVersionGroupId": null, - "contextLength": 16385, + "name": "Anthropic: Claude 3.5 Haiku (self-moderated)", + "shortName": "Claude 3.5 Haiku (self-moderated)", + "author": "anthropic", + "description": "Claude 3.5 Haiku features offers enhanced capabilities in speed, coding accuracy, and tool use. Engineered to excel in real-time applications, it delivers quick response times that are essential for dynamic tasks such as chat interactions and immediate coding suggestions.\n\nThis makes it highly suitable for environments that demand both speed and precision, such as software development, customer service bots, and data management systems.\n\nThis model is currently pointing to [Claude 3.5 Haiku (2024-10-22)](/anthropic/claude-3-5-haiku-20241022).", + "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", + "contextLength": 200000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "GPT", + "group": "Claude", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-3.5-turbo-0125", + "permaslug": "anthropic/claude-3-5-haiku", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "9be0749c-ce48-4f4a-9e29-b74dbc29d7d6", - "name": "OpenAI | openai/gpt-3.5-turbo-0125", - "contextLength": 16385, + "id": "ced809d4-9715-4634-8b4d-ea0e09d6bb85", + "name": "Anthropic | anthropic/claude-3-5-haiku:beta", + "contextLength": 200000, "model": { - "slug": "openai/gpt-3.5-turbo-0125", + "slug": "anthropic/claude-3.5-haiku", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-05-28T00:00:00+00:00", + "createdAt": "2024-11-04T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-3.5 Turbo 16k", - "shortName": "GPT-3.5 Turbo 16k", - "author": "openai", - "description": "The latest GPT-3.5 Turbo model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more. Training data: up to Sep 2021.\n\nThis version has a higher accuracy at responding in requested formats and a fix for a bug which caused a text encoding issue for non-English language function calls.", - "modelVersionGroupId": null, - "contextLength": 16385, + "name": "Anthropic: Claude 3.5 Haiku", + "shortName": "Claude 3.5 Haiku", + "author": "anthropic", + "description": "Claude 3.5 Haiku features offers enhanced capabilities in speed, coding accuracy, and tool use. Engineered to excel in real-time applications, it delivers quick response times that are essential for dynamic tasks such as chat interactions and immediate coding suggestions.\n\nThis makes it highly suitable for environments that demand both speed and precision, such as software development, customer service bots, and data management systems.\n\nThis model is currently pointing to [Claude 3.5 Haiku (2024-10-22)](/anthropic/claude-3-5-haiku-20241022).", + "modelVersionGroupId": "028ec497-a034-40fd-81fe-f51d0a0c640c", + "contextLength": 200000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "GPT", + "group": "Claude", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-3.5-turbo-0125", + "permaslug": "anthropic/claude-3-5-haiku", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/gpt-3.5-turbo-0125", - "modelVariantPermaslug": "openai/gpt-3.5-turbo-0125", - "providerName": "OpenAI", + "modelVariantSlug": "anthropic/claude-3.5-haiku:beta", + "modelVariantPermaslug": "anthropic/claude-3-5-haiku:beta", + "providerName": "Anthropic", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Anthropic", + "displayName": "Anthropic", + "slug": "anthropic", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", "paidModels": { "training": false, "retainsPrompts": true, @@ -56099,27 +59054,26 @@ export default { "hasCompletions": true, "isAbortable": true, "moderationRequired": true, - "group": "OpenAI", + "group": "Anthropic", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": "https://status.anthropic.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true + "url": "/images/icons/Anthropic.svg" } }, - "providerDisplayName": "OpenAI", - "providerModelId": "gpt-3.5-turbo-0125", - "providerGroup": "OpenAI", + "providerDisplayName": "Anthropic", + "providerModelId": "claude-3-5-haiku-20241022", + "providerGroup": "Anthropic", "quantization": "unknown", - "variant": "standard", + "variant": "beta", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": 8192, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -56128,21 +59082,14 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format" + "top_k", + "stop" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", "paidModels": { "training": false, "retainsPrompts": true, @@ -56154,10 +59101,12 @@ export default { "retentionDays": 30 }, "pricing": { - "prompt": "0.0000005", - "completion": "0.0000015", + "prompt": "0.0000008", + "completion": "0.000004", "image": "0", "request": "0", + "inputCacheRead": "0.00000008", + "inputCacheWrite": "0.000001", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -56179,26 +59128,30 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 200, - "newest": 318, - "throughputHighToLow": 45, - "latencyLowToHigh": 60, - "pricingLowToHigh": 190, - "pricingHighToLow": 131 + "topWeekly": 47, + "newest": 177, + "throughputHighToLow": 153, + "latencyLowToHigh": 187, + "pricingLowToHigh": 221, + "pricingHighToLow": 93 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { - "name": "OpenAI", - "slug": "openAi", + "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", + "slug": "anthropic", "quantization": "unknown", - "context": 16385, - "maxCompletionTokens": 4096, - "providerModelId": "gpt-3.5-turbo-0125", + "context": 200000, + "maxCompletionTokens": 8192, + "providerModelId": "claude-3-5-haiku-20241022", "pricing": { - "prompt": "0.0000005", - "completion": "0.0000015", + "prompt": "0.0000008", + "completion": "0.000004", "image": "0", "request": "0", + "inputCacheRead": "0.00000008", + "inputCacheWrite": "0.000001", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -56209,178 +59162,254 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format" + "top_k", + "stop" ], - "inputCost": 0.5, - "outputCost": 1.5 + "inputCost": 0.8, + "outputCost": 4, + "throughput": 57.0545, + "latency": 1171.5 + }, + { + "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", + "slug": "vertex", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 8192, + "providerModelId": "claude-3-5-haiku@20241022", + "pricing": { + "prompt": "0.0000008", + "completion": "0.000004", + "image": "0", + "request": "0", + "inputCacheRead": "0.00000008", + "inputCacheWrite": "0.000001", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 0.8, + "outputCost": 4, + "throughput": 61.668, + "latency": 2440.5 + }, + { + "name": "Amazon Bedrock", + "icon": "https://openrouter.ai/images/icons/Bedrock.svg", + "slug": "amazonBedrock", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 8192, + "providerModelId": "us.anthropic.claude-3-5-haiku-20241022-v1:0", + "pricing": { + "prompt": "0.0000008", + "completion": "0.000004", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 0.8, + "outputCost": 4, + "throughput": 55.818, + "latency": 1382 + }, + { + "name": "Amazon Bedrock (US-WEST)", + "icon": "", + "slug": "amazonBedrock (usWest)", + "quantization": null, + "context": 200000, + "maxCompletionTokens": 8192, + "providerModelId": "us.anthropic.claude-3-5-haiku-20241022-v1:0", + "pricing": { + "prompt": "0.0000008", + "completion": "0.000004", + "image": "0", + "request": "0", + "inputCacheRead": "0.00000008", + "inputCacheWrite": "0.000001", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 0.8, + "outputCost": 4, + "throughput": 55.934, + "latency": 1298.5 } ] }, { - "slug": "thudm/glm-z1-32b", - "hfSlug": "THUDM/GLM-Z1-32B-0414", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-04-17T21:09:08.005794+00:00", + "slug": "anthropic/claude-3-opus", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-03-05T00:00:00+00:00", "hfUpdatedAt": null, - "name": "THUDM: GLM Z1 32B", - "shortName": "GLM Z1 32B", - "author": "thudm", - "description": "GLM-Z1-32B-0414 is an enhanced reasoning variant of GLM-4-32B, built for deep mathematical, logical, and code-oriented problem solving. It applies extended reinforcement learning—both task-specific and general pairwise preference-based—to improve performance on complex multi-step tasks. Compared to the base GLM-4-32B model, Z1 significantly boosts capabilities in structured reasoning and formal domains.\n\nThe model supports enforced “thinking” steps via prompt engineering and offers improved coherence for long-form outputs. It’s optimized for use in agentic workflows, and includes support for long context (via YaRN), JSON tool calling, and fine-grained sampling configuration for stable inference. Ideal for use cases requiring deliberate, multi-step reasoning or formal derivations.", + "name": "Anthropic: Claude 3 Opus (self-moderated)", + "shortName": "Claude 3 Opus (self-moderated)", + "author": "anthropic", + "description": "Claude 3 Opus is Anthropic's most powerful model for highly complex tasks. It boasts top-level performance, intelligence, fluency, and understanding.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal", "modelVersionGroupId": null, - "contextLength": 32000, + "contextLength": 200000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "deepseek-r1", + "group": "Claude", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "thudm/glm-z1-32b-0414", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "permaslug": "anthropic/claude-3-opus", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "f32bfaa2-38d1-4b3b-b5b7-568d2df68ca4", - "name": "Novita | thudm/glm-z1-32b-0414", - "contextLength": 32000, + "id": "ee74a4e0-2863-4f51-99a5-997c31c48ae7", + "name": "Anthropic | anthropic/claude-3-opus:beta", + "contextLength": 200000, "model": { - "slug": "thudm/glm-z1-32b", - "hfSlug": "THUDM/GLM-Z1-32B-0414", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-04-17T21:09:08.005794+00:00", + "slug": "anthropic/claude-3-opus", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-03-05T00:00:00+00:00", "hfUpdatedAt": null, - "name": "THUDM: GLM Z1 32B", - "shortName": "GLM Z1 32B", - "author": "thudm", - "description": "GLM-Z1-32B-0414 is an enhanced reasoning variant of GLM-4-32B, built for deep mathematical, logical, and code-oriented problem solving. It applies extended reinforcement learning—both task-specific and general pairwise preference-based—to improve performance on complex multi-step tasks. Compared to the base GLM-4-32B model, Z1 significantly boosts capabilities in structured reasoning and formal domains.\n\nThe model supports enforced “thinking” steps via prompt engineering and offers improved coherence for long-form outputs. It’s optimized for use in agentic workflows, and includes support for long context (via YaRN), JSON tool calling, and fine-grained sampling configuration for stable inference. Ideal for use cases requiring deliberate, multi-step reasoning or formal derivations.", + "name": "Anthropic: Claude 3 Opus", + "shortName": "Claude 3 Opus", + "author": "anthropic", + "description": "Claude 3 Opus is Anthropic's most powerful model for highly complex tasks. It boasts top-level performance, intelligence, fluency, and understanding.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 200000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "deepseek-r1", + "group": "Claude", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "thudm/glm-z1-32b-0414", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "permaslug": "anthropic/claude-3-opus", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "thudm/glm-z1-32b", - "modelVariantPermaslug": "thudm/glm-z1-32b-0414", - "providerName": "Novita", + "modelVariantSlug": "anthropic/claude-3-opus:beta", + "modelVariantPermaslug": "anthropic/claude-3-opus:beta", + "providerName": "Anthropic", "providerInfo": { - "name": "Novita", - "displayName": "NovitaAI", - "slug": "novita", + "name": "Anthropic", + "displayName": "Anthropic", + "slug": "anthropic", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", "paidModels": { - "training": false - } + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": false, - "group": "Novita", + "moderationRequired": true, + "group": "Anthropic", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.novita.ai/", + "statusPageUrl": "https://status.anthropic.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", - "invertRequired": true + "url": "/images/icons/Anthropic.svg" } }, - "providerDisplayName": "NovitaAI", - "providerModelId": "thudm/glm-z1-32b-0414", - "providerGroup": "Novita", - "quantization": null, - "variant": "standard", + "providerDisplayName": "Anthropic", + "providerModelId": "claude-3-opus-20240229", + "providerGroup": "Anthropic", + "quantization": "unknown", + "variant": "beta", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", "top_k", - "min_p", - "repetition_penalty", - "logit_bias" + "stop" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "training": false + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.00000024", - "completion": "0.00000024", - "image": "0", + "prompt": "0.000015", + "completion": "0.000075", + "image": "0.024", "request": "0", + "inputCacheRead": "0.0000015", + "inputCacheWrite": "0.00001875", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -56389,192 +59418,201 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": true, + "supportsToolParameters": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 201, - "newest": 37, - "throughputHighToLow": 179, - "latencyLowToHigh": 220, - "pricingLowToHigh": 18, - "pricingHighToLow": 161 + "topWeekly": 135, + "newest": 279, + "throughputHighToLow": 247, + "latencyLowToHigh": 232, + "pricingLowToHigh": 309, + "pricingHighToLow": 8 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": null, - "context": 32000, - "maxCompletionTokens": null, - "providerModelId": "thudm/glm-z1-32b-0414", + "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", + "slug": "anthropic", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 4096, + "providerModelId": "claude-3-opus-20240229", + "pricing": { + "prompt": "0.000015", + "completion": "0.000075", + "image": "0.024", + "request": "0", + "inputCacheRead": "0.0000015", + "inputCacheWrite": "0.00001875", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 15, + "outputCost": 75, + "throughput": 30.035, + "latency": 1648.5 + }, + { + "name": "Google Vertex", + "icon": "https://openrouter.ai/images/icons/GoogleVertex.svg", + "slug": "vertex", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 4096, + "providerModelId": "claude-3-opus@20240229", "pricing": { - "prompt": "0.00000024", - "completion": "0.00000024", - "image": "0", + "prompt": "0.000015", + "completion": "0.000075", + "image": "0.024", "request": "0", + "inputCacheRead": "0.0000015", + "inputCacheWrite": "0.00001875", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", "top_k", - "min_p", - "repetition_penalty", - "logit_bias" + "stop" ], - "inputCost": 0.24, - "outputCost": 0.24 + "inputCost": 15, + "outputCost": 75, + "throughput": 17.365, + "latency": 3104 } ] }, { - "slug": "perplexity/sonar-reasoning-pro", - "hfSlug": "", - "updatedAt": "2025-05-05T14:36:52.712881+00:00", - "createdAt": "2025-03-07T02:08:28.125446+00:00", + "slug": "moonshotai/kimi-vl-a3b-thinking", + "hfSlug": "moonshotai/Kimi-VL-A3B-Thinking", + "updatedAt": "2025-04-10T17:10:17.814287+00:00", + "createdAt": "2025-04-10T17:07:21.175402+00:00", "hfUpdatedAt": null, - "name": "Perplexity: Sonar Reasoning Pro", - "shortName": "Sonar Reasoning Pro", - "author": "perplexity", - "description": "Note: Sonar Pro pricing includes Perplexity search pricing. See [details here](https://docs.perplexity.ai/guides/pricing#detailed-pricing-breakdown-for-sonar-reasoning-pro-and-sonar-pro)\n\nSonar Reasoning Pro is a premier reasoning model powered by DeepSeek R1 with Chain of Thought (CoT). Designed for advanced use cases, it supports in-depth, multi-step queries with a larger context window and can surface more citations per search, enabling more comprehensive and extensible responses.", + "name": "Moonshot AI: Kimi VL A3B Thinking (free)", + "shortName": "Kimi VL A3B Thinking (free)", + "author": "moonshotai", + "description": "Kimi-VL is a lightweight Mixture-of-Experts vision-language model that activates only 2.8B parameters per step while delivering strong performance on multimodal reasoning and long-context tasks. The Kimi-VL-A3B-Thinking variant, fine-tuned with chain-of-thought and reinforcement learning, excels in math and visual reasoning benchmarks like MathVision, MMMU, and MathVista, rivaling much larger models such as Qwen2.5-VL-7B and Gemma-3-12B. It supports 128K context and high-resolution input via its MoonViT encoder.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 131072, "inputModalities": [ - "text", - "image" + "image", + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, "group": "Other", - "instructType": "deepseek-r1", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "perplexity/sonar-reasoning-pro", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "permaslug": "moonshotai/kimi-vl-a3b-thinking", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "0d28660e-435a-4853-a2c6-9d916df28fc7", - "name": "Perplexity | perplexity/sonar-reasoning-pro", - "contextLength": 128000, + "id": "ba332390-9210-4e73-a2c4-0fab667c7fb1", + "name": "Chutes | moonshotai/kimi-vl-a3b-thinking:free", + "contextLength": 131072, "model": { - "slug": "perplexity/sonar-reasoning-pro", - "hfSlug": "", - "updatedAt": "2025-05-05T14:36:52.712881+00:00", - "createdAt": "2025-03-07T02:08:28.125446+00:00", + "slug": "moonshotai/kimi-vl-a3b-thinking", + "hfSlug": "moonshotai/Kimi-VL-A3B-Thinking", + "updatedAt": "2025-04-10T17:10:17.814287+00:00", + "createdAt": "2025-04-10T17:07:21.175402+00:00", "hfUpdatedAt": null, - "name": "Perplexity: Sonar Reasoning Pro", - "shortName": "Sonar Reasoning Pro", - "author": "perplexity", - "description": "Note: Sonar Pro pricing includes Perplexity search pricing. See [details here](https://docs.perplexity.ai/guides/pricing#detailed-pricing-breakdown-for-sonar-reasoning-pro-and-sonar-pro)\n\nSonar Reasoning Pro is a premier reasoning model powered by DeepSeek R1 with Chain of Thought (CoT). Designed for advanced use cases, it supports in-depth, multi-step queries with a larger context window and can surface more citations per search, enabling more comprehensive and extensible responses.", + "name": "Moonshot AI: Kimi VL A3B Thinking", + "shortName": "Kimi VL A3B Thinking", + "author": "moonshotai", + "description": "Kimi-VL is a lightweight Mixture-of-Experts vision-language model that activates only 2.8B parameters per step while delivering strong performance on multimodal reasoning and long-context tasks. The Kimi-VL-A3B-Thinking variant, fine-tuned with chain-of-thought and reinforcement learning, excels in math and visual reasoning benchmarks like MathVision, MMMU, and MathVista, rivaling much larger models such as Qwen2.5-VL-7B and Gemma-3-12B. It supports 128K context and high-resolution input via its MoonViT encoder.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 131072, "inputModalities": [ - "text", - "image" + "image", + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, "group": "Other", - "instructType": "deepseek-r1", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "perplexity/sonar-reasoning-pro", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "permaslug": "moonshotai/kimi-vl-a3b-thinking", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "perplexity/sonar-reasoning-pro", - "modelVariantPermaslug": "perplexity/sonar-reasoning-pro", - "providerName": "Perplexity", + "modelVariantSlug": "moonshotai/kimi-vl-a3b-thinking:free", + "modelVariantPermaslug": "moonshotai/kimi-vl-a3b-thinking:free", + "providerName": "Chutes", "providerInfo": { - "name": "Perplexity", - "displayName": "Perplexity", - "slug": "perplexity", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", - "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true } }, - "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, + "hasCompletions": true, + "isAbortable": true, "moderationRequired": false, - "group": "Perplexity", + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.perplexity.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Perplexity.svg" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "Perplexity", - "providerModelId": "sonar-reasoning-pro", - "providerGroup": "Perplexity", + "providerDisplayName": "Chutes", + "providerModelId": "moonshotai/Kimi-VL-A3B-Thinking", + "providerGroup": "Chutes", "quantization": null, - "variant": "standard", - "isFree": false, - "canAbort": false, + "variant": "free", + "isFree": true, + "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, - "maxPromptImages": null, + "maxPromptImages": 8, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", @@ -56582,47 +59620,38 @@ export default { "top_p", "reasoning", "include_reasoning", - "web_search_options", - "top_k", + "stop", "frequency_penalty", - "presence_penalty" + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", - "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true }, - "training": false + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.000002", - "completion": "0.000008", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", - "webSearch": "0.005", + "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [ - { - "type": "search-threshold", - "threshold": "high", - "request": "0.014" - }, - { - "type": "search-threshold", - "threshold": "medium", - "request": "0.01" - }, - { - "type": "search-threshold", - "threshold": "low", - "request": "0.006" - } - ], + "variablePricings": [], "isHidden": false, "isDeranked": false, "isDisabled": false, @@ -56631,164 +59660,137 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 202, - "newest": 98, - "throughputHighToLow": 163, - "latencyLowToHigh": 258, - "pricingLowToHigh": 249, - "pricingHighToLow": 69 + "topWeekly": 204, + "newest": 55, + "throughputHighToLow": 168, + "latencyLowToHigh": 282, + "pricingLowToHigh": 23, + "pricingHighToLow": 271 }, - "providers": [ - { - "name": "Perplexity", - "slug": "perplexity", - "quantization": null, - "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "sonar-reasoning-pro", - "pricing": { - "prompt": "0.000002", - "completion": "0.000008", - "image": "0", - "request": "0", - "webSearch": "0.005", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "web_search_options", - "top_k", - "frequency_penalty", - "presence_penalty" - ], - "inputCost": 2, - "outputCost": 8 - } - ] + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [] }, { - "slug": "infermatic/mn-inferor-12b", - "hfSlug": "Infermatic/MN-12B-Inferor-v0.0", + "slug": "google/learnlm-1.5-pro-experimental", + "hfSlug": "", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-13T02:20:28.884947+00:00", + "createdAt": "2024-11-21T19:15:51.117686+00:00", "hfUpdatedAt": null, - "name": "Infermatic: Mistral Nemo Inferor 12B", - "shortName": "Mistral Nemo Inferor 12B", - "author": "infermatic", - "description": "Inferor 12B is a merge of top roleplay models, expert on immersive narratives and storytelling.\n\nThis model was merged using the [Model Stock](https://arxiv.org/abs/2403.19522) merge method using [anthracite-org/magnum-v4-12b](https://openrouter.ai/anthracite-org/magnum-v4-72b) as a base.\n", + "name": "Google: LearnLM 1.5 Pro Experimental (free)", + "shortName": "LearnLM 1.5 Pro Experimental (free)", + "author": "google", + "description": "An experimental version of [Gemini 1.5 Pro](/google/gemini-pro-1.5) from Google.", "modelVersionGroupId": null, - "contextLength": 16384, + "contextLength": 32767, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "mistral", + "group": "Gemini", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "[INST]", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "infermatic/mn-inferor-12b", + "permaslug": "google/learnlm-1.5-pro-experimental", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "cc0c2222-fb27-4c56-8ce6-1f6a6be16a93", - "name": "Featherless | infermatic/mn-inferor-12b", - "contextLength": 16384, + "id": "bf41a79d-a83a-4b05-afad-b69bcc66dddc", + "name": "Google AI Studio | google/learnlm-1.5-pro-experimental:free", + "contextLength": 32767, "model": { - "slug": "infermatic/mn-inferor-12b", - "hfSlug": "Infermatic/MN-12B-Inferor-v0.0", + "slug": "google/learnlm-1.5-pro-experimental", + "hfSlug": "", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-13T02:20:28.884947+00:00", + "createdAt": "2024-11-21T19:15:51.117686+00:00", "hfUpdatedAt": null, - "name": "Infermatic: Mistral Nemo Inferor 12B", - "shortName": "Mistral Nemo Inferor 12B", - "author": "infermatic", - "description": "Inferor 12B is a merge of top roleplay models, expert on immersive narratives and storytelling.\n\nThis model was merged using the [Model Stock](https://arxiv.org/abs/2403.19522) merge method using [anthracite-org/magnum-v4-12b](https://openrouter.ai/anthracite-org/magnum-v4-72b) as a base.\n", + "name": "Google: LearnLM 1.5 Pro Experimental", + "shortName": "LearnLM 1.5 Pro Experimental", + "author": "google", + "description": "An experimental version of [Gemini 1.5 Pro](/google/gemini-pro-1.5) from Google.", "modelVersionGroupId": null, - "contextLength": 32000, + "contextLength": 40960, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "mistral", + "group": "Gemini", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "[INST]", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "infermatic/mn-inferor-12b", + "permaslug": "google/learnlm-1.5-pro-experimental", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "infermatic/mn-inferor-12b", - "modelVariantPermaslug": "infermatic/mn-inferor-12b", - "providerName": "Featherless", + "modelVariantSlug": "google/learnlm-1.5-pro-experimental:free", + "modelVariantPermaslug": "google/learnlm-1.5-pro-experimental:free", + "providerName": "Google AI Studio", "providerInfo": { - "name": "Featherless", - "displayName": "Featherless", - "slug": "featherless", + "name": "Google AI Studio", + "displayName": "Google AI Studio", + "slug": "google-ai-studio", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", + "termsOfServiceUrl": "https://cloud.google.com/terms/", + "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 55 + }, + "freeModels": { + "training": true, + "retainsPrompts": true, + "retentionDays": 55 } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, + "hasCompletions": false, "isAbortable": false, "moderationRequired": false, - "group": "Featherless", + "group": "Google AI Studio", "editors": [], "owners": [], - "isMultipartSupported": false, + "isMultipartSupported": true, "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" + "url": "/images/icons/GoogleAIStudio.svg" } }, - "providerDisplayName": "Featherless", - "providerModelId": "Infermatic/MN-12B-Inferor-v0.0", - "providerGroup": "Featherless", + "providerDisplayName": "Google AI Studio", + "providerModelId": "learnlm-1.5-pro-experimental", + "providerGroup": "Google AI Studio", "quantization": null, - "variant": "standard", - "isFree": false, + "variant": "free", + "isFree": true, "canAbort": false, - "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxPromptTokens": 32767, + "maxCompletionTokens": 8192, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -56798,26 +59800,32 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "top_k", - "min_p", - "seed" + "seed", + "response_format", + "structured_outputs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", + "termsOfServiceUrl": "https://cloud.google.com/terms/", + "privacyPolicyUrl": "https://cloud.google.com/terms/cloud-privacy-notice", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 55 }, - "training": false, - "retainsPrompts": false + "freeModels": { + "training": true, + "retainsPrompts": true, + "retentionDays": 55 + }, + "training": true, + "retainsPrompts": true, + "retentionDays": 55 }, "pricing": { - "prompt": "0.0000008", - "completion": "0.0000012", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -56830,10 +59838,10 @@ export default { "isDisabled": false, "supportsToolParameters": false, "supportsReasoning": false, - "supportsMultipart": false, + "supportsMultipart": true, "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, + "limitRpd": 80, + "hasCompletions": false, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -56841,59 +59849,28 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 203, - "newest": 171, - "throughputHighToLow": 282, - "latencyLowToHigh": 271, - "pricingLowToHigh": 207, - "pricingHighToLow": 109 + "topWeekly": 205, + "newest": 164, + "throughputHighToLow": 309, + "latencyLowToHigh": 316, + "pricingLowToHigh": 58, + "pricingHighToLow": 306 }, - "providers": [ - { - "name": "Featherless", - "slug": "featherless", - "quantization": null, - "context": 16384, - "maxCompletionTokens": 4096, - "providerModelId": "Infermatic/MN-12B-Inferor-v0.0", - "pricing": { - "prompt": "0.0000008", - "completion": "0.0000012", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "top_k", - "min_p", - "seed" - ], - "inputCost": 0.8, - "outputCost": 1.2 - } - ] + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", + "providers": [] }, { - "slug": "inception/mercury-coder-small-beta", - "hfSlug": "", - "updatedAt": "2025-04-30T17:36:42.979824+00:00", - "createdAt": "2025-04-30T17:24:40+00:00", + "slug": "deepseek/deepseek-r1-distill-qwen-1.5b", + "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-01-31T12:54:27.964156+00:00", "hfUpdatedAt": null, - "name": "Inception: Mercury Coder Small Beta", - "shortName": "Mercury Coder Small Beta", - "author": "inception", - "description": "Mercury Coder Small is the first diffusion large language model (dLLM). Applying a breakthrough discrete diffusion approach, the model runs 5-10x faster than even speed optimized models like Claude 3.5 Haiku and GPT-4o Mini while matching their performance. Mercury Coder Small's speed means that developers can stay in the flow while coding, enjoying rapid chat-based iteration and responsive code completion suggestions. On Copilot Arena, Mercury Coder ranks 1st in speed and ties for 2nd in quality. Read more in the [blog post here](https://www.inceptionlabs.ai/introducing-mercury).", - "modelVersionGroupId": null, - "contextLength": 32000, + "name": "DeepSeek: R1 Distill Qwen 1.5B", + "shortName": "R1 Distill Qwen 1.5B", + "author": "deepseek", + "description": "DeepSeek R1 Distill Qwen 1.5B is a distilled large language model based on [Qwen 2.5 Math 1.5B](https://huggingface.co/Qwen/Qwen2.5-Math-1.5B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It's a very small and efficient model which outperforms [GPT 4o 0513](/openai/gpt-4o-2024-05-13) on Math Benchmarks.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 28.9\n- AIME 2024 cons@64: 52.7\n- MATH-500 pass@1: 83.9\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", + "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", + "contextLength": 131072, "inputModalities": [ "text" ], @@ -56902,31 +59879,42 @@ export default { ], "hasTextOutput": true, "group": "Other", - "instructType": null, + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "inception/mercury-coder-small-beta", - "reasoningConfig": null, - "features": {}, + "permaslug": "deepseek/deepseek-r1-distill-qwen-1.5b", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "eb51fb37-11b0-42d1-924a-c25fa2375569", - "name": "Inception | inception/mercury-coder-small-beta", - "contextLength": 32000, + "id": "85bb5aae-0f60-4233-b275-d4e370685b3a", + "name": "Together | deepseek/deepseek-r1-distill-qwen-1.5b", + "contextLength": 131072, "model": { - "slug": "inception/mercury-coder-small-beta", - "hfSlug": "", - "updatedAt": "2025-04-30T17:36:42.979824+00:00", - "createdAt": "2025-04-30T17:24:40+00:00", + "slug": "deepseek/deepseek-r1-distill-qwen-1.5b", + "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-01-31T12:54:27.964156+00:00", "hfUpdatedAt": null, - "name": "Inception: Mercury Coder Small Beta", - "shortName": "Mercury Coder Small Beta", - "author": "inception", - "description": "Mercury Coder Small is the first diffusion large language model (dLLM). Applying a breakthrough discrete diffusion approach, the model runs 5-10x faster than even speed optimized models like Claude 3.5 Haiku and GPT-4o Mini while matching their performance. Mercury Coder Small's speed means that developers can stay in the flow while coding, enjoying rapid chat-based iteration and responsive code completion suggestions. On Copilot Arena, Mercury Coder ranks 1st in speed and ties for 2nd in quality. Read more in the [blog post here](https://www.inceptionlabs.ai/introducing-mercury).", - "modelVersionGroupId": null, - "contextLength": 32000, + "name": "DeepSeek: R1 Distill Qwen 1.5B", + "shortName": "R1 Distill Qwen 1.5B", + "author": "deepseek", + "description": "DeepSeek R1 Distill Qwen 1.5B is a distilled large language model based on [Qwen 2.5 Math 1.5B](https://huggingface.co/Qwen/Qwen2.5-Math-1.5B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It's a very small and efficient model which outperforms [GPT 4o 0513](/openai/gpt-4o-2024-05-13) on Math Benchmarks.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 28.9\n- AIME 2024 cons@64: 52.7\n- MATH-500 pass@1: 83.9\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", + "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", + "contextLength": 131072, "inputModalities": [ "text" ], @@ -56935,37 +59923,49 @@ export default { ], "hasTextOutput": true, "group": "Other", - "instructType": null, + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "inception/mercury-coder-small-beta", - "reasoningConfig": null, - "features": {} + "permaslug": "deepseek/deepseek-r1-distill-qwen-1.5b", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "inception/mercury-coder-small-beta", - "modelVariantPermaslug": "inception/mercury-coder-small-beta", - "providerName": "Inception", + "modelVariantSlug": "deepseek/deepseek-r1-distill-qwen-1.5b", + "modelVariantPermaslug": "deepseek/deepseek-r1-distill-qwen-1.5b", + "providerName": "Together", "providerInfo": { - "name": "Inception", - "displayName": "Inception", - "slug": "inception", + "name": "Together", + "displayName": "Together", + "slug": "together", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.inceptionlabs.ai/terms", - "privacyPolicyUrl": "https://www.inceptionlabs.ai/terms#privacy-policy", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false } }, + "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, + "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Inception", + "group": "Together", "editors": [], "owners": [], "isMultipartSupported": true, @@ -56973,32 +59973,40 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.inceptionlabs.ai/&size=256", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" } }, - "providerDisplayName": "Inception", - "providerModelId": "mercury-coder-small", - "providerGroup": "Inception", + "providerDisplayName": "Together", + "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B", + "providerGroup": "Together", "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 32768, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", "frequency_penalty", "presence_penalty", - "stop" + "top_k", + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.inceptionlabs.ai/terms", - "privacyPolicyUrl": "https://www.inceptionlabs.ai/terms#privacy-policy", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -57007,8 +60015,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000025", - "completion": "0.000001", + "prompt": "0.00000018", + "completion": "0.00000018", "image": "0", "request": "0", "webSearch": "0", @@ -57020,37 +60028,38 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 204, - "newest": 14, - "throughputHighToLow": 0, - "latencyLowToHigh": 69, - "pricingLowToHigh": 166, - "pricingHighToLow": 151 + "topWeekly": 206, + "newest": 130, + "throughputHighToLow": 3, + "latencyLowToHigh": 31, + "pricingLowToHigh": 139, + "pricingHighToLow": 181 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [ { - "name": "Inception", - "slug": "inception", + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", "quantization": null, - "context": 32000, - "maxCompletionTokens": null, - "providerModelId": "mercury-coder-small", + "context": 131072, + "maxCompletionTokens": 32768, + "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B", "pricing": { - "prompt": "0.00000025", - "completion": "0.000001", + "prompt": "0.00000018", + "completion": "0.00000018", "image": "0", "request": "0", "webSearch": "0", @@ -57059,134 +60068,127 @@ export default { }, "supportedParameters": [ "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", "frequency_penalty", "presence_penalty", - "stop" + "top_k", + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" ], - "inputCost": 0.25, - "outputCost": 1 + "inputCost": 0.18, + "outputCost": 0.18, + "throughput": 335.4975, + "latency": 315 } ] }, { - "slug": "perplexity/sonar-reasoning", + "slug": "mistralai/pixtral-large-2411", "hfSlug": "", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-01-29T06:11:47.38089+00:00", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-11-19T00:49:48.873161+00:00", "hfUpdatedAt": null, - "name": "Perplexity: Sonar Reasoning", - "shortName": "Sonar Reasoning", - "author": "perplexity", - "description": "Sonar Reasoning is a reasoning model provided by Perplexity based on [DeepSeek R1](/deepseek/deepseek-r1).\n\nIt allows developers to utilize long chain of thought with built-in web search. Sonar Reasoning is uncensored and hosted in US datacenters. ", + "name": "Mistral: Pixtral Large 2411", + "shortName": "Pixtral Large 2411", + "author": "mistralai", + "description": "Pixtral Large is a 124B parameter, open-weight, multimodal model built on top of [Mistral Large 2](/mistralai/mistral-large-2411). The model is able to understand documents, charts and natural images.\n\nThe model is available under the Mistral Research License (MRL) for research and educational use, and the Mistral Commercial License for experimentation, testing, and production for commercial purposes.\n\n", "modelVersionGroupId": null, - "contextLength": 127000, + "contextLength": 131072, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "deepseek-r1", + "group": "Mistral", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "perplexity/sonar-reasoning", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "permaslug": "mistralai/pixtral-large-2411", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "dbca9058-f3e6-4487-9651-f2f4c9f44190", - "name": "Perplexity | perplexity/sonar-reasoning", - "contextLength": 127000, + "id": "1a41639e-c1cf-422e-a871-27bc67f03928", + "name": "Mistral | mistralai/pixtral-large-2411", + "contextLength": 131072, "model": { - "slug": "perplexity/sonar-reasoning", + "slug": "mistralai/pixtral-large-2411", "hfSlug": "", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-01-29T06:11:47.38089+00:00", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-11-19T00:49:48.873161+00:00", "hfUpdatedAt": null, - "name": "Perplexity: Sonar Reasoning", - "shortName": "Sonar Reasoning", - "author": "perplexity", - "description": "Sonar Reasoning is a reasoning model provided by Perplexity based on [DeepSeek R1](/deepseek/deepseek-r1).\n\nIt allows developers to utilize long chain of thought with built-in web search. Sonar Reasoning is uncensored and hosted in US datacenters. ", + "name": "Mistral: Pixtral Large 2411", + "shortName": "Pixtral Large 2411", + "author": "mistralai", + "description": "Pixtral Large is a 124B parameter, open-weight, multimodal model built on top of [Mistral Large 2](/mistralai/mistral-large-2411). The model is able to understand documents, charts and natural images.\n\nThe model is available under the Mistral Research License (MRL) for research and educational use, and the Mistral Commercial License for experimentation, testing, and production for commercial purposes.\n\n", "modelVersionGroupId": null, - "contextLength": 127000, + "contextLength": 128000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "deepseek-r1", + "group": "Mistral", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "perplexity/sonar-reasoning", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "permaslug": "mistralai/pixtral-large-2411", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "perplexity/sonar-reasoning", - "modelVariantPermaslug": "perplexity/sonar-reasoning", - "providerName": "Perplexity", + "modelVariantSlug": "mistralai/pixtral-large-2411", + "modelVariantPermaslug": "mistralai/pixtral-large-2411", + "providerName": "Mistral", "providerInfo": { - "name": "Perplexity", - "displayName": "Perplexity", - "slug": "perplexity", + "name": "Mistral", + "displayName": "Mistral", + "slug": "mistral", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", - "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": true, + "retentionDays": 30 } }, - "headquarters": "US", + "headquarters": "FR", "hasChatCompletions": true, "hasCompletions": false, "isAbortable": false, "moderationRequired": false, - "group": "Perplexity", + "group": "Mistral", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.perplexity.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Perplexity.svg" + "url": "/images/icons/Mistral.png" } }, - "providerDisplayName": "Perplexity", - "providerModelId": "sonar-reasoning", - "providerGroup": "Perplexity", + "providerDisplayName": "Mistral", + "providerModelId": "pixtral-large-2411", + "providerGroup": "Mistral", "quantization": null, "variant": "standard", "isFree": false, @@ -57196,120 +60198,117 @@ export default { "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", - "web_search_options", - "top_k", + "stop", "frequency_penalty", - "presence_penalty" + "presence_penalty", + "response_format", + "structured_outputs", + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", - "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "training": false + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.000001", - "completion": "0.000005", - "image": "0", - "request": "0.005", + "prompt": "0.000002", + "completion": "0.000006", + "image": "0.002888", + "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [ - { - "type": "search-threshold", - "threshold": "high", - "request": "0.012" - }, - { - "type": "search-threshold", - "threshold": "medium", - "request": "0.008" - }, - { - "type": "search-threshold", - "threshold": "low", - "request": "0.005" - } - ], + "variablePricings": [], "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": true, + "supportsToolParameters": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": false, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 205, - "newest": 137, - "throughputHighToLow": 88, - "latencyLowToHigh": 230, - "pricingLowToHigh": 289, - "pricingHighToLow": 25 + "topWeekly": 207, + "newest": 169, + "throughputHighToLow": 227, + "latencyLowToHigh": 237, + "pricingLowToHigh": 245, + "pricingHighToLow": 74 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { - "name": "Perplexity", - "slug": "perplexity", + "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", + "slug": "mistral", "quantization": null, - "context": 127000, + "context": 131072, "maxCompletionTokens": null, - "providerModelId": "sonar-reasoning", + "providerModelId": "pixtral-large-2411", "pricing": { - "prompt": "0.000001", - "completion": "0.000005", - "image": "0", - "request": "0.005", + "prompt": "0.000002", + "completion": "0.000006", + "image": "0.002888", + "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", - "web_search_options", - "top_k", + "stop", "frequency_penalty", - "presence_penalty" + "presence_penalty", + "response_format", + "structured_outputs", + "seed" ], - "inputCost": 1, - "outputCost": 5 + "inputCost": 2, + "outputCost": 6, + "throughput": 33.0465, + "latency": 1792 } ] }, { - "slug": "qwen/qwen-vl-max", - "hfSlug": "", + "slug": "qwen/qwen2.5-vl-3b-instruct", + "hfSlug": "Qwen/Qwen2.5-VL-3B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-01T18:25:04.223655+00:00", + "createdAt": "2025-03-26T18:42:53.41832+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen VL Max", - "shortName": "Qwen VL Max", + "name": "Qwen: Qwen2.5 VL 3B Instruct (free)", + "shortName": "Qwen2.5 VL 3B Instruct (free)", "author": "qwen", - "description": "Qwen VL Max is a visual understanding model with 7500 tokens context length. It excels in delivering optimal performance for a broader spectrum of complex tasks.\n", + "description": "Qwen2.5 VL 3B is a multimodal LLM from the Qwen Team with the following key enhancements:\n\n- SoTA understanding of images of various resolution & ratio: Qwen2.5-VL achieves state-of-the-art performance on visual understanding benchmarks, including MathVista, DocVQA, RealWorldQA, MTVQA, etc.\n\n- Agent that can operate your mobiles, robots, etc.: with the abilities of complex reasoning and decision making, Qwen2.5-VL can be integrated with devices like mobile phones, robots, etc., for automatic operation based on visual environment and text instructions.\n\n- Multilingual Support: to serve global users, besides English and Chinese, Qwen2.5-VL now supports the understanding of texts in different languages inside images, including most European languages, Japanese, Korean, Arabic, Vietnamese, etc.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2-vl/) and [GitHub repo](https://github.com/QwenLM/Qwen2-VL).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", "modelVersionGroupId": null, - "contextLength": 7500, + "contextLength": 64000, "inputModalities": [ "text", "image" @@ -57325,25 +60324,25 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-vl-max-2025-01-25", + "permaslug": "qwen/qwen2.5-vl-3b-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "40a33c72-3801-49f3-ac2a-966d0b249981", - "name": "Alibaba | qwen/qwen-vl-max-2025-01-25", - "contextLength": 7500, + "id": "751df51c-753b-4849-b95a-bd5281cec4ff", + "name": "Chutes | qwen/qwen2.5-vl-3b-instruct:free", + "contextLength": 64000, "model": { - "slug": "qwen/qwen-vl-max", - "hfSlug": "", + "slug": "qwen/qwen2.5-vl-3b-instruct", + "hfSlug": "Qwen/Qwen2.5-VL-3B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-01T18:25:04.223655+00:00", + "createdAt": "2025-03-26T18:42:53.41832+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen VL Max", - "shortName": "Qwen VL Max", + "name": "Qwen: Qwen2.5 VL 3B Instruct", + "shortName": "Qwen2.5 VL 3B Instruct", "author": "qwen", - "description": "Qwen VL Max is a visual understanding model with 7500 tokens context length. It excels in delivering optimal performance for a broader spectrum of complex tasks.\n", + "description": "Qwen2.5 VL 3B is a multimodal LLM from the Qwen Team with the following key enhancements:\n\n- SoTA understanding of images of various resolution & ratio: Qwen2.5-VL achieves state-of-the-art performance on visual understanding benchmarks, including MathVista, DocVQA, RealWorldQA, MTVQA, etc.\n\n- Agent that can operate your mobiles, robots, etc.: with the abilities of complex reasoning and decision making, Qwen2.5-VL can be integrated with devices like mobile phones, robots, etc., for automatic operation based on visual environment and text instructions.\n\n- Multilingual Support: to serve global users, besides English and Chinese, Qwen2.5-VL now supports the understanding of texts in different languages inside images, including most European languages, Japanese, Korean, Arabic, Vietnamese, etc.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2-vl/) and [GitHub repo](https://github.com/QwenLM/Qwen2-VL).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", "modelVersionGroupId": null, - "contextLength": 7500, + "contextLength": 64000, "inputModalities": [ "text", "image" @@ -57359,31 +60358,30 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-vl-max-2025-01-25", + "permaslug": "qwen/qwen2.5-vl-3b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "qwen/qwen-vl-max", - "modelVariantPermaslug": "qwen/qwen-vl-max-2025-01-25", - "providerName": "Alibaba", + "modelVariantSlug": "qwen/qwen2.5-vl-3b-instruct:free", + "modelVariantPermaslug": "qwen/qwen2.5-vl-3b-instruct:free", + "providerName": "Chutes", "providerInfo": { - "name": "Alibaba", - "displayName": "Alibaba", - "slug": "alibaba", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", - "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true } }, - "headquarters": "CN", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, + "isAbortable": true, "moderationRequired": false, - "group": "Alibaba", + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, @@ -57391,42 +60389,50 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.alibabacloud.com/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "Alibaba", - "providerModelId": "qwen-vl-max", - "providerGroup": "Alibaba", + "providerDisplayName": "Chutes", + "providerModelId": "Qwen/Qwen2.5-VL-3B-Instruct", + "providerGroup": "Chutes", "quantization": null, - "variant": "standard", - "isFree": false, - "canAbort": false, - "maxPromptTokens": 6000, - "maxCompletionTokens": 1500, + "variant": "free", + "isFree": true, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "seed", - "response_format", - "presence_penalty" + "top_k", + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-product-terms-of-service-v-3-8-0", - "privacyPolicyUrl": "https://www.alibabacloud.com/help/en/legal/latest/alibaba-cloud-international-website-privacy-policy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true }, - "training": false + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.0000008", - "completion": "0.0000032", - "image": "0.001024", + "prompt": "0", + "completion": "0", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -57441,63 +60447,37 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 206, - "newest": 123, - "throughputHighToLow": 251, - "latencyLowToHigh": 138, - "pricingLowToHigh": 217, - "pricingHighToLow": 100 + "topWeekly": 208, + "newest": 71, + "throughputHighToLow": 63, + "latencyLowToHigh": 267, + "pricingLowToHigh": 30, + "pricingHighToLow": 278 }, - "providers": [ - { - "name": "Alibaba", - "slug": "alibaba", - "quantization": null, - "context": 7500, - "maxCompletionTokens": 1500, - "providerModelId": "qwen-vl-max", - "pricing": { - "prompt": "0.0000008", - "completion": "0.0000032", - "image": "0.001024", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "seed", - "response_format", - "presence_penalty" - ], - "inputCost": 0.8, - "outputCost": 3.2 - } - ] + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", + "providers": [] }, { - "slug": "cognitivecomputations/dolphin3.0-mistral-24b", - "hfSlug": "cognitivecomputations/Dolphin3.0-Mistral-24B", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-13T15:53:39.941479+00:00", + "slug": "openai/gpt-4o-mini-search-preview", + "hfSlug": "", + "updatedAt": "2025-04-22T22:34:08.214467+00:00", + "createdAt": "2025-03-12T22:22:02.718344+00:00", "hfUpdatedAt": null, - "name": "Dolphin3.0 Mistral 24B (free)", - "shortName": "Dolphin3.0 Mistral 24B (free)", - "author": "cognitivecomputations", - "description": "Dolphin 3.0 is the next generation of the Dolphin series of instruct-tuned models. Designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.\n\nDolphin aims to be a general purpose instruct model, similar to the models behind ChatGPT, Claude, Gemini. \n\nPart of the [Dolphin 3.0 Collection](https://huggingface.co/collections/cognitivecomputations/dolphin-30-677ab47f73d7ff66743979a3) Curated and trained by [Eric Hartford](https://huggingface.co/ehartford), [Ben Gitter](https://huggingface.co/bigstorm), [BlouseJury](https://huggingface.co/BlouseJury) and [Cognitive Computations](https://huggingface.co/cognitivecomputations)", + "name": "OpenAI: GPT-4o-mini Search Preview", + "shortName": "GPT-4o-mini Search Preview", + "author": "openai", + "description": "GPT-4o mini Search Preview is a specialized model for web search in Chat Completions. It is trained to understand and execute web search queries.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -57505,32 +60485,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cognitivecomputations/dolphin3.0-mistral-24b", + "permaslug": "openai/gpt-4o-mini-search-preview-2025-03-11", "reasoningConfig": null, "features": {}, - "endpoint": { - "id": "4c7841c6-0d75-49bc-be70-478c2ae4f12b", - "name": "Chutes | cognitivecomputations/dolphin3.0-mistral-24b:free", - "contextLength": 32768, + "endpoint": { + "id": "5154b382-e458-4539-bf6d-cbadfbaa0600", + "name": "OpenAI | openai/gpt-4o-mini-search-preview-2025-03-11", + "contextLength": 128000, "model": { - "slug": "cognitivecomputations/dolphin3.0-mistral-24b", - "hfSlug": "cognitivecomputations/Dolphin3.0-Mistral-24B", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-13T15:53:39.941479+00:00", + "slug": "openai/gpt-4o-mini-search-preview", + "hfSlug": "", + "updatedAt": "2025-04-22T22:34:08.214467+00:00", + "createdAt": "2025-03-12T22:22:02.718344+00:00", "hfUpdatedAt": null, - "name": "Dolphin3.0 Mistral 24B", - "shortName": "Dolphin3.0 Mistral 24B", - "author": "cognitivecomputations", - "description": "Dolphin 3.0 is the next generation of the Dolphin series of instruct-tuned models. Designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.\n\nDolphin aims to be a general purpose instruct model, similar to the models behind ChatGPT, Claude, Gemini. \n\nPart of the [Dolphin 3.0 Collection](https://huggingface.co/collections/cognitivecomputations/dolphin-30-677ab47f73d7ff66743979a3) Curated and trained by [Eric Hartford](https://huggingface.co/ehartford), [Ben Gitter](https://huggingface.co/bigstorm), [BlouseJury](https://huggingface.co/BlouseJury) and [Cognitive Computations](https://huggingface.co/cognitivecomputations)", + "name": "OpenAI: GPT-4o-mini Search Preview", + "shortName": "GPT-4o-mini Search Preview", + "author": "openai", + "description": "GPT-4o mini Search Preview is a specialized model for web search in Chat Completions. It is trained to understand and execute web search queries.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -57538,94 +60518,112 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cognitivecomputations/dolphin3.0-mistral-24b", + "permaslug": "openai/gpt-4o-mini-search-preview-2025-03-11", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "cognitivecomputations/dolphin3.0-mistral-24b:free", - "modelVariantPermaslug": "cognitivecomputations/dolphin3.0-mistral-24b:free", - "providerName": "Chutes", + "modelVariantSlug": "openai/gpt-4o-mini-search-preview", + "modelVariantPermaslug": "openai/gpt-4o-mini-search-preview-2025-03-11", + "providerName": "OpenAI", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { - "training": true, - "retainsPrompts": true - } + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": false, - "group": "Chutes", + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Chutes", - "providerModelId": "cognitivecomputations/Dolphin3.0-Mistral-24B", - "providerGroup": "Chutes", + "providerDisplayName": "OpenAI", + "providerModelId": "gpt-4o-mini-search-preview-2025-03-11", + "providerGroup": "OpenAI", "quantization": null, - "variant": "free", - "isFree": true, + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 16384, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "web_search_options", "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" + "response_format", + "structured_outputs" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "training": true, - "retainsPrompts": true + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0", - "completion": "0", - "image": "0", - "request": "0", + "prompt": "0.00000015", + "completion": "0.0000006", + "image": "0.000217", + "request": "0.0275", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [], + "variablePricings": [ + { + "type": "search-threshold", + "threshold": "high", + "request": "0.03" + }, + { + "type": "search-threshold", + "threshold": "medium", + "request": "0.0275" + }, + { + "type": "search-threshold", + "threshold": "low", + "request": "0.025" + } + ], "isHidden": false, "isDeranked": false, "isDisabled": false, @@ -57642,156 +60640,184 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 207, - "newest": 114, - "throughputHighToLow": 116, - "latencyLowToHigh": 227, - "pricingLowToHigh": 47, - "pricingHighToLow": 295 + "topWeekly": 209, + "newest": 90, + "throughputHighToLow": 14, + "latencyLowToHigh": 286, + "pricingLowToHigh": 311, + "pricingHighToLow": 7 }, - "providers": [] + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", + "providers": [ + { + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": null, + "context": 128000, + "maxCompletionTokens": 16384, + "providerModelId": "gpt-4o-mini-search-preview-2025-03-11", + "pricing": { + "prompt": "0.00000015", + "completion": "0.0000006", + "image": "0.000217", + "request": "0.0275", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "web_search_options", + "max_tokens", + "response_format", + "structured_outputs" + ], + "inputCost": 0.15, + "outputCost": 0.6, + "throughput": 187.4315, + "latency": 3358 + } + ] }, { - "slug": "anthropic/claude-3-opus", + "slug": "cohere/command-r", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-03-05T00:00:00+00:00", + "createdAt": "2024-03-14T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3 Opus (self-moderated)", - "shortName": "Claude 3 Opus (self-moderated)", - "author": "anthropic", - "description": "Claude 3 Opus is Anthropic's most powerful model for highly complex tasks. It boasts top-level performance, intelligence, fluency, and understanding.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal", + "name": "Cohere: Command R", + "shortName": "Command R", + "author": "cohere", + "description": "Command-R is a 35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents.\n\nRead the launch post [here](https://txt.cohere.com/command-r/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", "modelVersionGroupId": null, - "contextLength": 200000, + "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", + "group": "Cohere", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-opus", + "permaslug": "cohere/command-r", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "ee74a4e0-2863-4f51-99a5-997c31c48ae7", - "name": "Anthropic | anthropic/claude-3-opus:beta", - "contextLength": 200000, + "id": "520aa031-d968-4fa1-9648-70c2d4a23a0a", + "name": "Cohere | cohere/command-r", + "contextLength": 128000, "model": { - "slug": "anthropic/claude-3-opus", + "slug": "cohere/command-r", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-03-05T00:00:00+00:00", + "createdAt": "2024-03-14T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude 3 Opus", - "shortName": "Claude 3 Opus", - "author": "anthropic", - "description": "Claude 3 Opus is Anthropic's most powerful model for highly complex tasks. It boasts top-level performance, intelligence, fluency, and understanding.\n\nSee the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)\n\n#multimodal", + "name": "Cohere: Command R", + "shortName": "Command R", + "author": "cohere", + "description": "Command-R is a 35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents.\n\nRead the launch post [here](https://txt.cohere.com/command-r/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", "modelVersionGroupId": null, - "contextLength": 200000, + "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", + "group": "Cohere", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-3-opus", + "permaslug": "cohere/command-r", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "anthropic/claude-3-opus:beta", - "modelVariantPermaslug": "anthropic/claude-3-opus:beta", - "providerName": "Anthropic", + "modelVariantSlug": "cohere/command-r", + "modelVariantPermaslug": "cohere/command-r", + "providerName": "Cohere", "providerInfo": { - "name": "Anthropic", - "displayName": "Anthropic", - "slug": "anthropic", + "name": "Cohere", + "displayName": "Cohere", + "slug": "cohere", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "privacyPolicyUrl": "https://cohere.com/privacy", + "termsOfServiceUrl": "https://cohere.com/terms-of-use", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 - }, - "requiresUserIds": true + } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, + "hasCompletions": false, "isAbortable": true, - "moderationRequired": true, - "group": "Anthropic", + "moderationRequired": false, + "group": "Cohere", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.anthropic.com/", + "statusPageUrl": "https://status.cohere.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Anthropic.svg" + "url": "/images/icons/Cohere.png" } }, - "providerDisplayName": "Anthropic", - "providerModelId": "claude-3-opus-20240229", - "providerGroup": "Anthropic", + "providerDisplayName": "Cohere", + "providerModelId": "command-r", + "providerGroup": "Cohere", "quantization": "unknown", - "variant": "beta", + "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": 4000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "top_k", - "stop" + "seed", + "response_format", + "structured_outputs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "privacyPolicyUrl": "https://cohere.com/privacy", + "termsOfServiceUrl": "https://cohere.com/terms-of-use", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 }, - "requiresUserIds": true, "training": false, "retainsPrompts": true, "retentionDays": 30 }, "pricing": { - "prompt": "0.000015", - "completion": "0.000075", - "image": "0.024", + "prompt": "0.0000005", + "completion": "0.0000015", + "image": "0", "request": "0", - "inputCacheRead": "0.0000015", - "inputCacheWrite": "0.00001875", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -57805,96 +60831,73 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 138, - "newest": 281, - "throughputHighToLow": 249, - "latencyLowToHigh": 216, - "pricingLowToHigh": 309, - "pricingHighToLow": 8 + "topWeekly": 210, + "newest": 276, + "throughputHighToLow": 44, + "latencyLowToHigh": 3, + "pricingLowToHigh": 188, + "pricingHighToLow": 127 }, + "authorIcon": "https://openrouter.ai/images/icons/Cohere.png", "providers": [ { - "name": "Anthropic", - "slug": "anthropic", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 4096, - "providerModelId": "claude-3-opus-20240229", - "pricing": { - "prompt": "0.000015", - "completion": "0.000075", - "image": "0.024", - "request": "0", - "inputCacheRead": "0.0000015", - "inputCacheWrite": "0.00001875", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "top_k", - "stop" - ], - "inputCost": 15, - "outputCost": 75 - }, - { - "name": "Google Vertex", - "slug": "vertex", + "name": "Cohere", + "icon": "https://openrouter.ai/images/icons/Cohere.png", + "slug": "cohere", "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 4096, - "providerModelId": "claude-3-opus@20240229", + "context": 128000, + "maxCompletionTokens": 4000, + "providerModelId": "command-r", "pricing": { - "prompt": "0.000015", - "completion": "0.000075", - "image": "0.024", + "prompt": "0.0000005", + "completion": "0.0000015", + "image": "0", "request": "0", - "inputCacheRead": "0.0000015", - "inputCacheWrite": "0.00001875", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "top_k", - "stop" + "seed", + "response_format", + "structured_outputs" ], - "inputCost": 15, - "outputCost": 75 + "inputCost": 0.5, + "outputCost": 1.5, + "throughput": 139.458, + "latency": 175 } ] }, { - "slug": "microsoft/phi-3-mini-128k-instruct", - "hfSlug": "microsoft/Phi-3-mini-128k-instruct", + "slug": "nothingiisreal/mn-celeste-12b", + "hfSlug": "nothingiisreal/MN-12B-Celeste-V1.9", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-05-26T00:00:00+00:00", + "createdAt": "2024-08-02T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Microsoft: Phi-3 Mini 128K Instruct", - "shortName": "Phi-3 Mini 128K Instruct", - "author": "microsoft", - "description": "Phi-3 Mini is a powerful 3.8B parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. This model is static, trained on an offline dataset with an October 2023 cutoff date.", + "name": "Mistral Nemo 12B Celeste", + "shortName": "Mistral Nemo 12B Celeste", + "author": "nothingiisreal", + "description": "A specialized story writing and roleplaying model based on Mistral's NeMo 12B Instruct. Fine-tuned on curated datasets including Reddit Writing Prompts and Opus Instruct 25K.\n\nThis model excels at creative writing, offering improved NSFW capabilities, with smarter and more active narration. It demonstrates remarkable versatility in both SFW and NSFW scenarios, with strong Out of Character (OOC) steering capabilities, allowing fine-tuned control over narrative direction and character behavior.\n\nCheck out the model's [HuggingFace page](https://huggingface.co/nothingiisreal/MN-12B-Celeste-V1.9) for details on what parameters and prompts work best!", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 16384, "inputModalities": [ "text" ], @@ -57902,35 +60905,36 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "phi3", + "group": "Mistral", + "instructType": "chatml", "defaultSystem": null, "defaultStops": [ - "<|end|>", + "<|im_start|>", + "<|im_end|>", "<|endoftext|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "microsoft/phi-3-mini-128k-instruct", + "permaslug": "nothingiisreal/mn-celeste-12b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "b8da4002-63a2-416a-b5f5-b92d2be049dd", - "name": "Azure | microsoft/phi-3-mini-128k-instruct", - "contextLength": 128000, + "id": "294ab81c-7834-4c4d-8309-302582d5c5b8", + "name": "Featherless | nothingiisreal/mn-celeste-12b", + "contextLength": 16384, "model": { - "slug": "microsoft/phi-3-mini-128k-instruct", - "hfSlug": "microsoft/Phi-3-mini-128k-instruct", + "slug": "nothingiisreal/mn-celeste-12b", + "hfSlug": "nothingiisreal/MN-12B-Celeste-V1.9", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-05-26T00:00:00+00:00", + "createdAt": "2024-08-02T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Microsoft: Phi-3 Mini 128K Instruct", - "shortName": "Phi-3 Mini 128K Instruct", - "author": "microsoft", - "description": "Phi-3 Mini is a powerful 3.8B parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. This model is static, trained on an offline dataset with an October 2023 cutoff date.", + "name": "Mistral Nemo 12B Celeste", + "shortName": "Mistral Nemo 12B Celeste", + "author": "nothingiisreal", + "description": "A specialized story writing and roleplaying model based on Mistral's NeMo 12B Instruct. Fine-tuned on curated datasets including Reddit Writing Prompts and Opus Instruct 25K.\n\nThis model excels at creative writing, offering improved NSFW capabilities, with smarter and more active narration. It demonstrates remarkable versatility in both SFW and NSFW scenarios, with strong Out of Character (OOC) steering capabilities, allowing fine-tuned control over narrative direction and character behavior.\n\nCheck out the model's [HuggingFace page](https://huggingface.co/nothingiisreal/MN-12B-Celeste-V1.9) for details on what parameters and prompts work best!", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 32000, "inputModalities": [ "text" ], @@ -57938,32 +60942,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "phi3", + "group": "Mistral", + "instructType": "chatml", "defaultSystem": null, "defaultStops": [ - "<|end|>", + "<|im_start|>", + "<|im_end|>", "<|endoftext|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "microsoft/phi-3-mini-128k-instruct", + "permaslug": "nothingiisreal/mn-celeste-12b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "microsoft/phi-3-mini-128k-instruct", - "modelVariantPermaslug": "microsoft/phi-3-mini-128k-instruct", - "providerName": "Azure", + "modelVariantSlug": "nothingiisreal/mn-celeste-12b", + "modelVariantPermaslug": "nothingiisreal/mn-celeste-12b", + "providerName": "Featherless", "providerInfo": { - "name": "Azure", - "displayName": "Azure", - "slug": "azure", + "name": "Featherless", + "displayName": "Featherless", + "slug": "featherless", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.microsoft.com/en-us/legal/terms-of-use?oneroute=true", - "privacyPolicyUrl": "https://www.microsoft.com/en-us/privacy/privacystatement", - "dataPolicyUrl": "https://learn.microsoft.com/en-us/legal/cognitive-services/openai/data-privacy?tabs=azure-portal", + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -57971,44 +60975,48 @@ export default { }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": true, + "hasCompletions": true, + "isAbortable": false, "moderationRequired": false, - "group": "Azure", + "group": "Featherless", "editors": [], "owners": [], "isMultipartSupported": false, - "statusPageUrl": "https://status.azure.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Azure.svg" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" } }, - "providerDisplayName": "Azure", - "providerModelId": "phi3-mini-128k", - "providerGroup": "Azure", - "quantization": "unknown", + "providerDisplayName": "Featherless", + "providerModelId": "nothingiisreal/MN-12B-Celeste-V1.9", + "providerGroup": "Featherless", + "quantization": "fp8", "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", - "top_p" + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "top_k", + "min_p", + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.microsoft.com/en-us/legal/terms-of-use?oneroute=true", - "privacyPolicyUrl": "https://www.microsoft.com/en-us/privacy/privacystatement", - "dataPolicyUrl": "https://learn.microsoft.com/en-us/legal/cognitive-services/openai/data-privacy?tabs=azure-portal", + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -58017,8 +61025,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", + "prompt": "0.0000008", + "completion": "0.0000012", "image": "0", "request": "0", "webSearch": "0", @@ -58029,12 +61037,12 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": false, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -58042,24 +61050,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 209, - "newest": 253, - "throughputHighToLow": 90, - "latencyLowToHigh": 1, - "pricingLowToHigh": 117, - "pricingHighToLow": 205 + "topWeekly": 211, + "newest": 226, + "throughputHighToLow": 284, + "latencyLowToHigh": 200, + "pricingLowToHigh": 208, + "pricingHighToLow": 112 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { - "name": "Azure", - "slug": "azure", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "phi3-mini-128k", + "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", + "slug": "featherless", + "quantization": "fp8", + "context": 16384, + "maxCompletionTokens": 4096, + "providerModelId": "nothingiisreal/MN-12B-Celeste-V1.9", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000001", + "prompt": "0.0000008", + "completion": "0.0000012", "image": "0", "request": "0", "webSearch": "0", @@ -58067,14 +61077,21 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", - "top_p" + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "top_k", + "min_p", + "seed" ], - "inputCost": 0.1, - "outputCost": 0.1 + "inputCost": 0.8, + "outputCost": 1.2, + "throughput": 17.57, + "latency": 1457 } ] }, @@ -58235,16 +61252,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 28, + "topWeekly": 26, "newest": 229, - "throughputHighToLow": 180, - "latencyLowToHigh": 132, + "throughputHighToLow": 147, + "latencyLowToHigh": 172, "pricingLowToHigh": 67, "pricingHighToLow": 242 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", "slug": "inferenceNet", "quantization": "fp16", "context": 16384, @@ -58273,10 +61292,13 @@ export default { "top_logprobs" ], "inputCost": 0.02, - "outputCost": 0.03 + "outputCost": 0.03, + "throughput": 51.254, + "latency": 1218 }, { "name": "DeepInfra (Turbo)", + "icon": "", "slug": "deepInfra (turbo)", "quantization": "fp8", "context": 131072, @@ -58307,10 +61329,13 @@ export default { "min_p" ], "inputCost": 0.02, - "outputCost": 0.03 + "outputCost": 0.03, + "throughput": 101.6575, + "latency": 250 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "fp8", "context": 16384, @@ -58339,10 +61364,13 @@ export default { "logit_bias" ], "inputCost": 0.02, - "outputCost": 0.05 + "outputCost": 0.05, + "throughput": 78.6405, + "latency": 720 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 131072, @@ -58371,10 +61399,13 @@ export default { "top_logprobs" ], "inputCost": 0.02, - "outputCost": 0.06 + "outputCost": 0.06, + "throughput": 54.269, + "latency": 482 }, { "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", "slug": "lambda", "quantization": "bf16", "context": 131072, @@ -58403,10 +61434,13 @@ export default { "response_format" ], "inputCost": 0.02, - "outputCost": 0.04 + "outputCost": 0.04, + "throughput": 162.1145, + "latency": 563 }, { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "bf16", "context": 131072, @@ -58438,10 +61472,13 @@ export default { "min_p" ], "inputCost": 0.03, - "outputCost": 0.05 + "outputCost": 0.05, + "throughput": 51.918, + "latency": 338 }, { "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", "slug": "cloudflare", "quantization": "fp8", "context": 32000, @@ -58469,10 +61506,13 @@ export default { "presence_penalty" ], "inputCost": 0.04, - "outputCost": 0.38 + "outputCost": 0.38, + "throughput": 23.5165, + "latency": 784 }, { "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", "slug": "groq", "quantization": null, "context": 131072, @@ -58503,10 +61543,13 @@ export default { "seed" ], "inputCost": 0.05, - "outputCost": 0.08 + "outputCost": 0.08, + "throughput": 1512.5825, + "latency": 1437 }, { "name": "NextBit", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nextbit256.com/&size=256", "slug": "nextBit", "quantization": "fp8", "context": 131072, @@ -58532,10 +61575,13 @@ export default { "structured_outputs" ], "inputCost": 0.06, - "outputCost": 0.1 + "outputCost": 0.1, + "throughput": 114.779, + "latency": 1597 }, { "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", "slug": "hyperbolic", "quantization": "fp8", "context": 131072, @@ -58566,10 +61612,13 @@ export default { "repetition_penalty" ], "inputCost": 0.1, - "outputCost": 0.1 + "outputCost": 0.1, + "throughput": 268.153, + "latency": 981 }, { "name": "Cerebras", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cerebras.ai/&size=256", "slug": "cerebras", "quantization": "fp16", "context": 32000, @@ -58597,10 +61646,13 @@ export default { "top_logprobs" ], "inputCost": 0.1, - "outputCost": 0.1 + "outputCost": 0.1, + "throughput": 3911.239, + "latency": 188 }, { "name": "Friendli", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://friendli.ai/&size=256", "slug": "friendli", "quantization": null, "context": 131072, @@ -58632,10 +61684,13 @@ export default { "structured_outputs" ], "inputCost": 0.1, - "outputCost": 0.1 + "outputCost": 0.1, + "throughput": 282.859, + "latency": 304 }, { "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", "slug": "sambaNova", "quantization": "bf16", "context": 16384, @@ -58658,10 +61713,13 @@ export default { "stop" ], "inputCost": 0.1, - "outputCost": 0.2 + "outputCost": 0.2, + "throughput": 879.206, + "latency": 329 }, { "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", "slug": "klusterAi", "quantization": "fp8", "context": 131000, @@ -58678,13 +61736,27 @@ export default { }, "supportedParameters": [ "max_tokens", - "temperature" + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", + "seed" ], "inputCost": 0.18, - "outputCost": 0.18 + "outputCost": 0.18, + "throughput": 62.505, + "latency": 551 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "fp8", "context": 131072, @@ -58715,10 +61787,13 @@ export default { "response_format" ], "inputCost": 0.18, - "outputCost": 0.18 + "outputCost": 0.18, + "throughput": 252.455, + "latency": 313 }, { "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", "slug": "fireworks", "quantization": "unknown", "context": 131072, @@ -58749,10 +61824,13 @@ export default { "top_logprobs" ], "inputCost": 0.2, - "outputCost": 0.2 + "outputCost": 0.2, + "throughput": 255.586, + "latency": 361 }, { "name": "Avian.io", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://avian.io/&size=256", "slug": "avianIo", "quantization": "fp8", "context": 131072, @@ -58782,17 +61860,17 @@ export default { ] }, { - "slug": "anthracite-org/magnum-v4-72b", - "hfSlug": "anthracite-org/magnum-v4-72b", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-10-22T00:00:00+00:00", + "slug": "arcee-ai/coder-large", + "hfSlug": "", + "updatedAt": "2025-05-05T21:48:22.361364+00:00", + "createdAt": "2025-05-05T20:57:43.438481+00:00", "hfUpdatedAt": null, - "name": "Magnum v4 72B", - "shortName": "Magnum v4 72B", - "author": "anthracite-org", - "description": "This is a series of models designed to replicate the prose quality of the Claude 3 models, specifically Sonnet(https://openrouter.ai/anthropic/claude-3.5-sonnet) and Opus(https://openrouter.ai/anthropic/claude-3-opus).\n\nThe model is fine-tuned on top of [Qwen2.5 72B](https://openrouter.ai/qwen/qwen-2.5-72b-instruct).", + "name": "Arcee AI: Coder Large", + "shortName": "Coder Large", + "author": "arcee-ai", + "description": "Coder‑Large is a 32 B‑parameter offspring of Qwen 2.5‑Instruct that has been further trained on permissively‑licensed GitHub, CodeSearchNet and synthetic bug‑fix corpora. It supports a 32k context window, enabling multi‑file refactoring or long diff review in a single call, and understands 30‑plus programming languages with special attention to TypeScript, Go and Terraform. Internal benchmarks show 5–8 pt gains over CodeLlama‑34 B‑Python on HumanEval and competitive BugFix scores thanks to a reinforcement pass that rewards compilable output. The model emits structured explanations alongside code blocks by default, making it suitable for educational tooling as well as production copilot scenarios. Cost‑wise, Together AI prices it well below proprietary incumbents, so teams can scale interactive coding without runaway spend. ", "modelVersionGroupId": null, - "contextLength": 16384, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -58800,34 +61878,30 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthracite-org/magnum-v4-72b", + "permaslug": "arcee-ai/coder-large", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "bdcb63ce-ae8d-449d-819b-4a229625421e", - "name": "Mancer | anthracite-org/magnum-v4-72b", - "contextLength": 16384, + "id": "650746f1-a509-49e4-9660-3ecc91c025f5", + "name": "Together | arcee-ai/coder-large", + "contextLength": 32768, "model": { - "slug": "anthracite-org/magnum-v4-72b", - "hfSlug": "anthracite-org/magnum-v4-72b", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-10-22T00:00:00+00:00", + "slug": "arcee-ai/coder-large", + "hfSlug": "", + "updatedAt": "2025-05-05T21:48:22.361364+00:00", + "createdAt": "2025-05-05T20:57:43.438481+00:00", "hfUpdatedAt": null, - "name": "Magnum v4 72B", - "shortName": "Magnum v4 72B", - "author": "anthracite-org", - "description": "This is a series of models designed to replicate the prose quality of the Claude 3 models, specifically Sonnet(https://openrouter.ai/anthropic/claude-3.5-sonnet) and Opus(https://openrouter.ai/anthropic/claude-3-opus).\n\nThe model is fine-tuned on top of [Qwen2.5 72B](https://openrouter.ai/qwen/qwen-2.5-72b-instruct).", + "name": "Arcee AI: Coder Large", + "shortName": "Coder Large", + "author": "arcee-ai", + "description": "Coder‑Large is a 32 B‑parameter offspring of Qwen 2.5‑Instruct that has been further trained on permissively‑licensed GitHub, CodeSearchNet and synthetic bug‑fix corpora. It supports a 32k context window, enabling multi‑file refactoring or long diff review in a single call, and understands 30‑plus programming languages with special attention to TypeScript, Go and Terraform. Internal benchmarks show 5–8 pt gains over CodeLlama‑34 B‑Python on HumanEval and competitive BugFix scores thanks to a reinforcement pass that rewards compilable output. The model emits structured explanations alongside code blocks by default, making it suitable for educational tooling as well as production copilot scenarios. Cost‑wise, Together AI prices it well below proprietary incumbents, so teams can scale interactive coding without runaway spend. ", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ @@ -58837,42 +61911,39 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthracite-org/magnum-v4-72b", + "permaslug": "arcee-ai/coder-large", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "anthracite-org/magnum-v4-72b", - "modelVariantPermaslug": "anthracite-org/magnum-v4-72b", - "providerName": "Mancer", + "modelVariantSlug": "arcee-ai/coder-large", + "modelVariantPermaslug": "arcee-ai/coder-large", + "providerName": "Together", "providerInfo": { - "name": "Mancer", - "displayName": "Mancer", - "slug": "mancer", + "name": "Together", + "displayName": "Together", + "slug": "together", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://mancer.tech/terms", - "privacyPolicyUrl": "https://mancer.tech/privacy", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false } }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Mancer", + "group": "Together", "editors": [], "owners": [], "isMultipartSupported": true, @@ -58880,18 +61951,18 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" } }, - "providerDisplayName": "Mancer", - "providerModelId": "magnum-72b-v4", - "providerGroup": "Mancer", - "quantization": "fp6", + "providerDisplayName": "Together", + "providerModelId": "arcee-ai/coder-large", + "providerGroup": "Together", + "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 1024, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -58901,33 +61972,32 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "top_k", "repetition_penalty", "logit_bias", - "top_k", "min_p", - "seed", - "top_a" + "response_format" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://mancer.tech/terms", - "privacyPolicyUrl": "https://mancer.tech/privacy", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0.0000015", - "completion": "0.00000225", + "prompt": "0.0000005", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", - "discount": 0.25 + "discount": 0 }, "variablePricings": [], "isHidden": false, @@ -58941,95 +62011,32 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 211, - "newest": 181, - "throughputHighToLow": 290, - "latencyLowToHigh": 96, - "pricingLowToHigh": 237, - "pricingHighToLow": 81 + "topWeekly": 213, + "newest": 7, + "throughputHighToLow": 81, + "latencyLowToHigh": 22, + "pricingLowToHigh": 182, + "pricingHighToLow": 134 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://arcee.ai/\\u0026size=256", "providers": [ { - "name": "Mancer", - "slug": "mancer", - "quantization": "fp6", - "context": 16384, - "maxCompletionTokens": 1024, - "providerModelId": "magnum-72b-v4", - "pricing": { - "prompt": "0.0000015", - "completion": "0.00000225", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0.25 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "logit_bias", - "top_k", - "min_p", - "seed", - "top_a" - ], - "inputCost": 1.5, - "outputCost": 2.25 - }, - { - "name": "Mancer (private)", - "slug": "mancer (private)", - "quantization": "fp6", - "context": 16384, - "maxCompletionTokens": 1024, - "providerModelId": "magnum-72b-v4", - "pricing": { - "prompt": "0.000002", - "completion": "0.000003", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "logit_bias", - "top_k", - "min_p", - "seed", - "top_a" - ], - "inputCost": 2, - "outputCost": 3 - }, - { - "name": "Infermatic", - "slug": "infermatic", - "quantization": "fp8", + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": null, "context": 32768, "maxCompletionTokens": null, - "providerModelId": "anthracite-org-magnum-v4-72b-FP8-Dynamic", + "providerModelId": "arcee-ai/coder-large", "pricing": { - "prompt": "0.000003", - "completion": "0.000003", + "prompt": "0.0000005", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -59043,60 +62050,31 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "logit_bias", "top_k", - "min_p", - "seed" - ], - "inputCost": 3, - "outputCost": 3 - }, - { - "name": "Featherless", - "slug": "featherless", - "quantization": "fp8", - "context": 16384, - "maxCompletionTokens": 4096, - "providerModelId": "anthracite-org/magnum-v4-72b", - "pricing": { - "prompt": "0.000004", - "completion": "0.000006", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", "repetition_penalty", - "top_k", + "logit_bias", "min_p", - "seed" + "response_format" ], - "inputCost": 4, - "outputCost": 6 + "inputCost": 0.5, + "outputCost": 0.8, + "throughput": 93.636, + "latency": 284 } ] }, { - "slug": "perplexity/r1-1776", - "hfSlug": "perplexity-ai/r1-1776", + "slug": "thudm/glm-z1-32b", + "hfSlug": "THUDM/GLM-Z1-32B-0414", "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-02-19T22:42:09.214134+00:00", + "createdAt": "2025-04-17T21:09:08.005794+00:00", "hfUpdatedAt": null, - "name": "Perplexity: R1 1776", - "shortName": "R1 1776", - "author": "perplexity", - "description": "R1 1776 is a version of DeepSeek-R1 that has been post-trained to remove censorship constraints related to topics restricted by the Chinese government. The model retains its original reasoning capabilities while providing direct responses to a wider range of queries. R1 1776 is an offline chat model that does not use the perplexity search subsystem.\n\nThe model was tested on a multilingual dataset of over 1,000 examples covering sensitive topics to measure its likelihood of refusal or overly filtered responses. [Evaluation Results](https://cdn-uploads.huggingface.co/production/uploads/675c8332d01f593dc90817f5/GiN2VqC5hawUgAGJ6oHla.png) Its performance on math and reasoning benchmarks remains similar to the base R1 model. [Reasoning Performance](https://cdn-uploads.huggingface.co/production/uploads/675c8332d01f593dc90817f5/n4Z9Byqp2S7sKUvCvI40R.png)\n\nRead more on the [Blog Post](https://perplexity.ai/hub/blog/open-sourcing-r1-1776)", + "name": "THUDM: GLM Z1 32B", + "shortName": "GLM Z1 32B", + "author": "thudm", + "description": "GLM-Z1-32B-0414 is an enhanced reasoning variant of GLM-4-32B, built for deep mathematical, logical, and code-oriented problem solving. It applies extended reinforcement learning—both task-specific and general pairwise preference-based—to improve performance on complex multi-step tasks. Compared to the base GLM-4-32B model, Z1 significantly boosts capabilities in structured reasoning and formal domains.\n\nThe model supports enforced “thinking” steps via prompt engineering and offers improved coherence for long-form outputs. It’s optimized for use in agentic workflows, and includes support for long context (via YaRN), JSON tool calling, and fine-grained sampling configuration for stable inference. Ideal for use cases requiring deliberate, multi-step reasoning or formal derivations.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 32000, "inputModalities": [ "text" ], @@ -59104,7 +62082,7 @@ export default { "text" ], "hasTextOutput": true, - "group": "DeepSeek", + "group": "Other", "instructType": "deepseek-r1", "defaultSystem": null, "defaultStops": [ @@ -59114,7 +62092,7 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "perplexity/r1-1776", + "permaslug": "thudm/glm-z1-32b-0414", "reasoningConfig": { "startToken": "", "endToken": "" @@ -59126,21 +62104,21 @@ export default { } }, "endpoint": { - "id": "62190e52-ae3c-4b5b-913c-4a821a16de0b", - "name": "Perplexity | perplexity/r1-1776", - "contextLength": 128000, + "id": "f32bfaa2-38d1-4b3b-b5b7-568d2df68ca4", + "name": "Novita | thudm/glm-z1-32b-0414", + "contextLength": 32000, "model": { - "slug": "perplexity/r1-1776", - "hfSlug": "perplexity-ai/r1-1776", + "slug": "thudm/glm-z1-32b", + "hfSlug": "THUDM/GLM-Z1-32B-0414", "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-02-19T22:42:09.214134+00:00", + "createdAt": "2025-04-17T21:09:08.005794+00:00", "hfUpdatedAt": null, - "name": "Perplexity: R1 1776", - "shortName": "R1 1776", - "author": "perplexity", - "description": "R1 1776 is a version of DeepSeek-R1 that has been post-trained to remove censorship constraints related to topics restricted by the Chinese government. The model retains its original reasoning capabilities while providing direct responses to a wider range of queries. R1 1776 is an offline chat model that does not use the perplexity search subsystem.\n\nThe model was tested on a multilingual dataset of over 1,000 examples covering sensitive topics to measure its likelihood of refusal or overly filtered responses. [Evaluation Results](https://cdn-uploads.huggingface.co/production/uploads/675c8332d01f593dc90817f5/GiN2VqC5hawUgAGJ6oHla.png) Its performance on math and reasoning benchmarks remains similar to the base R1 model. [Reasoning Performance](https://cdn-uploads.huggingface.co/production/uploads/675c8332d01f593dc90817f5/n4Z9Byqp2S7sKUvCvI40R.png)\n\nRead more on the [Blog Post](https://perplexity.ai/hub/blog/open-sourcing-r1-1776)", + "name": "THUDM: GLM Z1 32B", + "shortName": "GLM Z1 32B", + "author": "thudm", + "description": "GLM-Z1-32B-0414 is an enhanced reasoning variant of GLM-4-32B, built for deep mathematical, logical, and code-oriented problem solving. It applies extended reinforcement learning—both task-specific and general pairwise preference-based—to improve performance on complex multi-step tasks. Compared to the base GLM-4-32B model, Z1 significantly boosts capabilities in structured reasoning and formal domains.\n\nThe model supports enforced “thinking” steps via prompt engineering and offers improved coherence for long-form outputs. It’s optimized for use in agentic workflows, and includes support for long context (via YaRN), JSON tool calling, and fine-grained sampling configuration for stable inference. Ideal for use cases requiring deliberate, multi-step reasoning or formal derivations.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -59148,7 +62126,7 @@ export default { "text" ], "hasTextOutput": true, - "group": "DeepSeek", + "group": "Other", "instructType": "deepseek-r1", "defaultSystem": null, "defaultStops": [ @@ -59158,7 +62136,7 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "perplexity/r1-1776", + "permaslug": "thudm/glm-z1-32b-0414", "reasoningConfig": { "startToken": "", "endToken": "" @@ -59170,44 +62148,45 @@ export default { } } }, - "modelVariantSlug": "perplexity/r1-1776", - "modelVariantPermaslug": "perplexity/r1-1776", - "providerName": "Perplexity", + "modelVariantSlug": "thudm/glm-z1-32b", + "modelVariantPermaslug": "thudm/glm-z1-32b-0414", + "providerName": "Novita", "providerInfo": { - "name": "Perplexity", - "displayName": "Perplexity", - "slug": "perplexity", + "name": "Novita", + "displayName": "NovitaAI", + "slug": "novita", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", - "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { "training": false } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, + "hasCompletions": true, + "isAbortable": true, "moderationRequired": false, - "group": "Perplexity", + "group": "Novita", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.perplexity.ai/", + "statusPageUrl": "https://status.novita.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Perplexity.svg" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "invertRequired": true } }, - "providerDisplayName": "Perplexity", - "providerModelId": "r1-1776", - "providerGroup": "Perplexity", + "providerDisplayName": "NovitaAI", + "providerModelId": "thudm/glm-z1-32b-0414", + "providerGroup": "Novita", "quantization": null, "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, "maxPromptImages": null, @@ -59218,23 +62197,28 @@ export default { "top_p", "reasoning", "include_reasoning", - "top_k", + "stop", "frequency_penalty", - "presence_penalty" + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", - "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { "training": false }, "training": false }, "pricing": { - "prompt": "0.000002", - "completion": "0.000008", + "prompt": "0.00000024", + "completion": "0.00000024", "image": "0", "request": "0", "webSearch": "0", @@ -59250,61 +62234,35 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 212, - "newest": 111, - "throughputHighToLow": 171, - "latencyLowToHigh": 174, - "pricingLowToHigh": 251, - "pricingHighToLow": 71 + "topWeekly": 214, + "newest": 37, + "throughputHighToLow": 181, + "latencyLowToHigh": 215, + "pricingLowToHigh": 18, + "pricingHighToLow": 161 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { - "name": "Perplexity", - "slug": "perplexity", - "quantization": null, - "context": 128000, - "maxCompletionTokens": null, - "providerModelId": "r1-1776", - "pricing": { - "prompt": "0.000002", - "completion": "0.000008", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "top_k", - "frequency_penalty", - "presence_penalty" - ], - "inputCost": 2, - "outputCost": 8 - }, - { - "name": "Together", - "slug": "together", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", "quantization": null, - "context": 163840, + "context": 32000, "maxCompletionTokens": null, - "providerModelId": "perplexity-ai/r1-1776", + "providerModelId": "thudm/glm-z1-32b-0414", "pricing": { - "prompt": "0.000003", - "completion": "0.000007", + "prompt": "0.00000024", + "completion": "0.00000024", "image": "0", "request": "0", "webSearch": "0", @@ -59320,29 +62278,31 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "seed", "top_k", - "repetition_penalty", - "logit_bias", "min_p", - "response_format" + "repetition_penalty", + "logit_bias" ], - "inputCost": 3, - "outputCost": 7 + "inputCost": 0.24, + "outputCost": 0.24, + "throughput": 21.073, + "latency": 1632 } ] }, { - "slug": "openai/gpt-4o-mini-search-preview", - "hfSlug": "", - "updatedAt": "2025-04-22T22:34:08.214467+00:00", - "createdAt": "2025-03-12T22:22:02.718344+00:00", + "slug": "thudm/glm-4-32b", + "hfSlug": "THUDM/GLM-4-32B-0414", + "updatedAt": "2025-04-17T21:06:30.647205+00:00", + "createdAt": "2025-04-17T20:15:15.766619+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4o-mini Search Preview", - "shortName": "GPT-4o-mini Search Preview", - "author": "openai", - "description": "GPT-4o mini Search Preview is a specialized model for web search in Chat Completions. It is trained to understand and execute web search queries.", + "name": "THUDM: GLM 4 32B", + "shortName": "GLM 4 32B", + "author": "thudm", + "description": "GLM-4-32B-0414 is a 32B bilingual (Chinese-English) open-weight language model optimized for code generation, function calling, and agent-style tasks. Pretrained on 15T of high-quality and reasoning-heavy data, it was further refined using human preference alignment, rejection sampling, and reinforcement learning. The model excels in complex reasoning, artifact generation, and structured output tasks, achieving performance comparable to GPT-4o and DeepSeek-V3-0324 across several benchmarks.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 32000, "inputModalities": [ "text" ], @@ -59350,32 +62310,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-4o-mini-search-preview-2025-03-11", + "permaslug": "thudm/glm-4-32b-0414", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "5154b382-e458-4539-bf6d-cbadfbaa0600", - "name": "OpenAI | openai/gpt-4o-mini-search-preview-2025-03-11", - "contextLength": 128000, + "id": "fa10cafb-307f-48a7-9b4a-3291cde14225", + "name": "Novita | thudm/glm-4-32b-0414", + "contextLength": 32000, "model": { - "slug": "openai/gpt-4o-mini-search-preview", - "hfSlug": "", - "updatedAt": "2025-04-22T22:34:08.214467+00:00", - "createdAt": "2025-03-12T22:22:02.718344+00:00", + "slug": "thudm/glm-4-32b", + "hfSlug": "THUDM/GLM-4-32B-0414", + "updatedAt": "2025-04-17T21:06:30.647205+00:00", + "createdAt": "2025-04-17T20:15:15.766619+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4o-mini Search Preview", - "shortName": "GPT-4o-mini Search Preview", - "author": "openai", - "description": "GPT-4o mini Search Preview is a specialized model for web search in Chat Completions. It is trained to understand and execute web search queries.", + "name": "THUDM: GLM 4 32B", + "shortName": "GLM 4 32B", + "author": "thudm", + "description": "GLM-4-32B-0414 is a 32B bilingual (Chinese-English) open-weight language model optimized for code generation, function calling, and agent-style tasks. Pretrained on 15T of high-quality and reasoning-heavy data, it was further refined using human preference alignment, rejection sampling, and reinforcement learning. The model excels in complex reasoning, artifact generation, and structured output tasks, achieving performance comparable to GPT-4o and DeepSeek-V3-0324 across several benchmarks.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -59383,112 +62343,93 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-4o-mini-search-preview-2025-03-11", + "permaslug": "thudm/glm-4-32b-0414", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/gpt-4o-mini-search-preview", - "modelVariantPermaslug": "openai/gpt-4o-mini-search-preview-2025-03-11", - "providerName": "OpenAI", + "modelVariantSlug": "thudm/glm-4-32b", + "modelVariantPermaslug": "thudm/glm-4-32b-0414", + "providerName": "Novita", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Novita", + "displayName": "NovitaAI", + "slug": "novita", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "training": false + } }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", + "moderationRequired": false, + "group": "Novita", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": "https://status.novita.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "invertRequired": true } }, - "providerDisplayName": "OpenAI", - "providerModelId": "gpt-4o-mini-search-preview-2025-03-11", - "providerGroup": "OpenAI", + "providerDisplayName": "NovitaAI", + "providerModelId": "thudm/glm-4-32b-0414", + "providerGroup": "Novita", "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 16384, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "web_search_options", "max_tokens", - "response_format", - "structured_outputs" + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, "pricing": { - "prompt": "0.00000015", - "completion": "0.0000006", - "image": "0.000217", - "request": "0.0275", + "prompt": "0.00000024", + "completion": "0.00000024", + "image": "0", + "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [ - { - "type": "search-threshold", - "threshold": "high", - "request": "0.03" - }, - { - "type": "search-threshold", - "threshold": "medium", - "request": "0.0275" - }, - { - "type": "search-threshold", - "threshold": "low", - "request": "0.025" - } - ], + "variablePricings": [], "isHidden": false, "isDeranked": false, "isDisabled": false, @@ -59500,58 +62441,70 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 213, - "newest": 90, - "throughputHighToLow": 17, - "latencyLowToHigh": 281, - "pricingLowToHigh": 311, - "pricingHighToLow": 7 + "topWeekly": 128, + "newest": 39, + "throughputHighToLow": 192, + "latencyLowToHigh": 112, + "pricingLowToHigh": 19, + "pricingHighToLow": 162 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { - "name": "OpenAI", - "slug": "openAi", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", "quantization": null, - "context": 128000, - "maxCompletionTokens": 16384, - "providerModelId": "gpt-4o-mini-search-preview-2025-03-11", + "context": 32000, + "maxCompletionTokens": null, + "providerModelId": "thudm/glm-4-32b-0414", "pricing": { - "prompt": "0.00000015", - "completion": "0.0000006", - "image": "0.000217", - "request": "0.0275", + "prompt": "0.00000024", + "completion": "0.00000024", + "image": "0", + "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "web_search_options", "max_tokens", - "response_format", - "structured_outputs" + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" ], - "inputCost": 0.15, - "outputCost": 0.6 + "inputCost": 0.24, + "outputCost": 0.24, + "throughput": 22.1315, + "latency": 1098 } ] }, { - "slug": "cohere/command-r", - "hfSlug": null, + "slug": "anthracite-org/magnum-v4-72b", + "hfSlug": "anthracite-org/magnum-v4-72b", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-03-14T00:00:00+00:00", + "createdAt": "2024-10-22T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Cohere: Command R", - "shortName": "Command R", - "author": "cohere", - "description": "Command-R is a 35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents.\n\nRead the launch post [here](https://txt.cohere.com/command-r/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", + "name": "Magnum v4 72B", + "shortName": "Magnum v4 72B", + "author": "anthracite-org", + "description": "This is a series of models designed to replicate the prose quality of the Claude 3 models, specifically Sonnet(https://openrouter.ai/anthropic/claude-3.5-sonnet) and Opus(https://openrouter.ai/anthropic/claude-3-opus).\n\nThe model is fine-tuned on top of [Qwen2.5 72B](https://openrouter.ai/qwen/qwen-2.5-72b-instruct).", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 16384, "inputModalities": [ "text" ], @@ -59559,32 +62512,36 @@ export default { "text" ], "hasTextOutput": true, - "group": "Cohere", - "instructType": null, + "group": "Qwen", + "instructType": "chatml", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cohere/command-r", + "permaslug": "anthracite-org/magnum-v4-72b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "520aa031-d968-4fa1-9648-70c2d4a23a0a", - "name": "Cohere | cohere/command-r", - "contextLength": 128000, + "id": "bdcb63ce-ae8d-449d-819b-4a229625421e", + "name": "Mancer | anthracite-org/magnum-v4-72b", + "contextLength": 16384, "model": { - "slug": "cohere/command-r", - "hfSlug": null, + "slug": "anthracite-org/magnum-v4-72b", + "hfSlug": "anthracite-org/magnum-v4-72b", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-03-14T00:00:00+00:00", + "createdAt": "2024-10-22T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Cohere: Command R", - "shortName": "Command R", - "author": "cohere", - "description": "Command-R is a 35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents.\n\nRead the launch post [here](https://txt.cohere.com/command-r/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", + "name": "Magnum v4 72B", + "shortName": "Magnum v4 72B", + "author": "anthracite-org", + "description": "This is a series of models designed to replicate the prose quality of the Claude 3 models, specifically Sonnet(https://openrouter.ai/anthropic/claude-3.5-sonnet) and Opus(https://openrouter.ai/anthropic/claude-3-opus).\n\nThe model is fine-tuned on top of [Qwen2.5 72B](https://openrouter.ai/qwen/qwen-2.5-72b-instruct).", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -59592,133 +62549,242 @@ export default { "text" ], "hasTextOutput": true, - "group": "Cohere", - "instructType": null, + "group": "Qwen", + "instructType": "chatml", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cohere/command-r", + "permaslug": "anthracite-org/magnum-v4-72b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "cohere/command-r", - "modelVariantPermaslug": "cohere/command-r", - "providerName": "Cohere", + "modelVariantSlug": "anthracite-org/magnum-v4-72b", + "modelVariantPermaslug": "anthracite-org/magnum-v4-72b", + "providerName": "Mancer", "providerInfo": { - "name": "Cohere", - "displayName": "Cohere", - "slug": "cohere", + "name": "Mancer", + "displayName": "Mancer", + "slug": "mancer", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://cohere.com/privacy", - "termsOfServiceUrl": "https://cohere.com/terms-of-use", + "termsOfServiceUrl": "https://mancer.tech/terms", + "privacyPolicyUrl": "https://mancer.tech/privacy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true } }, - "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, + "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Cohere", + "group": "Mancer", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.cohere.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Cohere.png" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256" } }, - "providerDisplayName": "Cohere", - "providerModelId": "command-r", - "providerGroup": "Cohere", - "quantization": "unknown", + "providerDisplayName": "Mancer", + "providerModelId": "magnum-72b-v4", + "providerGroup": "Mancer", + "quantization": "fp6", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4000, + "maxCompletionTokens": 1024, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", + "repetition_penalty", + "logit_bias", "top_k", + "min_p", "seed", - "response_format", - "structured_outputs" + "top_a" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://cohere.com/privacy", - "termsOfServiceUrl": "https://cohere.com/terms-of-use", + "termsOfServiceUrl": "https://mancer.tech/terms", + "privacyPolicyUrl": "https://mancer.tech/privacy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.0000005", - "completion": "0.0000015", + "prompt": "0.000001875", + "completion": "0.00000225", "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", - "discount": 0 + "discount": 0.25 }, "variablePricings": [], "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 214, - "newest": 276, - "throughputHighToLow": 47, - "latencyLowToHigh": 7, - "pricingLowToHigh": 187, - "pricingHighToLow": 128 + "topWeekly": 216, + "newest": 182, + "throughputHighToLow": 290, + "latencyLowToHigh": 80, + "pricingLowToHigh": 238, + "pricingHighToLow": 80 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { - "name": "Cohere", - "slug": "cohere", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 4000, - "providerModelId": "command-r", + "name": "Mancer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", + "slug": "mancer", + "quantization": "fp6", + "context": 16384, + "maxCompletionTokens": 1024, + "providerModelId": "magnum-72b-v4", + "pricing": { + "prompt": "0.000001875", + "completion": "0.00000225", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0.25 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "logit_bias", + "top_k", + "min_p", + "seed", + "top_a" + ], + "inputCost": 1.88, + "outputCost": 2.25, + "throughput": 16.3, + "latency": 587 + }, + { + "name": "Mancer (private)", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", + "slug": "mancer (private)", + "quantization": "fp6", + "context": 16384, + "maxCompletionTokens": 1024, + "providerModelId": "magnum-72b-v4", + "pricing": { + "prompt": "0.0000025", + "completion": "0.000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "logit_bias", + "top_k", + "min_p", + "seed", + "top_a" + ], + "inputCost": 2.5, + "outputCost": 3, + "throughput": 16.3245, + "latency": 603 + }, + { + "name": "Infermatic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://infermatic.ai/&size=256", + "slug": "infermatic", + "quantization": "fp8", + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "anthracite-org-magnum-v4-72b-FP8-Dynamic", + "pricing": { + "prompt": "0.000003", + "completion": "0.000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "logit_bias", + "top_k", + "min_p", + "seed" + ], + "inputCost": 3, + "outputCost": 3, + "throughput": 17.816, + "latency": 3142 + }, + { + "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", + "slug": "featherless", + "quantization": "fp8", + "context": 16384, + "maxCompletionTokens": 4096, + "providerModelId": "anthracite-org/magnum-v4-72b", "pricing": { - "prompt": "0.0000005", - "completion": "0.0000015", + "prompt": "0.000004", + "completion": "0.000006", "image": "0", "request": "0", "webSearch": "0", @@ -59726,35 +62792,36 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", + "repetition_penalty", "top_k", - "seed", - "response_format", - "structured_outputs" + "min_p", + "seed" ], - "inputCost": 0.5, - "outputCost": 1.5 + "inputCost": 4, + "outputCost": 6, + "throughput": 13.7545, + "latency": 9657 } ] }, { - "slug": "thudm/glm-4-32b", - "hfSlug": "THUDM/GLM-4-32B-0414", - "updatedAt": "2025-04-17T21:06:30.647205+00:00", - "createdAt": "2025-04-17T20:15:15.766619+00:00", + "slug": "deepseek/deepseek-r1-distill-qwen-14b", + "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-01-29T23:39:00.13687+00:00", "hfUpdatedAt": null, - "name": "THUDM: GLM 4 32B", - "shortName": "GLM 4 32B", - "author": "thudm", - "description": "GLM-4-32B-0414 is a 32B bilingual (Chinese-English) open-weight language model optimized for code generation, function calling, and agent-style tasks. Pretrained on 15T of high-quality and reasoning-heavy data, it was further refined using human preference alignment, rejection sampling, and reinforcement learning. The model excels in complex reasoning, artifact generation, and structured output tasks, achieving performance comparable to GPT-4o and DeepSeek-V3-0324 across several benchmarks.", - "modelVersionGroupId": null, - "contextLength": 32000, + "name": "DeepSeek: R1 Distill Qwen 14B (free)", + "shortName": "R1 Distill Qwen 14B (free)", + "author": "deepseek", + "description": "DeepSeek R1 Distill Qwen 14B is a distilled large language model based on [Qwen 2.5 14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 69.7\n- MATH-500 pass@1: 93.9\n- CodeForces Rating: 1481\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", + "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", + "contextLength": 64000, "inputModalities": [ "text" ], @@ -59762,32 +62829,43 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Qwen", + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "thudm/glm-4-32b-0414", - "reasoningConfig": null, - "features": {}, + "warningMessage": "", + "permaslug": "deepseek/deepseek-r1-distill-qwen-14b", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "fa10cafb-307f-48a7-9b4a-3291cde14225", - "name": "Novita | thudm/glm-4-32b-0414", - "contextLength": 32000, + "id": "d498727c-ad05-4d5c-997f-3abdf71ab15c", + "name": "Chutes | deepseek/deepseek-r1-distill-qwen-14b:free", + "contextLength": 64000, "model": { - "slug": "thudm/glm-4-32b", - "hfSlug": "THUDM/GLM-4-32B-0414", - "updatedAt": "2025-04-17T21:06:30.647205+00:00", - "createdAt": "2025-04-17T20:15:15.766619+00:00", + "slug": "deepseek/deepseek-r1-distill-qwen-14b", + "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-01-29T23:39:00.13687+00:00", "hfUpdatedAt": null, - "name": "THUDM: GLM 4 32B", - "shortName": "GLM 4 32B", - "author": "thudm", - "description": "GLM-4-32B-0414 is a 32B bilingual (Chinese-English) open-weight language model optimized for code generation, function calling, and agent-style tasks. Pretrained on 15T of high-quality and reasoning-heavy data, it was further refined using human preference alignment, rejection sampling, and reinforcement learning. The model excels in complex reasoning, artifact generation, and structured output tasks, achieving performance comparable to GPT-4o and DeepSeek-V3-0324 across several benchmarks.", - "modelVersionGroupId": null, - "contextLength": 32768, + "name": "DeepSeek: R1 Distill Qwen 14B", + "shortName": "R1 Distill Qwen 14B", + "author": "deepseek", + "description": "DeepSeek R1 Distill Qwen 14B is a distilled large language model based on [Qwen 2.5 14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 69.7\n- MATH-500 pass@1: 93.9\n- CodeForces Rating: 1481\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", + "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", + "contextLength": 131072, "inputModalities": [ "text" ], @@ -59795,55 +62873,64 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Qwen", + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "thudm/glm-4-32b-0414", - "reasoningConfig": null, - "features": {} + "warningMessage": "", + "permaslug": "deepseek/deepseek-r1-distill-qwen-14b", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "thudm/glm-4-32b", - "modelVariantPermaslug": "thudm/glm-4-32b-0414", - "providerName": "Novita", + "modelVariantSlug": "deepseek/deepseek-r1-distill-qwen-14b:free", + "modelVariantPermaslug": "deepseek/deepseek-r1-distill-qwen-14b:free", + "providerName": "Chutes", "providerInfo": { - "name": "Novita", - "displayName": "NovitaAI", - "slug": "novita", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Novita", + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.novita.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "NovitaAI", - "providerModelId": "thudm/glm-4-32b-0414", - "providerGroup": "Novita", - "quantization": null, - "variant": "standard", - "isFree": false, + "providerDisplayName": "Chutes", + "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B", + "providerGroup": "Chutes", + "quantization": "bf16", + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, @@ -59853,6 +62940,8 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -59860,21 +62949,24 @@ export default { "top_k", "min_p", "repetition_penalty", - "logit_bias" + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true }, - "training": false + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.00000024", - "completion": "0.00000024", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -59886,37 +62978,38 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 128, - "newest": 39, - "throughputHighToLow": 200, - "latencyLowToHigh": 184, - "pricingLowToHigh": 19, - "pricingHighToLow": 162 + "topWeekly": 187, + "newest": 135, + "throughputHighToLow": 124, + "latencyLowToHigh": 192, + "pricingLowToHigh": 51, + "pricingHighToLow": 184 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [ { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": null, - "context": 32000, - "maxCompletionTokens": null, - "providerModelId": "thudm/glm-4-32b-0414", + "context": 64000, + "maxCompletionTokens": 64000, + "providerModelId": "deepseek/deepseek-r1-distill-qwen-14b", "pricing": { - "prompt": "0.00000024", - "completion": "0.00000024", + "prompt": "0.00000015", + "completion": "0.00000015", "image": "0", "request": "0", "webSearch": "0", @@ -59927,6 +63020,8 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -59936,23 +63031,62 @@ export default { "repetition_penalty", "logit_bias" ], - "inputCost": 0.24, - "outputCost": 0.24 + "inputCost": 0.15, + "outputCost": 0.15, + "throughput": 45.768, + "latency": 1328 + }, + { + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 32768, + "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B", + "pricing": { + "prompt": "0.0000016", + "completion": "0.0000016", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" + ], + "inputCost": 1.6, + "outputCost": 1.6, + "throughput": 170.562, + "latency": 362 } ] }, { - "slug": "nothingiisreal/mn-celeste-12b", - "hfSlug": "nothingiisreal/MN-12B-Celeste-V1.9", + "slug": "openai/gpt-3.5-turbo-0125", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-02T00:00:00+00:00", + "createdAt": "2023-05-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Mistral Nemo 12B Celeste", - "shortName": "Mistral Nemo 12B Celeste", - "author": "nothingiisreal", - "description": "A specialized story writing and roleplaying model based on Mistral's NeMo 12B Instruct. Fine-tuned on curated datasets including Reddit Writing Prompts and Opus Instruct 25K.\n\nThis model excels at creative writing, offering improved NSFW capabilities, with smarter and more active narration. It demonstrates remarkable versatility in both SFW and NSFW scenarios, with strong Out of Character (OOC) steering capabilities, allowing fine-tuned control over narrative direction and character behavior.\n\nCheck out the model's [HuggingFace page](https://huggingface.co/nothingiisreal/MN-12B-Celeste-V1.9) for details on what parameters and prompts work best!", + "name": "OpenAI: GPT-3.5 Turbo 16k", + "shortName": "GPT-3.5 Turbo 16k", + "author": "openai", + "description": "The latest GPT-3.5 Turbo model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more. Training data: up to Sep 2021.\n\nThis version has a higher accuracy at responding in requested formats and a fix for a bug which caused a text encoding issue for non-English language function calls.", "modelVersionGroupId": null, - "contextLength": 16384, + "contextLength": 16385, "inputModalities": [ "text" ], @@ -59960,36 +63094,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "chatml", + "group": "GPT", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "nothingiisreal/mn-celeste-12b", + "permaslug": "openai/gpt-3.5-turbo-0125", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "294ab81c-7834-4c4d-8309-302582d5c5b8", - "name": "Featherless | nothingiisreal/mn-celeste-12b", - "contextLength": 16384, + "id": "9be0749c-ce48-4f4a-9e29-b74dbc29d7d6", + "name": "OpenAI | openai/gpt-3.5-turbo-0125", + "contextLength": 16385, "model": { - "slug": "nothingiisreal/mn-celeste-12b", - "hfSlug": "nothingiisreal/MN-12B-Celeste-V1.9", + "slug": "openai/gpt-3.5-turbo-0125", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-02T00:00:00+00:00", + "createdAt": "2023-05-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Mistral Nemo 12B Celeste", - "shortName": "Mistral Nemo 12B Celeste", - "author": "nothingiisreal", - "description": "A specialized story writing and roleplaying model based on Mistral's NeMo 12B Instruct. Fine-tuned on curated datasets including Reddit Writing Prompts and Opus Instruct 25K.\n\nThis model excels at creative writing, offering improved NSFW capabilities, with smarter and more active narration. It demonstrates remarkable versatility in both SFW and NSFW scenarios, with strong Out of Character (OOC) steering capabilities, allowing fine-tuned control over narrative direction and character behavior.\n\nCheck out the model's [HuggingFace page](https://huggingface.co/nothingiisreal/MN-12B-Celeste-V1.9) for details on what parameters and prompts work best!", + "name": "OpenAI: GPT-3.5 Turbo 16k", + "shortName": "GPT-3.5 Turbo 16k", + "author": "openai", + "description": "The latest GPT-3.5 Turbo model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more. Training data: up to Sep 2021.\n\nThis version has a higher accuracy at responding in requested formats and a fix for a bug which caused a text encoding issue for non-English language function calls.", "modelVersionGroupId": null, - "contextLength": 32000, + "contextLength": 16385, "inputModalities": [ "text" ], @@ -59997,91 +63127,98 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "chatml", + "group": "GPT", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "nothingiisreal/mn-celeste-12b", + "permaslug": "openai/gpt-3.5-turbo-0125", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "nothingiisreal/mn-celeste-12b", - "modelVariantPermaslug": "nothingiisreal/mn-celeste-12b", - "providerName": "Featherless", + "modelVariantSlug": "openai/gpt-3.5-turbo-0125", + "modelVariantPermaslug": "openai/gpt-3.5-turbo-0125", + "providerName": "OpenAI", "providerInfo": { - "name": "Featherless", - "displayName": "Featherless", - "slug": "featherless", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false - } + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, - "moderationRequired": false, - "group": "Featherless", + "isAbortable": true, + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], - "isMultipartSupported": false, - "statusPageUrl": null, + "isMultipartSupported": true, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Featherless", - "providerModelId": "nothingiisreal/MN-12B-Celeste-V1.9", - "providerGroup": "Featherless", - "quantization": "fp8", + "providerDisplayName": "OpenAI", + "providerModelId": "gpt-3.5-turbo-0125", + "providerGroup": "OpenAI", + "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "top_k", - "min_p", - "seed" + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, + "requiresUserIds": true, "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.0000008", - "completion": "0.0000012", + "prompt": "0.0000005", + "completion": "0.0000015", "image": "0", "request": "0", "webSearch": "0", @@ -60092,9 +63229,9 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, - "supportsMultipart": false, + "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, @@ -60105,24 +63242,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 216, - "newest": 224, - "throughputHighToLow": 283, - "latencyLowToHigh": 261, - "pricingLowToHigh": 209, - "pricingHighToLow": 111 + "topWeekly": 218, + "newest": 316, + "throughputHighToLow": 50, + "latencyLowToHigh": 41, + "pricingLowToHigh": 191, + "pricingHighToLow": 130 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "Featherless", - "slug": "featherless", - "quantization": "fp8", - "context": 16384, + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": "unknown", + "context": 16385, "maxCompletionTokens": 4096, - "providerModelId": "nothingiisreal/MN-12B-Celeste-V1.9", + "providerModelId": "gpt-3.5-turbo-0125", "pricing": { - "prompt": "0.0000008", - "completion": "0.0000012", + "prompt": "0.0000005", + "completion": "0.0000015", "image": "0", "request": "0", "webSearch": "0", @@ -60130,34 +63269,39 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "top_k", - "min_p", - "seed" + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" ], - "inputCost": 0.8, - "outputCost": 1.2 + "inputCost": 0.5, + "outputCost": 1.5, + "throughput": 130.096, + "latency": 397 } ] }, { - "slug": "deepseek/deepseek-r1-distill-qwen-14b", - "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-01-29T23:39:00.13687+00:00", + "slug": "microsoft/phi-3-mini-128k-instruct", + "hfSlug": "microsoft/Phi-3-mini-128k-instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-05-26T00:00:00+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: R1 Distill Qwen 14B (free)", - "shortName": "R1 Distill Qwen 14B (free)", - "author": "deepseek", - "description": "DeepSeek R1 Distill Qwen 14B is a distilled large language model based on [Qwen 2.5 14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 69.7\n- MATH-500 pass@1: 93.9\n- CodeForces Rating: 1481\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", - "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", - "contextLength": 64000, + "name": "Microsoft: Phi-3 Mini 128K Instruct", + "shortName": "Phi-3 Mini 128K Instruct", + "author": "microsoft", + "description": "Phi-3 Mini is a powerful 3.8B parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. This model is static, trained on an offline dataset with an October 2023 cutoff date.", + "modelVersionGroupId": null, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -60165,43 +63309,35 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "deepseek-r1", + "group": "Other", + "instructType": "phi3", "defaultSystem": null, "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" + "<|end|>", + "<|endoftext|>" ], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "deepseek/deepseek-r1-distill-qwen-14b", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "warningMessage": null, + "permaslug": "microsoft/phi-3-mini-128k-instruct", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "d498727c-ad05-4d5c-997f-3abdf71ab15c", - "name": "Chutes | deepseek/deepseek-r1-distill-qwen-14b:free", - "contextLength": 64000, + "id": "b8da4002-63a2-416a-b5f5-b92d2be049dd", + "name": "Azure | microsoft/phi-3-mini-128k-instruct", + "contextLength": 128000, "model": { - "slug": "deepseek/deepseek-r1-distill-qwen-14b", - "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-01-29T23:39:00.13687+00:00", + "slug": "microsoft/phi-3-mini-128k-instruct", + "hfSlug": "microsoft/Phi-3-mini-128k-instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-05-26T00:00:00+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: R1 Distill Qwen 14B", - "shortName": "R1 Distill Qwen 14B", - "author": "deepseek", - "description": "DeepSeek R1 Distill Qwen 14B is a distilled large language model based on [Qwen 2.5 14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 69.7\n- MATH-500 pass@1: 93.9\n- CodeForces Rating: 1481\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", - "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", - "contextLength": 131072, + "name": "Microsoft: Phi-3 Mini 128K Instruct", + "shortName": "Phi-3 Mini 128K Instruct", + "author": "microsoft", + "description": "Phi-3 Mini is a powerful 3.8B parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. This model is static, trained on an offline dataset with an October 2023 cutoff date.", + "modelVersionGroupId": null, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -60209,100 +63345,87 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "deepseek-r1", + "group": "Other", + "instructType": "phi3", "defaultSystem": null, "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" + "<|end|>", + "<|endoftext|>" ], "hidden": false, "router": null, - "warningMessage": "", - "permaslug": "deepseek/deepseek-r1-distill-qwen-14b", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "warningMessage": null, + "permaslug": "microsoft/phi-3-mini-128k-instruct", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "deepseek/deepseek-r1-distill-qwen-14b:free", - "modelVariantPermaslug": "deepseek/deepseek-r1-distill-qwen-14b:free", - "providerName": "Chutes", + "modelVariantSlug": "microsoft/phi-3-mini-128k-instruct", + "modelVariantPermaslug": "microsoft/phi-3-mini-128k-instruct", + "providerName": "Azure", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "Azure", + "displayName": "Azure", + "slug": "azure", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://www.microsoft.com/en-us/legal/terms-of-use?oneroute=true", + "privacyPolicyUrl": "https://www.microsoft.com/en-us/privacy/privacystatement", + "dataPolicyUrl": "https://learn.microsoft.com/en-us/legal/cognitive-services/openai/data-privacy?tabs=azure-portal", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false } }, + "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, + "hasCompletions": false, "isAbortable": true, "moderationRequired": false, - "group": "Chutes", + "group": "Azure", "editors": [], "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, + "isMultipartSupported": false, + "statusPageUrl": "https://status.azure.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "/images/icons/Azure.svg" } }, - "providerDisplayName": "Chutes", - "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B", - "providerGroup": "Chutes", - "quantization": "bf16", - "variant": "free", - "isFree": true, + "providerDisplayName": "Azure", + "providerModelId": "phi3-mini-128k", + "providerGroup": "Azure", + "quantization": "unknown", + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" + "top_p" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://www.microsoft.com/en-us/legal/terms-of-use?oneroute=true", + "privacyPolicyUrl": "https://www.microsoft.com/en-us/privacy/privacystatement", + "dataPolicyUrl": "https://learn.microsoft.com/en-us/legal/cognitive-services/openai/data-privacy?tabs=azure-portal", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.0000001", + "completion": "0.0000001", "image": "0", "request": "0", "webSearch": "0", @@ -60313,12 +63436,12 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": true, - "supportsMultipart": true, + "supportsToolParameters": true, + "supportsReasoning": false, + "supportsMultipart": false, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -60326,58 +63449,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 176, - "newest": 135, - "throughputHighToLow": 125, - "latencyLowToHigh": 229, - "pricingLowToHigh": 51, - "pricingHighToLow": 184 + "topWeekly": 219, + "newest": 253, + "throughputHighToLow": 78, + "latencyLowToHigh": 0, + "pricingLowToHigh": 117, + "pricingHighToLow": 205 }, + "authorIcon": "https://openrouter.ai/images/icons/Microsoft.svg", "providers": [ { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": null, - "context": 64000, - "maxCompletionTokens": 64000, - "providerModelId": "deepseek/deepseek-r1-distill-qwen-14b", - "pricing": { - "prompt": "0.00000015", - "completion": "0.00000015", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" - ], - "inputCost": 0.15, - "outputCost": 0.15 - }, - { - "name": "Together", - "slug": "together", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 32768, - "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B", + "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", + "slug": "azure", + "quantization": "unknown", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "phi3-mini-128k", "pricing": { - "prompt": "0.0000016", - "completion": "0.0000016", + "prompt": "0.0000001", + "completion": "0.0000001", "image": "0", "request": "0", "webSearch": "0", @@ -60385,22 +63476,16 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "top_p" ], - "inputCost": 1.6, - "outputCost": 1.6 + "inputCost": 0.1, + "outputCost": 0.1, + "throughput": 85.0695, + "latency": 109 } ] }, @@ -60572,16 +63657,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 218, - "newest": 301, - "throughputHighToLow": 24, - "latencyLowToHigh": 12, + "topWeekly": 220, + "newest": 300, + "throughputHighToLow": 21, + "latencyLowToHigh": 10, "pricingLowToHigh": 225, "pricingHighToLow": 99 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", "slug": "openAi", "quantization": "unknown", "context": 16385, @@ -60613,22 +63700,24 @@ export default { "structured_outputs" ], "inputCost": 1, - "outputCost": 2 + "outputCost": 2, + "throughput": 175.824, + "latency": 218 } ] }, { - "slug": "arcee-ai/coder-large", - "hfSlug": "", - "updatedAt": "2025-05-05T21:48:22.361364+00:00", - "createdAt": "2025-05-05T20:57:43.438481+00:00", + "slug": "nousresearch/deephermes-3-llama-3-8b-preview", + "hfSlug": "NousResearch/DeepHermes-3-Llama-3-8B-Preview", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-28T05:09:32.50188+00:00", "hfUpdatedAt": null, - "name": "Arcee AI: Coder Large", - "shortName": "Coder Large", - "author": "arcee-ai", - "description": "Coder‑Large is a 32 B‑parameter offspring of Qwen 2.5‑Instruct that has been further trained on permissively‑licensed GitHub, CodeSearchNet and synthetic bug‑fix corpora. It supports a 32k context window, enabling multi‑file refactoring or long diff review in a single call, and understands 30‑plus programming languages with special attention to TypeScript, Go and Terraform. Internal benchmarks show 5–8 pt gains over CodeLlama‑34 B‑Python on HumanEval and competitive BugFix scores thanks to a reinforcement pass that rewards compilable output. The model emits structured explanations alongside code blocks by default, making it suitable for educational tooling as well as production copilot scenarios. Cost‑wise, Together AI prices it well below proprietary incumbents, so teams can scale interactive coding without runaway spend. ", + "name": "Nous: DeepHermes 3 Llama 3 8B Preview (free)", + "shortName": "DeepHermes 3 Llama 3 8B Preview (free)", + "author": "nousresearch", + "description": "DeepHermes 3 Preview is the latest version of our flagship Hermes series of LLMs by Nous Research, and one of the first models in the world to unify Reasoning (long chains of thought that improve answer accuracy) and normal LLM response modes into one model. We have also improved LLM annotation, judgement, and function calling.\n\nDeepHermes 3 Preview is one of the first LLM models to unify both \"intuitive\", traditional mode responses and long chain of thought reasoning responses into a single model, toggled by a system prompt.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -60643,25 +63732,25 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "arcee-ai/coder-large", + "permaslug": "nousresearch/deephermes-3-llama-3-8b-preview", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "650746f1-a509-49e4-9660-3ecc91c025f5", - "name": "Together | arcee-ai/coder-large", - "contextLength": 32768, + "id": "0e363cf8-2ac9-4f88-b85a-0020a1f60ce7", + "name": "Chutes | nousresearch/deephermes-3-llama-3-8b-preview:free", + "contextLength": 131072, "model": { - "slug": "arcee-ai/coder-large", - "hfSlug": "", - "updatedAt": "2025-05-05T21:48:22.361364+00:00", - "createdAt": "2025-05-05T20:57:43.438481+00:00", + "slug": "nousresearch/deephermes-3-llama-3-8b-preview", + "hfSlug": "NousResearch/DeepHermes-3-Llama-3-8B-Preview", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-28T05:09:32.50188+00:00", "hfUpdatedAt": null, - "name": "Arcee AI: Coder Large", - "shortName": "Coder Large", - "author": "arcee-ai", - "description": "Coder‑Large is a 32 B‑parameter offspring of Qwen 2.5‑Instruct that has been further trained on permissively‑licensed GitHub, CodeSearchNet and synthetic bug‑fix corpora. It supports a 32k context window, enabling multi‑file refactoring or long diff review in a single call, and understands 30‑plus programming languages with special attention to TypeScript, Go and Terraform. Internal benchmarks show 5–8 pt gains over CodeLlama‑34 B‑Python on HumanEval and competitive BugFix scores thanks to a reinforcement pass that rewards compilable output. The model emits structured explanations alongside code blocks by default, making it suitable for educational tooling as well as production copilot scenarios. Cost‑wise, Together AI prices it well below proprietary incumbents, so teams can scale interactive coding without runaway spend. ", + "name": "Nous: DeepHermes 3 Llama 3 8B Preview", + "shortName": "DeepHermes 3 Llama 3 8B Preview", + "author": "nousresearch", + "description": "DeepHermes 3 Preview is the latest version of our flagship Hermes series of LLMs by Nous Research, and one of the first models in the world to unify Reasoning (long chains of thought that improve answer accuracy) and normal LLM response modes into one model. We have also improved LLM annotation, judgement, and function calling.\n\nDeepHermes 3 Preview is one of the first LLM models to unify both \"intuitive\", traditional mode responses and long chain of thought reasoning responses into a single model, toggled by a system prompt.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -60676,32 +63765,30 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "arcee-ai/coder-large", + "permaslug": "nousresearch/deephermes-3-llama-3-8b-preview", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "arcee-ai/coder-large", - "modelVariantPermaslug": "arcee-ai/coder-large", - "providerName": "Together", + "modelVariantSlug": "nousresearch/deephermes-3-llama-3-8b-preview:free", + "modelVariantPermaslug": "nousresearch/deephermes-3-llama-3-8b-preview:free", + "providerName": "Chutes", "providerInfo": { - "name": "Together", - "displayName": "Together", - "slug": "together", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Together", + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, @@ -60709,235 +63796,13 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" - } - }, - "providerDisplayName": "Together", - "providerModelId": "arcee-ai/coder-large", - "providerGroup": "Together", - "quantization": null, - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", - "paidModels": { - "training": false, - "retainsPrompts": false - }, - "training": false, - "retainsPrompts": false - }, - "pricing": { - "prompt": "0.0000005", - "completion": "0.0000008", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 219, - "newest": 7, - "throughputHighToLow": 77, - "latencyLowToHigh": 103, - "pricingLowToHigh": 182, - "pricingHighToLow": 134 - }, - "providers": [ - { - "name": "Together", - "slug": "together", - "quantization": null, - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "arcee-ai/coder-large", - "pricing": { - "prompt": "0.0000005", - "completion": "0.0000008", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" - ], - "inputCost": 0.5, - "outputCost": 0.8 - } - ] - }, - { - "slug": "qwen/qwen3-0.6b-04-28", - "hfSlug": "Qwen/Qwen3-0.6B", - "updatedAt": "2025-05-12T00:34:37.771851+00:00", - "createdAt": "2025-04-30T20:05:26.503076+00:00", - "hfUpdatedAt": null, - "name": "Qwen: Qwen3 0.6B (free)", - "shortName": "Qwen3 0.6B (free)", - "author": "qwen", - "description": "Qwen3-0.6B is a lightweight, 0.6 billion parameter language model in the Qwen3 series, offering support for both general-purpose dialogue and structured reasoning through a dual-mode (thinking/non-thinking) architecture. Despite its small size, it supports long contexts up to 32,768 tokens and provides multilingual, tool-use, and instruction-following capabilities.", - "modelVersionGroupId": null, - "contextLength": 32000, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", - "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "qwen/qwen3-0.6b-04-28", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, - "endpoint": { - "id": "54c54d56-2d0d-4fa1-95ae-c74383a6e77e", - "name": "Novita | qwen/qwen3-0.6b-04-28:free", - "contextLength": 32000, - "model": { - "slug": "qwen/qwen3-0.6b-04-28", - "hfSlug": "Qwen/Qwen3-0.6B", - "updatedAt": "2025-05-12T00:34:37.771851+00:00", - "createdAt": "2025-04-30T20:05:26.503076+00:00", - "hfUpdatedAt": null, - "name": "Qwen: Qwen3 0.6B", - "shortName": "Qwen3 0.6B", - "author": "qwen", - "description": "Qwen3-0.6B is a lightweight, 0.6 billion parameter language model in the Qwen3 series, offering support for both general-purpose dialogue and structured reasoning through a dual-mode (thinking/non-thinking) architecture. Despite its small size, it supports long contexts up to 32,768 tokens and provides multilingual, tool-use, and instruction-following capabilities.", - "modelVersionGroupId": null, - "contextLength": 32000, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", - "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "qwen/qwen3-0.6b-04-28", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } - }, - "modelVariantSlug": "qwen/qwen3-0.6b-04-28:free", - "modelVariantPermaslug": "qwen/qwen3-0.6b-04-28:free", - "providerName": "Novita", - "providerInfo": { - "name": "Novita", - "displayName": "NovitaAI", - "slug": "novita", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", - "paidModels": { - "training": false - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "Novita", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": "https://status.novita.ai/", - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "NovitaAI", - "providerModelId": "qwen/qwen3-0.6b-fp8", - "providerGroup": "Novita", - "quantization": "fp8", + "providerDisplayName": "Chutes", + "providerModelId": "NousResearch/DeepHermes-3-Llama-3-8B-Preview", + "providerGroup": "Chutes", + "quantization": "bf16", "variant": "free", "isFree": true, "canAbort": true, @@ -60949,8 +63814,6 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -60958,17 +63821,20 @@ export default { "top_k", "min_p", "repetition_penalty", - "logit_bias" + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true }, - "training": false + "training": true, + "retainsPrompts": true }, "pricing": { "prompt": "0", @@ -60984,40 +63850,40 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 220, - "newest": 13, - "throughputHighToLow": 5, - "latencyLowToHigh": 85, - "pricingLowToHigh": 3, - "pricingHighToLow": 251 + "topWeekly": 221, + "newest": 105, + "throughputHighToLow": 24, + "latencyLowToHigh": 219, + "pricingLowToHigh": 45, + "pricingHighToLow": 293 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nousresearch.com/\\u0026size=256", "providers": [] }, { - "slug": "meta-llama/llama-3.1-405b", - "hfSlug": "meta-llama/llama-3.1-405B", + "slug": "deepseek/deepseek-coder", + "hfSlug": "deepseek-ai/DeepSeek-Coder-V2-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-02T00:00:00+00:00", + "createdAt": "2024-05-14T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.1 405B (base) (free)", - "shortName": "Llama 3.1 405B (base) (free)", - "author": "meta-llama", - "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This is the base 405B pre-trained version.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "modelVersionGroupId": "1fd9d06b-aa20-4c7d-a0b1-d3d9b5aae712", - "contextLength": 64000, + "name": "DeepSeek-Coder-V2", + "shortName": "DeepSeek-Coder-V2", + "author": "deepseek", + "description": "DeepSeek-Coder-V2, an open-source Mixture-of-Experts (MoE) code language model. It is further pre-trained from an intermediate checkpoint of DeepSeek-V2 with additional 6 trillion tokens.\n\nThe original V1 model was trained from scratch on 2T tokens, with a composition of 87% code and 13% natural language in both English and Chinese. It was pre-trained on project-level code corpus by employing a extra fill-in-the-blank task.", + "modelVersionGroupId": null, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -61025,32 +63891,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "none", + "group": "Other", + "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.1-405b", + "permaslug": "deepseek/deepseek-coder", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "95fab27a-2e59-403e-9767-ed8a3447ba71", - "name": "Chutes | meta-llama/llama-3.1-405b:free", - "contextLength": 64000, + "id": "91fc52a2-a00f-4b45-b3a6-0d74a352bc07", + "name": "Nebius | deepseek/deepseek-coder", + "contextLength": 128000, "model": { - "slug": "meta-llama/llama-3.1-405b", - "hfSlug": "meta-llama/llama-3.1-405B", + "slug": "deepseek/deepseek-coder", + "hfSlug": "deepseek-ai/DeepSeek-Coder-V2-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-02T00:00:00+00:00", + "createdAt": "2024-05-14T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.1 405B (base)", - "shortName": "Llama 3.1 405B (base)", - "author": "meta-llama", - "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This is the base 405B pre-trained version.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "modelVersionGroupId": "1fd9d06b-aa20-4c7d-a0b1-d3d9b5aae712", - "contextLength": 131072, + "name": "DeepSeek-Coder-V2", + "shortName": "DeepSeek-Coder-V2", + "author": "deepseek", + "description": "DeepSeek-Coder-V2, an open-source Mixture-of-Experts (MoE) code language model. It is further pre-trained from an intermediate checkpoint of DeepSeek-V2 with additional 6 trillion tokens.\n\nThe original V1 model was trained from scratch on 2T tokens, with a composition of 87% code and 13% natural language in both English and Chinese. It was pre-trained on project-level code corpus by employing a extra fill-in-the-blank task.", + "modelVersionGroupId": null, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -61058,37 +63924,39 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "none", + "group": "Other", + "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.1-405b", + "permaslug": "deepseek/deepseek-coder", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "meta-llama/llama-3.1-405b:free", - "modelVariantPermaslug": "meta-llama/llama-3.1-405b:free", - "providerName": "Chutes", + "modelVariantSlug": "deepseek/deepseek-coder", + "modelVariantPermaslug": "deepseek/deepseek-coder", + "providerName": "Nebius", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "Nebius", + "displayName": "Nebius AI Studio", + "slug": "nebius", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false } }, + "headquarters": "DE", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": true, + "isAbortable": false, "moderationRequired": false, - "group": "Chutes", + "group": "Nebius", "editors": [], "owners": [], "isMultipartSupported": true, @@ -61096,16 +63964,17 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "invertRequired": true } }, - "providerDisplayName": "Chutes", - "providerModelId": "chutesai/Llama-3.1-405B-FP8", - "providerGroup": "Chutes", + "providerDisplayName": "Nebius AI Studio", + "providerModelId": "deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct", + "providerGroup": "Nebius", "quantization": "fp8", - "variant": "free", - "isFree": true, - "canAbort": true, + "variant": "standard", + "isFree": false, + "canAbort": false, "maxPromptTokens": null, "maxCompletionTokens": null, "maxPromptImages": null, @@ -61119,26 +63988,25 @@ export default { "presence_penalty", "seed", "top_k", - "min_p", - "repetition_penalty", - "logprobs", "logit_bias", + "logprobs", "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.00000004", + "completion": "0.00000012", "image": "0", "request": "0", "webSearch": "0", @@ -61163,24 +64031,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 109, - "newest": 225, - "throughputHighToLow": 268, - "latencyLowToHigh": 305, - "pricingLowToHigh": 66, - "pricingHighToLow": 78 + "topWeekly": 222, + "newest": 256, + "throughputHighToLow": 55, + "latencyLowToHigh": 15, + "pricingLowToHigh": 92, + "pricingHighToLow": 226 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [ { - "name": "Hyperbolic", - "slug": "hyperbolic", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", "quantization": "fp8", - "context": 32768, + "context": 128000, "maxCompletionTokens": null, - "providerModelId": "meta-llama/Meta-Llama-3.1-405B-FP8", + "providerModelId": "deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct", "pricing": { - "prompt": "0.000002", - "completion": "0.000002", + "prompt": "0.00000004", + "completion": "0.00000012", "image": "0", "request": "0", "webSearch": "0", @@ -61194,154 +64064,138 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "logprobs", - "top_logprobs", "seed", - "logit_bias", "top_k", - "min_p", - "repetition_penalty" - ], - "inputCost": 2, - "outputCost": 2 - }, - { - "name": "Hyperbolic", - "slug": "hyperbolic", - "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": 32768, - "providerModelId": "meta-llama/Meta-Llama-3.1-405B", - "pricing": { - "prompt": "0.000004", - "completion": "0.000004", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "logprobs", - "top_logprobs", - "seed", "logit_bias", - "top_k", - "min_p", - "repetition_penalty" + "logprobs", + "top_logprobs" ], - "inputCost": 4, - "outputCost": 4 + "inputCost": 0.04, + "outputCost": 0.12, + "throughput": 123.565, + "latency": 233 } ] }, { - "slug": "mistralai/pixtral-large-2411", - "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-19T00:49:48.873161+00:00", + "slug": "perplexity/r1-1776", + "hfSlug": "perplexity-ai/r1-1776", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-02-19T22:42:09.214134+00:00", "hfUpdatedAt": null, - "name": "Mistral: Pixtral Large 2411", - "shortName": "Pixtral Large 2411", - "author": "mistralai", - "description": "Pixtral Large is a 124B parameter, open-weight, multimodal model built on top of [Mistral Large 2](/mistralai/mistral-large-2411). The model is able to understand documents, charts and natural images.\n\nThe model is available under the Mistral Research License (MRL) for research and educational use, and the Mistral Commercial License for experimentation, testing, and production for commercial purposes.\n\n", + "name": "Perplexity: R1 1776", + "shortName": "R1 1776", + "author": "perplexity", + "description": "R1 1776 is a version of DeepSeek-R1 that has been post-trained to remove censorship constraints related to topics restricted by the Chinese government. The model retains its original reasoning capabilities while providing direct responses to a wider range of queries. R1 1776 is an offline chat model that does not use the perplexity search subsystem.\n\nThe model was tested on a multilingual dataset of over 1,000 examples covering sensitive topics to measure its likelihood of refusal or overly filtered responses. [Evaluation Results](https://cdn-uploads.huggingface.co/production/uploads/675c8332d01f593dc90817f5/GiN2VqC5hawUgAGJ6oHla.png) Its performance on math and reasoning benchmarks remains similar to the base R1 model. [Reasoning Performance](https://cdn-uploads.huggingface.co/production/uploads/675c8332d01f593dc90817f5/n4Z9Byqp2S7sKUvCvI40R.png)\n\nRead more on the [Blog Post](https://perplexity.ai/hub/blog/open-sourcing-r1-1776)", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": null, + "group": "DeepSeek", + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/pixtral-large-2411", - "reasoningConfig": null, - "features": {}, + "permaslug": "perplexity/r1-1776", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "1a41639e-c1cf-422e-a871-27bc67f03928", - "name": "Mistral | mistralai/pixtral-large-2411", - "contextLength": 131072, + "id": "62190e52-ae3c-4b5b-913c-4a821a16de0b", + "name": "Perplexity | perplexity/r1-1776", + "contextLength": 128000, "model": { - "slug": "mistralai/pixtral-large-2411", - "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-19T00:49:48.873161+00:00", + "slug": "perplexity/r1-1776", + "hfSlug": "perplexity-ai/r1-1776", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-02-19T22:42:09.214134+00:00", "hfUpdatedAt": null, - "name": "Mistral: Pixtral Large 2411", - "shortName": "Pixtral Large 2411", - "author": "mistralai", - "description": "Pixtral Large is a 124B parameter, open-weight, multimodal model built on top of [Mistral Large 2](/mistralai/mistral-large-2411). The model is able to understand documents, charts and natural images.\n\nThe model is available under the Mistral Research License (MRL) for research and educational use, and the Mistral Commercial License for experimentation, testing, and production for commercial purposes.\n\n", + "name": "Perplexity: R1 1776", + "shortName": "R1 1776", + "author": "perplexity", + "description": "R1 1776 is a version of DeepSeek-R1 that has been post-trained to remove censorship constraints related to topics restricted by the Chinese government. The model retains its original reasoning capabilities while providing direct responses to a wider range of queries. R1 1776 is an offline chat model that does not use the perplexity search subsystem.\n\nThe model was tested on a multilingual dataset of over 1,000 examples covering sensitive topics to measure its likelihood of refusal or overly filtered responses. [Evaluation Results](https://cdn-uploads.huggingface.co/production/uploads/675c8332d01f593dc90817f5/GiN2VqC5hawUgAGJ6oHla.png) Its performance on math and reasoning benchmarks remains similar to the base R1 model. [Reasoning Performance](https://cdn-uploads.huggingface.co/production/uploads/675c8332d01f593dc90817f5/n4Z9Byqp2S7sKUvCvI40R.png)\n\nRead more on the [Blog Post](https://perplexity.ai/hub/blog/open-sourcing-r1-1776)", "modelVersionGroupId": null, "contextLength": 128000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": null, + "group": "DeepSeek", + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/pixtral-large-2411", - "reasoningConfig": null, - "features": {} + "permaslug": "perplexity/r1-1776", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "mistralai/pixtral-large-2411", - "modelVariantPermaslug": "mistralai/pixtral-large-2411", - "providerName": "Mistral", + "modelVariantSlug": "perplexity/r1-1776", + "modelVariantPermaslug": "perplexity/r1-1776", + "providerName": "Perplexity", "providerInfo": { - "name": "Mistral", - "displayName": "Mistral", - "slug": "mistral", + "name": "Perplexity", + "displayName": "Perplexity", + "slug": "perplexity", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", + "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false } }, - "headquarters": "FR", + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": false, "isAbortable": false, "moderationRequired": false, - "group": "Mistral", + "group": "Perplexity", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.perplexity.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Mistral.png" + "url": "/images/icons/Perplexity.svg" } }, - "providerDisplayName": "Mistral", - "providerModelId": "pixtral-large-2411", - "providerGroup": "Mistral", + "providerDisplayName": "Perplexity", + "providerModelId": "r1-1776", + "providerGroup": "Perplexity", "quantization": null, "variant": "standard", "isFree": false, @@ -61351,36 +64205,29 @@ export default { "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "stop", + "reasoning", + "include_reasoning", + "top_k", "frequency_penalty", - "presence_penalty", - "response_format", - "structured_outputs", - "seed" + "presence_penalty" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "termsOfServiceUrl": "https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service", + "privacyPolicyUrl": "https://www.perplexity.ai/hub/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, "pricing": { "prompt": "0.000002", - "completion": "0.000006", - "image": "0.002888", + "completion": "0.000008", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -61390,74 +64237,111 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, + "supportsToolParameters": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": false, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 222, - "newest": 169, - "throughputHighToLow": 243, - "latencyLowToHigh": 243, - "pricingLowToHigh": 245, - "pricingHighToLow": 74 + "topWeekly": 223, + "newest": 111, + "throughputHighToLow": 121, + "latencyLowToHigh": 164, + "pricingLowToHigh": 251, + "pricingHighToLow": 71 }, + "authorIcon": "https://openrouter.ai/images/icons/Perplexity.svg", "providers": [ { - "name": "Mistral", - "slug": "mistral", + "name": "Perplexity", + "icon": "https://openrouter.ai/images/icons/Perplexity.svg", + "slug": "perplexity", "quantization": null, - "context": 131072, + "context": 128000, "maxCompletionTokens": null, - "providerModelId": "pixtral-large-2411", + "providerModelId": "r1-1776", "pricing": { "prompt": "0.000002", - "completion": "0.000006", - "image": "0.002888", + "completion": "0.000008", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "top_k", + "frequency_penalty", + "presence_penalty" + ], + "inputCost": 2, + "outputCost": 8, + "throughput": 60.043, + "latency": 1052.5 + }, + { + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": null, + "context": 163840, + "maxCompletionTokens": null, + "providerModelId": "perplexity-ai/r1-1776", + "pricing": { + "prompt": "0.000003", + "completion": "0.000007", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "response_format", - "structured_outputs", - "seed" + "top_k", + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" ], - "inputCost": 2, - "outputCost": 6 + "inputCost": 3, + "outputCost": 7, + "throughput": 69.044, + "latency": 840 } ] }, { - "slug": "nousresearch/deephermes-3-llama-3-8b-preview", - "hfSlug": "NousResearch/DeepHermes-3-Llama-3-8B-Preview", + "slug": "cohere/command-r-plus-04-2024", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-28T05:09:32.50188+00:00", + "createdAt": "2024-04-02T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Nous: DeepHermes 3 Llama 3 8B Preview (free)", - "shortName": "DeepHermes 3 Llama 3 8B Preview (free)", - "author": "nousresearch", - "description": "DeepHermes 3 Preview is the latest version of our flagship Hermes series of LLMs by Nous Research, and one of the first models in the world to unify Reasoning (long chains of thought that improve answer accuracy) and normal LLM response modes into one model. We have also improved LLM annotation, judgement, and function calling.\n\nDeepHermes 3 Preview is one of the first LLM models to unify both \"intuitive\", traditional mode responses and long chain of thought reasoning responses into a single model, toggled by a system prompt.", + "name": "Cohere: Command R+ (04-2024)", + "shortName": "Command R+ (04-2024)", + "author": "cohere", + "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -61465,32 +64349,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "Cohere", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "nousresearch/deephermes-3-llama-3-8b-preview", + "permaslug": "cohere/command-r-plus-04-2024", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "0e363cf8-2ac9-4f88-b85a-0020a1f60ce7", - "name": "Chutes | nousresearch/deephermes-3-llama-3-8b-preview:free", - "contextLength": 131072, + "id": "bc75924a-9b83-418a-a36d-d8d0b152ab93", + "name": "Cohere | cohere/command-r-plus-04-2024", + "contextLength": 128000, "model": { - "slug": "nousresearch/deephermes-3-llama-3-8b-preview", - "hfSlug": "NousResearch/DeepHermes-3-Llama-3-8B-Preview", + "slug": "cohere/command-r-plus-04-2024", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-28T05:09:32.50188+00:00", + "createdAt": "2024-04-02T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Nous: DeepHermes 3 Llama 3 8B Preview", - "shortName": "DeepHermes 3 Llama 3 8B Preview", - "author": "nousresearch", - "description": "DeepHermes 3 Preview is the latest version of our flagship Hermes series of LLMs by Nous Research, and one of the first models in the world to unify Reasoning (long chains of thought that improve answer accuracy) and normal LLM response modes into one model. We have also improved LLM annotation, judgement, and function calling.\n\nDeepHermes 3 Preview is one of the first LLM models to unify both \"intuitive\", traditional mode responses and long chain of thought reasoning responses into a single model, toggled by a system prompt.", + "name": "Cohere: Command R+ (04-2024)", + "shortName": "Command R+ (04-2024)", + "author": "cohere", + "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -61498,87 +64382,91 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "Cohere", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "nousresearch/deephermes-3-llama-3-8b-preview", + "permaslug": "cohere/command-r-plus-04-2024", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "nousresearch/deephermes-3-llama-3-8b-preview:free", - "modelVariantPermaslug": "nousresearch/deephermes-3-llama-3-8b-preview:free", - "providerName": "Chutes", + "modelVariantSlug": "cohere/command-r-plus-04-2024", + "modelVariantPermaslug": "cohere/command-r-plus-04-2024", + "providerName": "Cohere", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "Cohere", + "displayName": "Cohere", + "slug": "cohere", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "privacyPolicyUrl": "https://cohere.com/privacy", + "termsOfServiceUrl": "https://cohere.com/terms-of-use", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": true, + "retentionDays": 30 } }, + "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, + "hasCompletions": false, "isAbortable": true, "moderationRequired": false, - "group": "Chutes", + "group": "Cohere", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.cohere.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "/images/icons/Cohere.png" } }, - "providerDisplayName": "Chutes", - "providerModelId": "NousResearch/DeepHermes-3-Llama-3-8B-Preview", - "providerGroup": "Chutes", - "quantization": "bf16", - "variant": "free", - "isFree": true, + "providerDisplayName": "Cohere", + "providerModelId": "command-r-plus-04-2024", + "providerGroup": "Cohere", + "quantization": "unknown", + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "seed", "top_k", - "min_p", - "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" + "seed", + "response_format", + "structured_outputs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "privacyPolicyUrl": "https://cohere.com/privacy", + "termsOfServiceUrl": "https://cohere.com/terms-of-use", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.000003", + "completion": "0.000015", "image": "0", "request": "0", "webSearch": "0", @@ -61589,38 +64477,76 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 223, - "newest": 105, - "throughputHighToLow": 41, - "latencyLowToHigh": 191, - "pricingLowToHigh": 45, - "pricingHighToLow": 293 + "topWeekly": 224, + "newest": 273, + "throughputHighToLow": 151, + "latencyLowToHigh": 35, + "pricingLowToHigh": 277, + "pricingHighToLow": 49 }, - "providers": [] + "authorIcon": "https://openrouter.ai/images/icons/Cohere.png", + "providers": [ + { + "name": "Cohere", + "icon": "https://openrouter.ai/images/icons/Cohere.png", + "slug": "cohere", + "quantization": "unknown", + "context": 128000, + "maxCompletionTokens": 4000, + "providerModelId": "command-r-plus-04-2024", + "pricing": { + "prompt": "0.000003", + "completion": "0.000015", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "seed", + "response_format", + "structured_outputs" + ], + "inputCost": 3, + "outputCost": 15, + "throughput": 57.274, + "latency": 393 + } + ] }, { - "slug": "arliai/qwq-32b-arliai-rpr-v1", - "hfSlug": "ArliAI/QwQ-32B-ArliAI-RpR-v1", + "slug": "open-r1/olympiccoder-32b", + "hfSlug": "open-r1/OlympicCoder-32B", "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-04-13T14:53:02.382429+00:00", + "createdAt": "2025-03-15T22:20:28.942272+00:00", "hfUpdatedAt": null, - "name": "ArliAI: QwQ 32B RpR v1 (free)", - "shortName": "QwQ 32B RpR v1 (free)", - "author": "arliai", - "description": "QwQ-32B-ArliAI-RpR-v1 is a 32B parameter model fine-tuned from Qwen/QwQ-32B using a curated creative writing and roleplay dataset originally developed for the RPMax series. It is designed to maintain coherence and reasoning across long multi-turn conversations by introducing explicit reasoning steps per dialogue turn, generated and refined using the base model itself.\n\nThe model was trained using RS-QLORA+ on 8K sequence lengths and supports up to 128K context windows (with practical performance around 32K). It is optimized for creative roleplay and dialogue generation, with an emphasis on minimizing cross-context repetition while preserving stylistic diversity.", + "name": "OlympicCoder 32B (free)", + "shortName": "OlympicCoder 32B (free)", + "author": "open-r1", + "description": "OlympicCoder-32B is a high-performing open-source model fine-tuned using the CodeForces-CoTs dataset, containing approximately 100,000 chain-of-thought programming samples. It excels at complex competitive programming benchmarks, such as IOI 2024 and Codeforces-style challenges, frequently surpassing state-of-the-art closed-source models. OlympicCoder-32B provides advanced reasoning, coherent multi-step problem-solving, and robust code generation capabilities, demonstrating significant potential for olympiad-level competitive programming applications.", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ @@ -61640,7 +64566,7 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "arliai/qwq-32b-arliai-rpr-v1", + "permaslug": "open-r1/olympiccoder-32b", "reasoningConfig": { "startToken": "", "endToken": "" @@ -61652,19 +64578,19 @@ export default { } }, "endpoint": { - "id": "d6cbde45-bd1e-47b2-949e-61d0453fc229", - "name": "Chutes | arliai/qwq-32b-arliai-rpr-v1:free", + "id": "cf95d8ad-af21-42a1-9fa4-c68d7df33805", + "name": "Chutes | open-r1/olympiccoder-32b:free", "contextLength": 32768, "model": { - "slug": "arliai/qwq-32b-arliai-rpr-v1", - "hfSlug": "ArliAI/QwQ-32B-ArliAI-RpR-v1", + "slug": "open-r1/olympiccoder-32b", + "hfSlug": "open-r1/OlympicCoder-32B", "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-04-13T14:53:02.382429+00:00", + "createdAt": "2025-03-15T22:20:28.942272+00:00", "hfUpdatedAt": null, - "name": "ArliAI: QwQ 32B RpR v1", - "shortName": "QwQ 32B RpR v1", - "author": "arliai", - "description": "QwQ-32B-ArliAI-RpR-v1 is a 32B parameter model fine-tuned from Qwen/QwQ-32B using a curated creative writing and roleplay dataset originally developed for the RPMax series. It is designed to maintain coherence and reasoning across long multi-turn conversations by introducing explicit reasoning steps per dialogue turn, generated and refined using the base model itself.\n\nThe model was trained using RS-QLORA+ on 8K sequence lengths and supports up to 128K context windows (with practical performance around 32K). It is optimized for creative roleplay and dialogue generation, with an emphasis on minimizing cross-context repetition while preserving stylistic diversity.", + "name": "OlympicCoder 32B", + "shortName": "OlympicCoder 32B", + "author": "open-r1", + "description": "OlympicCoder-32B is a high-performing open-source model fine-tuned using the CodeForces-CoTs dataset, containing approximately 100,000 chain-of-thought programming samples. It excels at complex competitive programming benchmarks, such as IOI 2024 and Codeforces-style challenges, frequently surpassing state-of-the-art closed-source models. OlympicCoder-32B provides advanced reasoning, coherent multi-step problem-solving, and robust code generation capabilities, demonstrating significant potential for olympiad-level competitive programming applications.", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ @@ -61684,7 +64610,7 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "arliai/qwq-32b-arliai-rpr-v1", + "permaslug": "open-r1/olympiccoder-32b", "reasoningConfig": { "startToken": "", "endToken": "" @@ -61696,8 +64622,8 @@ export default { } } }, - "modelVariantSlug": "arliai/qwq-32b-arliai-rpr-v1:free", - "modelVariantPermaslug": "arliai/qwq-32b-arliai-rpr-v1:free", + "modelVariantSlug": "open-r1/olympiccoder-32b:free", + "modelVariantPermaslug": "open-r1/olympiccoder-32b:free", "providerName": "Chutes", "providerInfo": { "name": "Chutes", @@ -61727,9 +64653,9 @@ export default { } }, "providerDisplayName": "Chutes", - "providerModelId": "ArliAI/QwQ-32B-ArliAI-RpR-v1", + "providerModelId": "open-r1/OlympicCoder-32B", "providerGroup": "Chutes", - "quantization": null, + "quantization": "bf16", "variant": "free", "isFree": true, "canAbort": true, @@ -61786,33 +64712,33 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 224, - "newest": 53, - "throughputHighToLow": 219, - "latencyLowToHigh": 210, - "pricingLowToHigh": 21, - "pricingHighToLow": 269 + "topWeekly": 225, + "newest": 81, + "throughputHighToLow": 143, + "latencyLowToHigh": 275, + "pricingLowToHigh": 36, + "pricingHighToLow": 284 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { - "slug": "cohere/command-r-plus-04-2024", - "hfSlug": null, + "slug": "meta-llama/llama-3.1-405b", + "hfSlug": "meta-llama/llama-3.1-405B", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-04-02T00:00:00+00:00", + "createdAt": "2024-08-02T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Cohere: Command R+ (04-2024)", - "shortName": "Command R+ (04-2024)", - "author": "cohere", - "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", - "modelVersionGroupId": null, - "contextLength": 128000, + "name": "Meta: Llama 3.1 405B (base) (free)", + "shortName": "Llama 3.1 405B (base) (free)", + "author": "meta-llama", + "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This is the base 405B pre-trained version.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "modelVersionGroupId": "1fd9d06b-aa20-4c7d-a0b1-d3d9b5aae712", + "contextLength": 64000, "inputModalities": [ "text" ], @@ -61820,32 +64746,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Cohere", - "instructType": null, + "group": "Llama3", + "instructType": "none", "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cohere/command-r-plus-04-2024", + "permaslug": "meta-llama/llama-3.1-405b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "bc75924a-9b83-418a-a36d-d8d0b152ab93", - "name": "Cohere | cohere/command-r-plus-04-2024", - "contextLength": 128000, + "id": "95fab27a-2e59-403e-9767-ed8a3447ba71", + "name": "Chutes | meta-llama/llama-3.1-405b:free", + "contextLength": 64000, "model": { - "slug": "cohere/command-r-plus-04-2024", - "hfSlug": null, + "slug": "meta-llama/llama-3.1-405b", + "hfSlug": "meta-llama/llama-3.1-405B", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-04-02T00:00:00+00:00", + "createdAt": "2024-08-02T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Cohere: Command R+ (04-2024)", - "shortName": "Command R+ (04-2024)", - "author": "cohere", - "description": "Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).\n\nIt offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", - "modelVersionGroupId": null, - "contextLength": 128000, + "name": "Meta: Llama 3.1 405B (base)", + "shortName": "Llama 3.1 405B (base)", + "author": "meta-llama", + "description": "Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors. This is the base 405B pre-trained version.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "modelVersionGroupId": "1fd9d06b-aa20-4c7d-a0b1-d3d9b5aae712", + "contextLength": 131072, "inputModalities": [ "text" ], @@ -61853,91 +64779,87 @@ export default { "text" ], "hasTextOutput": true, - "group": "Cohere", - "instructType": null, + "group": "Llama3", + "instructType": "none", "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cohere/command-r-plus-04-2024", + "permaslug": "meta-llama/llama-3.1-405b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "cohere/command-r-plus-04-2024", - "modelVariantPermaslug": "cohere/command-r-plus-04-2024", - "providerName": "Cohere", + "modelVariantSlug": "meta-llama/llama-3.1-405b:free", + "modelVariantPermaslug": "meta-llama/llama-3.1-405b:free", + "providerName": "Chutes", "providerInfo": { - "name": "Cohere", - "displayName": "Cohere", - "slug": "cohere", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://cohere.com/privacy", - "termsOfServiceUrl": "https://cohere.com/terms-of-use", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true } }, - "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, + "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Cohere", + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.cohere.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Cohere.png" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "Cohere", - "providerModelId": "command-r-plus-04-2024", - "providerGroup": "Cohere", - "quantization": "unknown", - "variant": "standard", - "isFree": false, + "providerDisplayName": "Chutes", + "providerModelId": "chutesai/Llama-3.1-405B-FP8", + "providerGroup": "Chutes", + "quantization": "fp8", + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4000, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "top_k", "seed", - "response_format", - "structured_outputs" + "top_k", + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://cohere.com/privacy", - "termsOfServiceUrl": "https://cohere.com/terms-of-use", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.000003", - "completion": "0.000015", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -61948,12 +64870,12 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { "supportedParameters": {}, @@ -61962,24 +64884,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 225, - "newest": 273, - "throughputHighToLow": 136, - "latencyLowToHigh": 45, - "pricingLowToHigh": 277, - "pricingHighToLow": 49 + "topWeekly": 110, + "newest": 224, + "throughputHighToLow": 270, + "latencyLowToHigh": 292, + "pricingLowToHigh": 66, + "pricingHighToLow": 78 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { - "name": "Cohere", - "slug": "cohere", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 4000, - "providerModelId": "command-r-plus-04-2024", + "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", + "slug": "hyperbolic", + "quantization": "fp8", + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Meta-Llama-3.1-405B-FP8", "pricing": { - "prompt": "0.000003", - "completion": "0.000015", + "prompt": "0.000002", + "completion": "0.000002", "image": "0", "request": "0", "webSearch": "0", @@ -61987,35 +64911,74 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", + "logprobs", + "top_logprobs", + "seed", + "logit_bias", "top_k", + "min_p", + "repetition_penalty" + ], + "inputCost": 2, + "outputCost": 2 + }, + { + "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", + "slug": "hyperbolic", + "quantization": "bf16", + "context": 32768, + "maxCompletionTokens": 32768, + "providerModelId": "meta-llama/Meta-Llama-3.1-405B", + "pricing": { + "prompt": "0.000004", + "completion": "0.000004", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "logprobs", + "top_logprobs", "seed", - "response_format", - "structured_outputs" + "logit_bias", + "top_k", + "min_p", + "repetition_penalty" ], - "inputCost": 3, - "outputCost": 15 + "inputCost": 4, + "outputCost": 4, + "throughput": 12.006, + "latency": 1129.5 } ] }, { - "slug": "alpindale/goliath-120b", - "hfSlug": "alpindale/goliath-120b", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-11-10T00:00:00+00:00", + "slug": "qwen/qwen3-0.6b-04-28", + "hfSlug": "Qwen/Qwen3-0.6B", + "updatedAt": "2025-05-12T00:34:37.771851+00:00", + "createdAt": "2025-04-30T20:05:26.503076+00:00", "hfUpdatedAt": null, - "name": "Goliath 120B", - "shortName": "Goliath 120B", - "author": "alpindale", - "description": "A large LLM created by combining two fine-tuned Llama 70B models into one 120B model. Combines Xwin and Euryale.\n\nCredits to\n- [@chargoddard](https://huggingface.co/chargoddard) for developing the framework used to merge the model - [mergekit](https://github.com/cg123/mergekit).\n- [@Undi95](https://huggingface.co/Undi95) for helping with the merge ratios.\n\n#merge", + "name": "Qwen: Qwen3 0.6B (free)", + "shortName": "Qwen3 0.6B (free)", + "author": "qwen", + "description": "Qwen3-0.6B is a lightweight, 0.6 billion parameter language model in the Qwen3 series, offering support for both general-purpose dialogue and structured reasoning through a dual-mode (thinking/non-thinking) architecture. Despite its small size, it supports long contexts up to 32,768 tokens and provides multilingual, tool-use, and instruction-following capabilities.", "modelVersionGroupId": null, - "contextLength": 6144, + "contextLength": 32000, "inputModalities": [ "text" ], @@ -62023,35 +64986,43 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama2", - "instructType": "airoboros", + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, "defaultStops": [ - "USER:", - "" + "<|im_start|>", + "<|im_end|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "alpindale/goliath-120b", - "reasoningConfig": null, - "features": {}, + "permaslug": "qwen/qwen3-0.6b-04-28", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "73654edf-bbfe-43bd-9c4c-13b858f00dc8", - "name": "Mancer | alpindale/goliath-120b", - "contextLength": 6144, + "id": "54c54d56-2d0d-4fa1-95ae-c74383a6e77e", + "name": "Novita | qwen/qwen3-0.6b-04-28:free", + "contextLength": 32000, "model": { - "slug": "alpindale/goliath-120b", - "hfSlug": "alpindale/goliath-120b", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-11-10T00:00:00+00:00", + "slug": "qwen/qwen3-0.6b-04-28", + "hfSlug": "Qwen/Qwen3-0.6B", + "updatedAt": "2025-05-12T00:34:37.771851+00:00", + "createdAt": "2025-04-30T20:05:26.503076+00:00", "hfUpdatedAt": null, - "name": "Goliath 120B", - "shortName": "Goliath 120B", - "author": "alpindale", - "description": "A large LLM created by combining two fine-tuned Llama 70B models into one 120B model. Combines Xwin and Euryale.\n\nCredits to\n- [@chargoddard](https://huggingface.co/chargoddard) for developing the framework used to merge the model - [mergekit](https://github.com/cg123/mergekit).\n- [@Undi95](https://huggingface.co/Undi95) for helping with the merge ratios.\n\n#merge", + "name": "Qwen: Qwen3 0.6B", + "shortName": "Qwen3 0.6B", + "author": "qwen", + "description": "Qwen3-0.6B is a lightweight, 0.6 billion parameter language model in the Qwen3 series, offering support for both general-purpose dialogue and structured reasoning through a dual-mode (thinking/non-thinking) architecture. Despite its small size, it supports long contexts up to 32,768 tokens and provides multilingual, tool-use, and instruction-following capabilities.", "modelVersionGroupId": null, - "contextLength": 6144, + "contextLength": 32000, "inputModalities": [ "text" ], @@ -62059,189 +65030,132 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama2", - "instructType": "airoboros", + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, "defaultStops": [ - "USER:", - "" + "<|im_start|>", + "<|im_end|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "alpindale/goliath-120b", - "reasoningConfig": null, - "features": {} + "permaslug": "qwen/qwen3-0.6b-04-28", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "alpindale/goliath-120b", - "modelVariantPermaslug": "alpindale/goliath-120b", - "providerName": "Mancer", + "modelVariantSlug": "qwen/qwen3-0.6b-04-28:free", + "modelVariantPermaslug": "qwen/qwen3-0.6b-04-28:free", + "providerName": "Novita", "providerInfo": { - "name": "Mancer", - "displayName": "Mancer", - "slug": "mancer", + "name": "Novita", + "displayName": "NovitaAI", + "slug": "novita", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://mancer.tech/terms", - "privacyPolicyUrl": "https://mancer.tech/privacy", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false } }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Mancer", + "group": "Novita", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.novita.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "invertRequired": true } }, - "providerDisplayName": "Mancer", - "providerModelId": "goliath-120b", - "providerGroup": "Mancer", - "quantization": "int4", - "variant": "standard", - "isFree": false, + "providerDisplayName": "NovitaAI", + "providerModelId": "qwen/qwen3-0.6b-fp8", + "providerGroup": "Novita", + "quantization": "fp8", + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 512, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "logit_bias", + "seed", "top_k", "min_p", - "seed", - "top_a" + "repetition_penalty", + "logit_bias" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://mancer.tech/terms", - "privacyPolicyUrl": "https://mancer.tech/privacy", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false }, - "training": true, - "retainsPrompts": true + "training": false }, "pricing": { - "prompt": "0.0000065625", - "completion": "0.000009375", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", - "discount": 0.25 + "discount": 0 }, "variablePricings": [], "isHidden": false, "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 226, - "newest": 298, - "throughputHighToLow": 286, - "latencyLowToHigh": 247, - "pricingLowToHigh": 294, - "pricingHighToLow": 24 + "topWeekly": 227, + "newest": 13, + "throughputHighToLow": 182, + "latencyLowToHigh": 108, + "pricingLowToHigh": 3, + "pricingHighToLow": 251 }, - "providers": [ - { - "name": "Mancer", - "slug": "mancer", - "quantization": "int4", - "context": 6144, - "maxCompletionTokens": 512, - "providerModelId": "goliath-120b", - "pricing": { - "prompt": "0.0000065625", - "completion": "0.000009375", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0.25 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "logit_bias", - "top_k", - "min_p", - "seed", - "top_a" - ], - "inputCost": 6.56, - "outputCost": 9.38 - }, - { - "name": "Mancer (private)", - "slug": "mancer (private)", - "quantization": "int4", - "context": 6144, - "maxCompletionTokens": 512, - "providerModelId": "goliath-120b", - "pricing": { - "prompt": "0.00000875", - "completion": "0.0000125", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "logit_bias", - "top_k", - "min_p", - "seed", - "top_a" - ], - "inputCost": 8.75, - "outputCost": 12.5 - } - ] + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", + "providers": [] }, { "slug": "inflection/inflection-3-pi", @@ -62395,16 +65309,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 227, - "newest": 191, - "throughputHighToLow": 205, + "topWeekly": 228, + "newest": 192, + "throughputHighToLow": 203, "latencyLowToHigh": 270, "pricingLowToHigh": 261, "pricingHighToLow": 56 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://inflection.ai/\\u0026size=256", "providers": [ { "name": "Inflection", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inflection.ai/&size=256", "slug": "inflection", "quantization": "unknown", "context": 8000, @@ -62426,22 +65342,24 @@ export default { "stop" ], "inputCost": 2.5, - "outputCost": 10 + "outputCost": 10, + "throughput": 41.514, + "latency": 2471 } ] }, { - "slug": "qwen/qwen3-1.7b", - "hfSlug": "Qwen/Qwen3-1.7B", - "updatedAt": "2025-05-12T00:34:59.533129+00:00", - "createdAt": "2025-04-30T16:43:08.565779+00:00", + "slug": "arliai/qwq-32b-arliai-rpr-v1", + "hfSlug": "ArliAI/QwQ-32B-ArliAI-RpR-v1", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-04-13T14:53:02.382429+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 1.7B (free)", - "shortName": "Qwen3 1.7B (free)", - "author": "qwen", - "description": "Qwen3-1.7B is a compact, 1.7 billion parameter dense language model from the Qwen3 series, featuring dual-mode operation for both efficient dialogue (non-thinking) and advanced reasoning (thinking). Despite its small size, it supports 32,768-token contexts and delivers strong multilingual, instruction-following, and agentic capabilities, including tool use and structured output.", + "name": "ArliAI: QwQ 32B RpR v1 (free)", + "shortName": "QwQ 32B RpR v1 (free)", + "author": "arliai", + "description": "QwQ-32B-ArliAI-RpR-v1 is a 32B parameter model fine-tuned from Qwen/QwQ-32B using a curated creative writing and roleplay dataset originally developed for the RPMax series. It is designed to maintain coherence and reasoning across long multi-turn conversations by introducing explicit reasoning steps per dialogue turn, generated and refined using the base model itself.\n\nThe model was trained using RS-QLORA+ on 8K sequence lengths and supports up to 128K context windows (with practical performance around 32K). It is optimized for creative roleplay and dialogue generation, with an emphasis on minimizing cross-context repetition while preserving stylistic diversity.", "modelVersionGroupId": null, - "contextLength": 32000, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -62449,17 +65367,17 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Other", + "instructType": "deepseek-r1", "defaultSystem": null, "defaultStops": [ - "<|im_start|>", - "<|im_end|>" + "<｜User｜>", + "<｜end▁of▁sentence｜>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen3-1.7b-04-28", + "permaslug": "arliai/qwq-32b-arliai-rpr-v1", "reasoningConfig": { "startToken": "", "endToken": "" @@ -62471,21 +65389,21 @@ export default { } }, "endpoint": { - "id": "390a675c-72d2-4cfd-895e-63653fd3bf29", - "name": "Novita | qwen/qwen3-1.7b-04-28:free", - "contextLength": 32000, + "id": "d6cbde45-bd1e-47b2-949e-61d0453fc229", + "name": "Chutes | arliai/qwq-32b-arliai-rpr-v1:free", + "contextLength": 32768, "model": { - "slug": "qwen/qwen3-1.7b", - "hfSlug": "Qwen/Qwen3-1.7B", - "updatedAt": "2025-05-12T00:34:59.533129+00:00", - "createdAt": "2025-04-30T16:43:08.565779+00:00", + "slug": "arliai/qwq-32b-arliai-rpr-v1", + "hfSlug": "ArliAI/QwQ-32B-ArliAI-RpR-v1", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-04-13T14:53:02.382429+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen3 1.7B", - "shortName": "Qwen3 1.7B", - "author": "qwen", - "description": "Qwen3-1.7B is a compact, 1.7 billion parameter dense language model from the Qwen3 series, featuring dual-mode operation for both efficient dialogue (non-thinking) and advanced reasoning (thinking). Despite its small size, it supports 32,768-token contexts and delivers strong multilingual, instruction-following, and agentic capabilities, including tool use and structured output.", + "name": "ArliAI: QwQ 32B RpR v1", + "shortName": "QwQ 32B RpR v1", + "author": "arliai", + "description": "QwQ-32B-ArliAI-RpR-v1 is a 32B parameter model fine-tuned from Qwen/QwQ-32B using a curated creative writing and roleplay dataset originally developed for the RPMax series. It is designed to maintain coherence and reasoning across long multi-turn conversations by introducing explicit reasoning steps per dialogue turn, generated and refined using the base model itself.\n\nThe model was trained using RS-QLORA+ on 8K sequence lengths and supports up to 128K context windows (with practical performance around 32K). It is optimized for creative roleplay and dialogue generation, with an emphasis on minimizing cross-context repetition while preserving stylistic diversity.", "modelVersionGroupId": null, - "contextLength": 32000, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -62493,17 +65411,17 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen3", - "instructType": "qwen3", + "group": "Other", + "instructType": "deepseek-r1", "defaultSystem": null, "defaultStops": [ - "<|im_start|>", - "<|im_end|>" + "<｜User｜>", + "<｜end▁of▁sentence｜>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen3-1.7b-04-28", + "permaslug": "arliai/qwq-32b-arliai-rpr-v1", "reasoningConfig": { "startToken": "", "endToken": "" @@ -62515,42 +65433,40 @@ export default { } } }, - "modelVariantSlug": "qwen/qwen3-1.7b:free", - "modelVariantPermaslug": "qwen/qwen3-1.7b-04-28:free", - "providerName": "Novita", + "modelVariantSlug": "arliai/qwq-32b-arliai-rpr-v1:free", + "modelVariantPermaslug": "arliai/qwq-32b-arliai-rpr-v1:free", + "providerName": "Chutes", "providerInfo": { - "name": "Novita", - "displayName": "NovitaAI", - "slug": "novita", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Novita", + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.novita.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "NovitaAI", - "providerModelId": "qwen/qwen3-1.7b-fp8", - "providerGroup": "Novita", - "quantization": "fp8", + "providerDisplayName": "Chutes", + "providerModelId": "ArliAI/QwQ-32B-ArliAI-RpR-v1", + "providerGroup": "Chutes", + "quantization": null, "variant": "free", "isFree": true, "canAbort": true, @@ -62571,17 +65487,20 @@ export default { "top_k", "min_p", "repetition_penalty", - "logit_bias" + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true }, - "training": false + "training": true, + "retainsPrompts": true }, "pricing": { "prompt": "0", @@ -62610,108 +65529,99 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 228, - "newest": 15, - "throughputHighToLow": 19, - "latencyLowToHigh": 99, - "pricingLowToHigh": 4, - "pricingHighToLow": 252 + "topWeekly": 229, + "newest": 53, + "throughputHighToLow": 218, + "latencyLowToHigh": 199, + "pricingLowToHigh": 21, + "pricingHighToLow": 269 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { - "slug": "meta-llama/llama-3.2-11b-vision-instruct", - "hfSlug": "meta-llama/Llama-3.2-11B-Vision-Instruct", + "slug": "mistralai/mistral-small-24b-instruct-2501", + "hfSlug": "mistralai/Mistral-Small-24B-Instruct-2501", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-25T00:00:00+00:00", + "createdAt": "2025-01-30T16:43:29.33592+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.2 11B Vision Instruct (free)", - "shortName": "Llama 3.2 11B Vision Instruct (free)", - "author": "meta-llama", - "description": "Llama 3.2 11B Vision is a multimodal model with 11 billion parameters, designed to handle tasks combining visual and textual data. It excels in tasks such as image captioning and visual question answering, bridging the gap between language generation and visual reasoning. Pre-trained on a massive dataset of image-text pairs, it performs well in complex, high-accuracy image analysis.\n\nIts ability to integrate visual understanding with language processing makes it an ideal solution for industries requiring comprehensive visual-linguistic AI applications, such as content creation, AI-driven customer service, and research.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "name": "Mistral: Mistral Small 3 (free)", + "shortName": "Mistral Small 3 (free)", + "author": "mistralai", + "description": "Mistral Small 3 is a 24B-parameter language model optimized for low-latency performance across common AI tasks. Released under the Apache 2.0 license, it features both pre-trained and instruction-tuned versions designed for efficient local deployment.\n\nThe model achieves 81% accuracy on the MMLU benchmark and performs competitively with larger models like Llama 3.3 70B and Qwen 32B, while operating at three times the speed on equivalent hardware. [Read the blog post about the model here.](https://mistral.ai/news/mistral-small-3/)", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 32768, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Mistral", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.2-11b-vision-instruct", + "permaslug": "mistralai/mistral-small-24b-instruct-2501", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "f7538554-912a-4cfe-a919-c5cfa3125617", - "name": "Together | meta-llama/llama-3.2-11b-vision-instruct:free", - "contextLength": 131072, + "id": "697b50b5-f19c-4b8b-b680-a0ca9b2e1f6a", + "name": "Chutes | mistralai/mistral-small-24b-instruct-2501:free", + "contextLength": 32768, "model": { - "slug": "meta-llama/llama-3.2-11b-vision-instruct", - "hfSlug": "meta-llama/Llama-3.2-11B-Vision-Instruct", + "slug": "mistralai/mistral-small-24b-instruct-2501", + "hfSlug": "mistralai/Mistral-Small-24B-Instruct-2501", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-25T00:00:00+00:00", + "createdAt": "2025-01-30T16:43:29.33592+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3.2 11B Vision Instruct", - "shortName": "Llama 3.2 11B Vision Instruct", - "author": "meta-llama", - "description": "Llama 3.2 11B Vision is a multimodal model with 11 billion parameters, designed to handle tasks combining visual and textual data. It excels in tasks such as image captioning and visual question answering, bridging the gap between language generation and visual reasoning. Pre-trained on a massive dataset of image-text pairs, it performs well in complex, high-accuracy image analysis.\n\nIts ability to integrate visual understanding with language processing makes it an ideal solution for industries requiring comprehensive visual-linguistic AI applications, such as content creation, AI-driven customer service, and research.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", + "name": "Mistral: Mistral Small 3", + "shortName": "Mistral Small 3", + "author": "mistralai", + "description": "Mistral Small 3 is a 24B-parameter language model optimized for low-latency performance across common AI tasks. Released under the Apache 2.0 license, it features both pre-trained and instruction-tuned versions designed for efficient local deployment.\n\nThe model achieves 81% accuracy on the MMLU benchmark and performs competitively with larger models like Llama 3.3 70B and Qwen 32B, while operating at three times the speed on equivalent hardware. [Read the blog post about the model here.](https://mistral.ai/news/mistral-small-3/)", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 32768, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Mistral", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3.2-11b-vision-instruct", + "permaslug": "mistralai/mistral-small-24b-instruct-2501", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "meta-llama/llama-3.2-11b-vision-instruct:free", - "modelVariantPermaslug": "meta-llama/llama-3.2-11b-vision-instruct:free", - "providerName": "Together", + "modelVariantSlug": "mistralai/mistral-small-24b-instruct-2501:free", + "modelVariantPermaslug": "mistralai/mistral-small-24b-instruct-2501:free", + "providerName": "Chutes", "providerInfo": { - "name": "Together", - "displayName": "Together", - "slug": "together", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Together", + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, @@ -62719,18 +65629,18 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "Together", - "providerModelId": "meta-llama/Llama-Vision-Free", - "providerGroup": "Together", - "quantization": "fp8", + "providerDisplayName": "Chutes", + "providerModelId": "unsloth/Mistral-Small-24B-Instruct-2501", + "providerGroup": "Chutes", + "quantization": null, "variant": "free", "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 2048, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -62740,23 +65650,24 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "seed", "top_k", + "min_p", "repetition_penalty", + "logprobs", "logit_bias", - "min_p", - "response_format" + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true }, - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true }, "pricing": { "prompt": "0", @@ -62779,63 +65690,66 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 107, - "newest": 199, - "throughputHighToLow": 303, + "topWeekly": 37, + "newest": 131, + "throughputHighToLow": 79, "latencyLowToHigh": 272, - "pricingLowToHigh": 62, - "pricingHighToLow": 224 + "pricingLowToHigh": 49, + "pricingHighToLow": 216 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/Llama-3.2-11B-Vision-Instruct", + "name": "Enfer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://enfer.ai/&size=256", + "slug": "enfer", + "quantization": null, + "context": 28000, + "maxCompletionTokens": 14000, + "providerModelId": "mistralai/mistral-small-24b-instruct-2501", "pricing": { - "prompt": "0.000000049", - "completion": "0.000000049", - "image": "0.00007948", + "prompt": "0.00000006", + "completion": "0.00000012", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" + "logit_bias", + "logprobs" ], - "inputCost": 0.05, - "outputCost": 0.05 + "inputCost": 0.06, + "outputCost": 0.12, + "throughput": 30.0135, + "latency": 6906.5 }, { - "name": "Cloudflare", - "slug": "cloudflare", - "quantization": null, - "context": 131072, + "name": "NextBit", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nextbit256.com/&size=256", + "slug": "nextBit", + "quantization": "bf16", + "context": 32768, "maxCompletionTokens": null, - "providerModelId": "@cf/meta/llama-3.2-11b-vision-instruct", + "providerModelId": "mistral:small3", "pricing": { - "prompt": "0.000000049", - "completion": "0.00000068", - "image": "0.001281", + "prompt": "0.00000007", + "completion": "0.00000013", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -62845,25 +65759,28 @@ export default { "max_tokens", "temperature", "top_p", - "top_k", - "seed", - "repetition_penalty", + "stop", "frequency_penalty", - "presence_penalty" + "presence_penalty", + "response_format", + "structured_outputs" ], - "inputCost": 0.05, - "outputCost": 0.68 + "inputCost": 0.07, + "outputCost": 0.13, + "throughput": 38.119, + "latency": 1667 }, { - "name": "Lambda", - "slug": "lambda", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": 131072, - "providerModelId": "llama3.2-11b-vision-instruct", + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "fp8", + "context": 32768, + "maxCompletionTokens": 16384, + "providerModelId": "mistralai/Mistral-Small-24B-Instruct-2501", "pricing": { - "prompt": "0.00000005", - "completion": "0.00000005", + "prompt": "0.00000007", + "completion": "0.00000014", "image": "0", "request": "0", "webSearch": "0", @@ -62877,25 +65794,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format" + "min_p" ], - "inputCost": 0.05, - "outputCost": 0.05 + "inputCost": 0.07, + "outputCost": 0.14, + "throughput": 77.2655, + "latency": 506 }, { - "name": "inference.net", - "slug": "inferenceNet", - "quantization": "fp16", - "context": 16384, - "maxCompletionTokens": 16384, - "providerModelId": "meta-llama/llama-3.2-11b-instruct/fp-16", + "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", + "slug": "mistral", + "quantization": null, + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "mistral-small-2501", "pricing": { - "prompt": "0.000000055", - "completion": "0.000000055", + "prompt": "0.0000001", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -62903,31 +65823,34 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "min_p", - "logit_bias", - "top_logprobs" + "response_format", + "structured_outputs", + "seed" ], - "inputCost": 0.06, - "outputCost": 0.06 + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 145.2055, + "latency": 260 }, { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": "fp8", + "name": "Ubicloud", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.ubicloud.com/&size=256", + "slug": "ubicloud", + "quantization": "bf16", "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/llama-3.2-11b-vision-instruct", + "maxCompletionTokens": 32768, + "providerModelId": "mistral-small-3", "pricing": { - "prompt": "0.00000006", - "completion": "0.00000006", + "prompt": "0.0000003", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -62941,26 +65864,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", "min_p", - "repetition_penalty", - "logit_bias" + "seed", + "response_format", + "top_k" ], - "inputCost": 0.06, - "outputCost": 0.06 + "inputCost": 0.3, + "outputCost": 0.3, + "throughput": 32.404, + "latency": 762 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", - "quantization": "fp8", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo", + "quantization": null, + "context": 32768, + "maxCompletionTokens": 2048, + "providerModelId": "mistralai/Mistral-Small-24B-Instruct-2501", "pricing": { - "prompt": "0.00000018", - "completion": "0.00000018", - "image": "0.001156", + "prompt": "0.0000008", + "completion": "0.0000008", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -62979,8 +65904,10 @@ export default { "min_p", "response_format" ], - "inputCost": 0.18, - "outputCost": 0.18 + "inputCost": 0.8, + "outputCost": 0.8, + "throughput": 80.305, + "latency": 654.5 } ] }, @@ -63141,16 +66068,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 230, - "newest": 207, + "topWeekly": 231, + "newest": 208, "throughputHighToLow": 315, - "latencyLowToHigh": 307, + "latencyLowToHigh": 305, "pricingLowToHigh": 307, "pricingHighToLow": 12 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", "slug": "openAi", "quantization": "unknown", "context": 128000, @@ -63171,259 +66100,22 @@ export default { "max_tokens" ], "inputCost": 15, - "outputCost": 60 - } - ] - }, - { - "slug": "neversleep/noromaid-20b", - "hfSlug": "NeverSleep/Noromaid-20b-v0.1.1", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-11-26T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Noromaid 20B", - "shortName": "Noromaid 20B", - "author": "neversleep", - "description": "A collab between IkariDev and Undi. This merge is suitable for RP, ERP, and general knowledge.\n\n#merge #uncensored", - "modelVersionGroupId": null, - "contextLength": 8192, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Llama2", - "instructType": "alpaca", - "defaultSystem": null, - "defaultStops": [ - "###", - "" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "neversleep/noromaid-20b", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "2a342807-616e-467d-ace8-10e1cc447a01", - "name": "Mancer | neversleep/noromaid-20b", - "contextLength": 8192, - "model": { - "slug": "neversleep/noromaid-20b", - "hfSlug": "NeverSleep/Noromaid-20b-v0.1.1", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-11-26T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Noromaid 20B", - "shortName": "Noromaid 20B", - "author": "neversleep", - "description": "A collab between IkariDev and Undi. This merge is suitable for RP, ERP, and general knowledge.\n\n#merge #uncensored", - "modelVersionGroupId": null, - "contextLength": 8192, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Llama2", - "instructType": "alpaca", - "defaultSystem": null, - "defaultStops": [ - "###", - "" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "neversleep/noromaid-20b", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "neversleep/noromaid-20b", - "modelVariantPermaslug": "neversleep/noromaid-20b", - "providerName": "Mancer", - "providerInfo": { - "name": "Mancer", - "displayName": "Mancer", - "slug": "mancer", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://mancer.tech/terms", - "privacyPolicyUrl": "https://mancer.tech/privacy", - "paidModels": { - "training": true, - "retainsPrompts": true - } - }, - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "Mancer", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256" - } - }, - "providerDisplayName": "Mancer", - "providerModelId": "noromaid", - "providerGroup": "Mancer", - "quantization": "unknown", - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 2048, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "logit_bias", - "top_k", - "min_p", - "seed", - "top_a" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://mancer.tech/terms", - "privacyPolicyUrl": "https://mancer.tech/privacy", - "paidModels": { - "training": true, - "retainsPrompts": true - }, - "training": true, - "retainsPrompts": true - }, - "pricing": { - "prompt": "0.00000075", - "completion": "0.0000015", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0.25 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 231, - "newest": 293, - "throughputHighToLow": 269, - "latencyLowToHigh": 212, - "pricingLowToHigh": 204, - "pricingHighToLow": 115 - }, - "providers": [ - { - "name": "Mancer", - "slug": "mancer", - "quantization": "unknown", - "context": 8192, - "maxCompletionTokens": 2048, - "providerModelId": "noromaid", - "pricing": { - "prompt": "0.00000075", - "completion": "0.0000015", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0.25 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "logit_bias", - "top_k", - "min_p", - "seed", - "top_a" - ], - "inputCost": 0.75, - "outputCost": 1.5 - }, - { - "name": "Mancer (private)", - "slug": "mancer (private)", - "quantization": "unknown", - "context": 8192, - "maxCompletionTokens": 2048, - "providerModelId": "noromaid", - "pricing": { - "prompt": "0.000001", - "completion": "0.000002", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "logit_bias", - "top_k", - "min_p", - "seed", - "top_a" - ], - "inputCost": 1, - "outputCost": 2 + "outputCost": 60, + "throughput": 121.715, + "latency": 7992 } ] }, { - "slug": "microsoft/phi-3-medium-128k-instruct", - "hfSlug": "microsoft/Phi-3-medium-128k-instruct", + "slug": "meta-llama/llama-guard-3-8b", + "hfSlug": "meta-llama/Llama-Guard-3-8B", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-05-24T00:00:00+00:00", + "createdAt": "2025-02-12T23:01:58.468577+00:00", "hfUpdatedAt": null, - "name": "Microsoft: Phi-3 Medium 128K Instruct", - "shortName": "Phi-3 Medium 128K Instruct", - "author": "microsoft", - "description": "Phi-3 128K Medium is a powerful 14-billion parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. In the MMLU-Pro eval, the model even comes close to a Llama3 70B level of performance.\n\nFor 4k context length, try [Phi-3 Medium 4K](/models/microsoft/phi-3-medium-4k-instruct).", + "name": "Llama Guard 3 8B", + "shortName": "Llama Guard 3 8B", + "author": "meta-llama", + "description": "Llama Guard 3 is a Llama-3.1-8B pretrained model, fine-tuned for content safety classification. Similar to previous versions, it can be used to classify content in both LLM inputs (prompt classification) and in LLM responses (response classification). It acts as an LLM – it generates text in its output that indicates whether a given prompt or response is safe or unsafe, and if unsafe, it also lists the content categories violated.\n\nLlama Guard 3 was aligned to safeguard against the MLCommons standardized hazards taxonomy and designed to support Llama 3.1 capabilities. Specifically, it provides content moderation in 8 languages, and was optimized to support safety and security for search and code interpreter tool calls.\n", "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ @@ -63433,35 +66125,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "phi3", + "group": "Llama3", + "instructType": "none", "defaultSystem": null, - "defaultStops": [ - "<|end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "microsoft/phi-3-medium-128k-instruct", + "permaslug": "meta-llama/llama-guard-3-8b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "1899ce7d-d477-442a-8528-50d8ec66df16", - "name": "Nebius | microsoft/phi-3-medium-128k-instruct", + "id": "b2fbbc15-67f2-4b72-94a0-112f1591f295", + "name": "Nebius | meta-llama/llama-guard-3-8b", "contextLength": 131072, "model": { - "slug": "microsoft/phi-3-medium-128k-instruct", - "hfSlug": "microsoft/Phi-3-medium-128k-instruct", + "slug": "meta-llama/llama-guard-3-8b", + "hfSlug": "meta-llama/Llama-Guard-3-8B", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-05-24T00:00:00+00:00", + "createdAt": "2025-02-12T23:01:58.468577+00:00", "hfUpdatedAt": null, - "name": "Microsoft: Phi-3 Medium 128K Instruct", - "shortName": "Phi-3 Medium 128K Instruct", - "author": "microsoft", - "description": "Phi-3 128K Medium is a powerful 14-billion parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. In the MMLU-Pro eval, the model even comes close to a Llama3 70B level of performance.\n\nFor 4k context length, try [Phi-3 Medium 4K](/models/microsoft/phi-3-medium-4k-instruct).", + "name": "Llama Guard 3 8B", + "shortName": "Llama Guard 3 8B", + "author": "meta-llama", + "description": "Llama Guard 3 is a Llama-3.1-8B pretrained model, fine-tuned for content safety classification. Similar to previous versions, it can be used to classify content in both LLM inputs (prompt classification) and in LLM responses (response classification). It acts as an LLM – it generates text in its output that indicates whether a given prompt or response is safe or unsafe, and if unsafe, it also lists the content categories violated.\n\nLlama Guard 3 was aligned to safeguard against the MLCommons standardized hazards taxonomy and designed to support Llama 3.1 capabilities. Specifically, it provides content moderation in 8 languages, and was optimized to support safety and security for search and code interpreter tool calls.\n", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 0, "inputModalities": [ "text" ], @@ -63469,22 +66158,19 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "phi3", + "group": "Llama3", + "instructType": "none", "defaultSystem": null, - "defaultStops": [ - "<|end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "microsoft/phi-3-medium-128k-instruct", + "permaslug": "meta-llama/llama-guard-3-8b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "microsoft/phi-3-medium-128k-instruct", - "modelVariantPermaslug": "microsoft/phi-3-medium-128k-instruct", + "modelVariantSlug": "meta-llama/llama-guard-3-8b", + "modelVariantPermaslug": "meta-llama/llama-guard-3-8b", "providerName": "Nebius", "providerInfo": { "name": "Nebius", @@ -63517,9 +66203,9 @@ export default { } }, "providerDisplayName": "Nebius AI Studio", - "providerModelId": "microsoft/Phi-3-medium-128k-instruct", + "providerModelId": "meta-llama/Llama-Guard-3-8B", "providerGroup": "Nebius", - "quantization": "fp8", + "quantization": null, "variant": "standard", "isFree": false, "canAbort": false, @@ -63553,8 +66239,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", + "prompt": "0.00000002", + "completion": "0.00000006", "image": "0", "request": "0", "webSearch": "0", @@ -63580,23 +66266,95 @@ export default { }, "sorting": { "topWeekly": 232, - "newest": 254, - "throughputHighToLow": 61, - "latencyLowToHigh": 3, - "pricingLowToHigh": 125, - "pricingHighToLow": 196 + "newest": 115, + "throughputHighToLow": 287, + "latencyLowToHigh": 16, + "pricingLowToHigh": 79, + "pricingHighToLow": 238 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": "fp8", - "context": 131072, + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": null, + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Llama-Guard-3-8B", + "pricing": { + "prompt": "0.00000002", + "completion": "0.00000006", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.02, + "outputCost": 0.06, + "throughput": 17.054, + "latency": 242 + }, + { + "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", + "slug": "deepInfra", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/Llama-Guard-3-8B", + "pricing": { + "prompt": "0.000000055", + "completion": "0.000000055", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "response_format", + "top_k", + "seed", + "min_p" + ], + "inputCost": 0.06, + "outputCost": 0.06, + "throughput": 44.2035, + "latency": 641.5 + }, + { + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": null, + "context": 8192, "maxCompletionTokens": null, - "providerModelId": "microsoft/Phi-3-medium-128k-instruct", + "providerModelId": "meta-llama/Meta-Llama-Guard-3-8B", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", + "prompt": "0.0000002", + "completion": "0.0000002", "image": "0", "request": "0", "webSearch": "0", @@ -63610,25 +66368,88 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", "top_k", + "repetition_penalty", "logit_bias", + "min_p", + "response_format" + ], + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 38.7615, + "latency": 576 + }, + { + "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", + "slug": "groq", + "quantization": null, + "context": 8192, + "maxCompletionTokens": 8192, + "providerModelId": "llama-guard-3-8b", + "pricing": { + "prompt": "0.0000002", + "completion": "0.0000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "top_logprobs", "logprobs", - "top_logprobs" + "logit_bias", + "seed" ], - "inputCost": 0.1, + "inputCost": 0.2, + "outputCost": 0.2 + }, + { + "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", + "slug": "sambaNova", + "quantization": null, + "context": 16384, + "maxCompletionTokens": 4096, + "providerModelId": "Meta-Llama-Guard-3-8B", + "pricing": { + "prompt": "0.0000003", + "completion": "0.0000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 0.3, "outputCost": 0.3 }, { - "name": "Azure", - "slug": "azure", - "quantization": "unknown", - "context": 128000, + "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", + "slug": "cloudflare", + "quantization": null, + "context": 0, "maxCompletionTokens": null, - "providerModelId": "phi3-medium-128k", + "providerModelId": "@cf/meta/llama-guard-3-8b", "pricing": { - "prompt": "0.000001", - "completion": "0.000001", + "prompt": "0.00000048", + "completion": "0.00000003", "image": "0", "request": "0", "webSearch": "0", @@ -63636,92 +66457,100 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", - "top_p" + "top_p", + "top_k", + "seed", + "repetition_penalty", + "frequency_penalty", + "presence_penalty" ], - "inputCost": 1, - "outputCost": 1 + "inputCost": 0.48, + "outputCost": 0.03 } ] }, { - "slug": "qwen/qwen2.5-vl-3b-instruct", - "hfSlug": "Qwen/Qwen2.5-VL-3B-Instruct", + "slug": "neversleep/noromaid-20b", + "hfSlug": "NeverSleep/Noromaid-20b-v0.1.1", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-26T18:42:53.41832+00:00", + "createdAt": "2023-11-26T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen2.5 VL 3B Instruct (free)", - "shortName": "Qwen2.5 VL 3B Instruct (free)", - "author": "qwen", - "description": "Qwen2.5 VL 3B is a multimodal LLM from the Qwen Team with the following key enhancements:\n\n- SoTA understanding of images of various resolution & ratio: Qwen2.5-VL achieves state-of-the-art performance on visual understanding benchmarks, including MathVista, DocVQA, RealWorldQA, MTVQA, etc.\n\n- Agent that can operate your mobiles, robots, etc.: with the abilities of complex reasoning and decision making, Qwen2.5-VL can be integrated with devices like mobile phones, robots, etc., for automatic operation based on visual environment and text instructions.\n\n- Multilingual Support: to serve global users, besides English and Chinese, Qwen2.5-VL now supports the understanding of texts in different languages inside images, including most European languages, Japanese, Korean, Arabic, Vietnamese, etc.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2-vl/) and [GitHub repo](https://github.com/QwenLM/Qwen2-VL).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "name": "Noromaid 20B", + "shortName": "Noromaid 20B", + "author": "neversleep", + "description": "A collab between IkariDev and Undi. This merge is suitable for RP, ERP, and general knowledge.\n\n#merge #uncensored", "modelVersionGroupId": null, - "contextLength": 64000, + "contextLength": 8192, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": null, + "group": "Llama2", + "instructType": "alpaca", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "###", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen2.5-vl-3b-instruct", + "permaslug": "neversleep/noromaid-20b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "751df51c-753b-4849-b95a-bd5281cec4ff", - "name": "Chutes | qwen/qwen2.5-vl-3b-instruct:free", - "contextLength": 64000, + "id": "2a342807-616e-467d-ace8-10e1cc447a01", + "name": "Mancer | neversleep/noromaid-20b", + "contextLength": 8192, "model": { - "slug": "qwen/qwen2.5-vl-3b-instruct", - "hfSlug": "Qwen/Qwen2.5-VL-3B-Instruct", + "slug": "neversleep/noromaid-20b", + "hfSlug": "NeverSleep/Noromaid-20b-v0.1.1", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-26T18:42:53.41832+00:00", + "createdAt": "2023-11-26T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen: Qwen2.5 VL 3B Instruct", - "shortName": "Qwen2.5 VL 3B Instruct", - "author": "qwen", - "description": "Qwen2.5 VL 3B is a multimodal LLM from the Qwen Team with the following key enhancements:\n\n- SoTA understanding of images of various resolution & ratio: Qwen2.5-VL achieves state-of-the-art performance on visual understanding benchmarks, including MathVista, DocVQA, RealWorldQA, MTVQA, etc.\n\n- Agent that can operate your mobiles, robots, etc.: with the abilities of complex reasoning and decision making, Qwen2.5-VL can be integrated with devices like mobile phones, robots, etc., for automatic operation based on visual environment and text instructions.\n\n- Multilingual Support: to serve global users, besides English and Chinese, Qwen2.5-VL now supports the understanding of texts in different languages inside images, including most European languages, Japanese, Korean, Arabic, Vietnamese, etc.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2-vl/) and [GitHub repo](https://github.com/QwenLM/Qwen2-VL).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "name": "Noromaid 20B", + "shortName": "Noromaid 20B", + "author": "neversleep", + "description": "A collab between IkariDev and Undi. This merge is suitable for RP, ERP, and general knowledge.\n\n#merge #uncensored", "modelVersionGroupId": null, - "contextLength": 64000, + "contextLength": 8192, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": null, + "group": "Llama2", + "instructType": "alpaca", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "###", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen2.5-vl-3b-instruct", + "permaslug": "neversleep/noromaid-20b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "qwen/qwen2.5-vl-3b-instruct:free", - "modelVariantPermaslug": "qwen/qwen2.5-vl-3b-instruct:free", - "providerName": "Chutes", + "modelVariantSlug": "neversleep/noromaid-20b", + "modelVariantPermaslug": "neversleep/noromaid-20b", + "providerName": "Mancer", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "Mancer", + "displayName": "Mancer", + "slug": "mancer", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://mancer.tech/terms", + "privacyPolicyUrl": "https://mancer.tech/privacy", "paidModels": { "training": true, "retainsPrompts": true @@ -63731,7 +66560,7 @@ export default { "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Chutes", + "group": "Mancer", "editors": [], "owners": [], "isMultipartSupported": true, @@ -63739,18 +66568,18 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256" } }, - "providerDisplayName": "Chutes", - "providerModelId": "Qwen/Qwen2.5-VL-3B-Instruct", - "providerGroup": "Chutes", - "quantization": null, - "variant": "free", - "isFree": true, + "providerDisplayName": "Mancer", + "providerModelId": "noromaid", + "providerGroup": "Mancer", + "quantization": "unknown", + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 2048, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -63760,18 +66589,18 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "min_p", "repetition_penalty", - "logprobs", "logit_bias", - "top_logprobs" + "top_k", + "min_p", + "seed", + "top_a" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://mancer.tech/terms", + "privacyPolicyUrl": "https://mancer.tech/privacy", "paidModels": { "training": true, "retainsPrompts": true @@ -63780,13 +66609,13 @@ export default { "retainsPrompts": true }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.0000009375", + "completion": "0.0000015", "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", - "discount": 0 + "discount": 0.25 }, "variablePricings": [], "isHidden": false, @@ -63800,33 +66629,106 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 233, - "newest": 71, - "throughputHighToLow": 58, - "latencyLowToHigh": 278, - "pricingLowToHigh": 30, - "pricingHighToLow": 278 + "newest": 293, + "throughputHighToLow": 271, + "latencyLowToHigh": 170, + "pricingLowToHigh": 216, + "pricingHighToLow": 102 }, - "providers": [] + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [ + { + "name": "Mancer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", + "slug": "mancer", + "quantization": "unknown", + "context": 8192, + "maxCompletionTokens": 2048, + "providerModelId": "noromaid", + "pricing": { + "prompt": "0.0000009375", + "completion": "0.0000015", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0.25 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "logit_bias", + "top_k", + "min_p", + "seed", + "top_a" + ], + "inputCost": 0.94, + "outputCost": 1.5, + "throughput": 22.182, + "latency": 1815 + }, + { + "name": "Mancer (private)", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", + "slug": "mancer (private)", + "quantization": "unknown", + "context": 8192, + "maxCompletionTokens": 2048, + "providerModelId": "noromaid", + "pricing": { + "prompt": "0.00000125", + "completion": "0.000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "logit_bias", + "top_k", + "min_p", + "seed", + "top_a" + ], + "inputCost": 1.25, + "outputCost": 2, + "throughput": 21.8425, + "latency": 2930.5 + } + ] }, { - "slug": "deepseek/deepseek-r1-distill-qwen-1.5b", - "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-01-31T12:54:27.964156+00:00", + "slug": "qwen/qwen3-1.7b", + "hfSlug": "Qwen/Qwen3-1.7B", + "updatedAt": "2025-05-12T00:34:59.533129+00:00", + "createdAt": "2025-04-30T16:43:08.565779+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: R1 Distill Qwen 1.5B", - "shortName": "R1 Distill Qwen 1.5B", - "author": "deepseek", - "description": "DeepSeek R1 Distill Qwen 1.5B is a distilled large language model based on [Qwen 2.5 Math 1.5B](https://huggingface.co/Qwen/Qwen2.5-Math-1.5B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It's a very small and efficient model which outperforms [GPT 4o 0513](/openai/gpt-4o-2024-05-13) on Math Benchmarks.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 28.9\n- AIME 2024 cons@64: 52.7\n- MATH-500 pass@1: 83.9\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", - "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", - "contextLength": 131072, + "name": "Qwen: Qwen3 1.7B (free)", + "shortName": "Qwen3 1.7B (free)", + "author": "qwen", + "description": "Qwen3-1.7B is a compact, 1.7 billion parameter dense language model from the Qwen3 series, featuring dual-mode operation for both efficient dialogue (non-thinking) and advanced reasoning (thinking). Despite its small size, it supports 32,768-token contexts and delivers strong multilingual, instruction-following, and agentic capabilities, including tool use and structured output.", + "modelVersionGroupId": null, + "contextLength": 32000, "inputModalities": [ "text" ], @@ -63834,17 +66736,17 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "deepseek-r1", + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" + "<|im_start|>", + "<|im_end|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "deepseek/deepseek-r1-distill-qwen-1.5b", + "permaslug": "qwen/qwen3-1.7b-04-28", "reasoningConfig": { "startToken": "", "endToken": "" @@ -63856,21 +66758,21 @@ export default { } }, "endpoint": { - "id": "85bb5aae-0f60-4233-b275-d4e370685b3a", - "name": "Together | deepseek/deepseek-r1-distill-qwen-1.5b", - "contextLength": 131072, + "id": "390a675c-72d2-4cfd-895e-63653fd3bf29", + "name": "Novita | qwen/qwen3-1.7b-04-28:free", + "contextLength": 32000, "model": { - "slug": "deepseek/deepseek-r1-distill-qwen-1.5b", - "hfSlug": "deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-01-31T12:54:27.964156+00:00", + "slug": "qwen/qwen3-1.7b", + "hfSlug": "Qwen/Qwen3-1.7B", + "updatedAt": "2025-05-12T00:34:59.533129+00:00", + "createdAt": "2025-04-30T16:43:08.565779+00:00", "hfUpdatedAt": null, - "name": "DeepSeek: R1 Distill Qwen 1.5B", - "shortName": "R1 Distill Qwen 1.5B", - "author": "deepseek", - "description": "DeepSeek R1 Distill Qwen 1.5B is a distilled large language model based on [Qwen 2.5 Math 1.5B](https://huggingface.co/Qwen/Qwen2.5-Math-1.5B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It's a very small and efficient model which outperforms [GPT 4o 0513](/openai/gpt-4o-2024-05-13) on Math Benchmarks.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 28.9\n- AIME 2024 cons@64: 52.7\n- MATH-500 pass@1: 83.9\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.", - "modelVersionGroupId": "92d90d33-1fa7-4537-b283-b8199ac69987", - "contextLength": 131072, + "name": "Qwen: Qwen3 1.7B", + "shortName": "Qwen3 1.7B", + "author": "qwen", + "description": "Qwen3-1.7B is a compact, 1.7 billion parameter dense language model from the Qwen3 series, featuring dual-mode operation for both efficient dialogue (non-thinking) and advanced reasoning (thinking). Despite its small size, it supports 32,768-token contexts and delivers strong multilingual, instruction-following, and agentic capabilities, including tool use and structured output.", + "modelVersionGroupId": null, + "contextLength": 32000, "inputModalities": [ "text" ], @@ -63878,17 +66780,17 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "deepseek-r1", + "group": "Qwen3", + "instructType": "qwen3", "defaultSystem": null, "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" + "<|im_start|>", + "<|im_end|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "deepseek/deepseek-r1-distill-qwen-1.5b", + "permaslug": "qwen/qwen3-1.7b-04-28", "reasoningConfig": { "startToken": "", "endToken": "" @@ -63900,20 +66802,19 @@ export default { } } }, - "modelVariantSlug": "deepseek/deepseek-r1-distill-qwen-1.5b", - "modelVariantPermaslug": "deepseek/deepseek-r1-distill-qwen-1.5b", - "providerName": "Together", + "modelVariantSlug": "qwen/qwen3-1.7b:free", + "modelVariantPermaslug": "qwen/qwen3-1.7b-04-28:free", + "providerName": "Novita", "providerInfo": { - "name": "Together", - "displayName": "Together", - "slug": "together", + "name": "Novita", + "displayName": "NovitaAI", + "slug": "novita", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": false + "training": false } }, "headquarters": "US", @@ -63921,26 +66822,27 @@ export default { "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Together", + "group": "Novita", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.novita.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "invertRequired": true } }, - "providerDisplayName": "Together", - "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B", - "providerGroup": "Together", - "quantization": null, - "variant": "standard", - "isFree": false, + "providerDisplayName": "NovitaAI", + "providerModelId": "qwen/qwen3-1.7b-fp8", + "providerGroup": "Novita", + "quantization": "fp8", + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 32768, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -63952,27 +66854,25 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "seed", "top_k", - "repetition_penalty", - "logit_bias", "min_p", - "response_format" + "repetition_penalty", + "logit_bias" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": false + "training": false }, - "training": false, - "retainsPrompts": false + "training": false }, "pricing": { - "prompt": "0.00000018", - "completion": "0.00000018", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -63991,138 +66891,115 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 234, - "newest": 130, - "throughputHighToLow": 2, - "latencyLowToHigh": 27, - "pricingLowToHigh": 142, - "pricingHighToLow": 178 + "newest": 15, + "throughputHighToLow": 175, + "latencyLowToHigh": 123, + "pricingLowToHigh": 4, + "pricingHighToLow": 252 }, - "providers": [ - { - "name": "Together", - "slug": "together", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 32768, - "providerModelId": "deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B", - "pricing": { - "prompt": "0.00000018", - "completion": "0.00000018", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" - ], - "inputCost": 0.18, - "outputCost": 0.18 - } - ] + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", + "providers": [] }, { - "slug": "mistralai/mistral-small-24b-instruct-2501", - "hfSlug": "mistralai/Mistral-Small-24B-Instruct-2501", + "slug": "meta-llama/llama-3.2-11b-vision-instruct", + "hfSlug": "meta-llama/Llama-3.2-11B-Vision-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-01-30T16:43:29.33592+00:00", + "createdAt": "2024-09-25T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral Small 3 (free)", - "shortName": "Mistral Small 3 (free)", - "author": "mistralai", - "description": "Mistral Small 3 is a 24B-parameter language model optimized for low-latency performance across common AI tasks. Released under the Apache 2.0 license, it features both pre-trained and instruction-tuned versions designed for efficient local deployment.\n\nThe model achieves 81% accuracy on the MMLU benchmark and performs competitively with larger models like Llama 3.3 70B and Qwen 32B, while operating at three times the speed on equivalent hardware. [Read the blog post about the model here.](https://mistral.ai/news/mistral-small-3/)", + "name": "Meta: Llama 3.2 11B Vision Instruct (free)", + "shortName": "Llama 3.2 11B Vision Instruct (free)", + "author": "meta-llama", + "description": "Llama 3.2 11B Vision is a multimodal model with 11 billion parameters, designed to handle tasks combining visual and textual data. It excels in tasks such as image captioning and visual question answering, bridging the gap between language generation and visual reasoning. Pre-trained on a massive dataset of image-text pairs, it performs well in complex, high-accuracy image analysis.\n\nIts ability to integrate visual understanding with language processing makes it an ideal solution for industries requiring comprehensive visual-linguistic AI applications, such as content creation, AI-driven customer service, and research.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 131072, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": null, + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-small-24b-instruct-2501", + "permaslug": "meta-llama/llama-3.2-11b-vision-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "697b50b5-f19c-4b8b-b680-a0ca9b2e1f6a", - "name": "Chutes | mistralai/mistral-small-24b-instruct-2501:free", - "contextLength": 32768, + "id": "f7538554-912a-4cfe-a919-c5cfa3125617", + "name": "Together | meta-llama/llama-3.2-11b-vision-instruct:free", + "contextLength": 131072, "model": { - "slug": "mistralai/mistral-small-24b-instruct-2501", - "hfSlug": "mistralai/Mistral-Small-24B-Instruct-2501", + "slug": "meta-llama/llama-3.2-11b-vision-instruct", + "hfSlug": "meta-llama/Llama-3.2-11B-Vision-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-01-30T16:43:29.33592+00:00", + "createdAt": "2024-09-25T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral Small 3", - "shortName": "Mistral Small 3", - "author": "mistralai", - "description": "Mistral Small 3 is a 24B-parameter language model optimized for low-latency performance across common AI tasks. Released under the Apache 2.0 license, it features both pre-trained and instruction-tuned versions designed for efficient local deployment.\n\nThe model achieves 81% accuracy on the MMLU benchmark and performs competitively with larger models like Llama 3.3 70B and Qwen 32B, while operating at three times the speed on equivalent hardware. [Read the blog post about the model here.](https://mistral.ai/news/mistral-small-3/)", + "name": "Meta: Llama 3.2 11B Vision Instruct", + "shortName": "Llama 3.2 11B Vision Instruct", + "author": "meta-llama", + "description": "Llama 3.2 11B Vision is a multimodal model with 11 billion parameters, designed to handle tasks combining visual and textual data. It excels in tasks such as image captioning and visual question answering, bridging the gap between language generation and visual reasoning. Pre-trained on a massive dataset of image-text pairs, it performs well in complex, high-accuracy image analysis.\n\nIts ability to integrate visual understanding with language processing makes it an ideal solution for industries requiring comprehensive visual-linguistic AI applications, such as content creation, AI-driven customer service, and research.\n\nClick here for the [original model card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/MODEL_CARD_VISION.md).\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://www.llama.com/llama3/use-policy/).", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 131072, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": null, + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-small-24b-instruct-2501", + "permaslug": "meta-llama/llama-3.2-11b-vision-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "mistralai/mistral-small-24b-instruct-2501:free", - "modelVariantPermaslug": "mistralai/mistral-small-24b-instruct-2501:free", - "providerName": "Chutes", + "modelVariantSlug": "meta-llama/llama-3.2-11b-vision-instruct:free", + "modelVariantPermaslug": "meta-llama/llama-3.2-11b-vision-instruct:free", + "providerName": "Together", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "Together", + "displayName": "Together", + "slug": "together", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false } }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Chutes", + "group": "Together", "editors": [], "owners": [], "isMultipartSupported": true, @@ -64130,18 +67007,18 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" } }, - "providerDisplayName": "Chutes", - "providerModelId": "unsloth/Mistral-Small-24B-Instruct-2501", - "providerGroup": "Chutes", - "quantization": null, + "providerDisplayName": "Together", + "providerModelId": "meta-llama/Llama-Vision-Free", + "providerGroup": "Together", + "quantization": "fp8", "variant": "free", "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 2048, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -64151,24 +67028,23 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", "top_k", - "min_p", "repetition_penalty", - "logprobs", "logit_bias", - "top_logprobs" + "min_p", + "response_format" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, "pricing": { "prompt": "0", @@ -64191,61 +67067,33 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 35, - "newest": 131, - "throughputHighToLow": 75, - "latencyLowToHigh": 264, - "pricingLowToHigh": 49, - "pricingHighToLow": 216 + "topWeekly": 103, + "newest": 202, + "throughputHighToLow": 54, + "latencyLowToHigh": 157, + "pricingLowToHigh": 62, + "pricingHighToLow": 223 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ - { - "name": "Enfer", - "slug": "enfer", - "quantization": null, - "context": 28000, - "maxCompletionTokens": 14000, - "providerModelId": "mistralai/mistral-small-24b-instruct-2501", - "pricing": { - "prompt": "0.00000006", - "completion": "0.00000012", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "logit_bias", - "logprobs" - ], - "inputCost": 0.06, - "outputCost": 0.12 - }, { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", - "quantization": "fp8", - "context": 32768, + "quantization": "bf16", + "context": 131072, "maxCompletionTokens": 16384, - "providerModelId": "mistralai/Mistral-Small-24B-Instruct-2501", + "providerModelId": "meta-llama/Llama-3.2-11B-Vision-Instruct", "pricing": { - "prompt": "0.00000007", - "completion": "0.00000014", - "image": "0", + "prompt": "0.000000049", + "completion": "0.000000049", + "image": "0.00007948", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -64264,20 +67112,23 @@ export default { "seed", "min_p" ], - "inputCost": 0.07, - "outputCost": 0.14 + "inputCost": 0.05, + "outputCost": 0.05, + "throughput": 12.962, + "latency": 2799 }, { - "name": "NextBit", - "slug": "nextBit", - "quantization": "bf16", - "context": 32768, + "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", + "slug": "cloudflare", + "quantization": null, + "context": 131072, "maxCompletionTokens": null, - "providerModelId": "mistral:small3", + "providerModelId": "@cf/meta/llama-3.2-11b-vision-instruct", "pricing": { - "prompt": "0.00000008", - "completion": "0.00000025", - "image": "0", + "prompt": "0.000000049", + "completion": "0.00000068", + "image": "0.001281", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -64287,25 +67138,28 @@ export default { "max_tokens", "temperature", "top_p", - "stop", + "top_k", + "seed", + "repetition_penalty", "frequency_penalty", - "presence_penalty", - "response_format", - "structured_outputs" + "presence_penalty" ], - "inputCost": 0.08, - "outputCost": 0.25 + "inputCost": 0.05, + "outputCost": 0.68, + "throughput": 36.1265, + "latency": 287 }, { - "name": "Mistral", - "slug": "mistral", - "quantization": null, - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "mistral-small-2501", + "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", + "slug": "lambda", + "quantization": "bf16", + "context": 131072, + "maxCompletionTokens": 131072, + "providerModelId": "llama3.2-11b-vision-instruct", "pricing": { - "prompt": "0.0000001", - "completion": "0.0000003", + "prompt": "0.00000005", + "completion": "0.00000005", "image": "0", "request": "0", "webSearch": "0", @@ -64313,31 +67167,34 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "response_format", - "structured_outputs", - "seed" + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" ], - "inputCost": 0.1, - "outputCost": 0.3 + "inputCost": 0.05, + "outputCost": 0.05, + "throughput": 119.7235, + "latency": 251.5 }, { - "name": "Ubicloud", - "slug": "ubicloud", - "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": 32768, - "providerModelId": "mistral-small-3", + "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", + "slug": "inferenceNet", + "quantization": "fp16", + "context": 16384, + "maxCompletionTokens": 16384, + "providerModelId": "meta-llama/llama-3.2-11b-instruct/fp-16", "pricing": { - "prompt": "0.0000003", - "completion": "0.0000003", + "prompt": "0.000000055", + "completion": "0.000000055", "image": "0", "request": "0", "webSearch": "0", @@ -64351,24 +67208,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "min_p", "seed", - "response_format", - "top_k" + "top_k", + "min_p", + "logit_bias", + "top_logprobs" ], - "inputCost": 0.3, - "outputCost": 0.3 + "inputCost": 0.06, + "outputCost": 0.06, + "throughput": 35.252, + "latency": 1401.5 }, { - "name": "Together", - "slug": "together", - "quantization": null, + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": "fp8", "context": 32768, - "maxCompletionTokens": 2048, - "providerModelId": "mistralai/Mistral-Small-24B-Instruct-2501", + "maxCompletionTokens": null, + "providerModelId": "meta-llama/llama-3.2-11b-vision-instruct", "pricing": { - "prompt": "0.0000008", - "completion": "0.0000008", + "prompt": "0.00000006", + "completion": "0.00000006", "image": "0", "request": "0", "webSearch": "0", @@ -64382,194 +67243,29 @@ export default { "stop", "frequency_penalty", "presence_penalty", + "seed", "top_k", - "repetition_penalty", - "logit_bias", "min_p", - "response_format" - ], - "inputCost": 0.8, - "outputCost": 0.8 - } - ] - }, - { - "slug": "arcee-ai/arcee-blitz", - "hfSlug": "arcee-ai/arcee-blitz", - "updatedAt": "2025-05-05T20:32:10.090817+00:00", - "createdAt": "2025-05-05T18:35:00.177097+00:00", - "hfUpdatedAt": null, - "name": "Arcee AI: Arcee Blitz", - "shortName": "Arcee Blitz", - "author": "arcee-ai", - "description": "Arcee Blitz is a 24 B‑parameter dense model distilled from DeepSeek and built on Mistral architecture for \"everyday\" chat. The distillation‑plus‑refinement pipeline trims compute while keeping DeepSeek‑style reasoning, so Blitz punches above its weight on MMLU, GSM‑8K and BBH compared with other mid‑size open models. With a default 128 k context window and competitive throughput, it serves as a cost‑efficient workhorse for summarization, brainstorming and light code help. Internally, Arcee uses Blitz as the default writer in Conductor pipelines when the heavier Virtuoso line is not required. Users therefore get near‑70 B quality at ~⅓ the latency and price. ", - "modelVersionGroupId": null, - "contextLength": 32768, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "arcee-ai/arcee-blitz", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "b9e4eb0d-b61e-4d25-a26d-71178306bed3", - "name": "Together | arcee-ai/arcee-blitz", - "contextLength": 32768, - "model": { - "slug": "arcee-ai/arcee-blitz", - "hfSlug": "arcee-ai/arcee-blitz", - "updatedAt": "2025-05-05T20:32:10.090817+00:00", - "createdAt": "2025-05-05T18:35:00.177097+00:00", - "hfUpdatedAt": null, - "name": "Arcee AI: Arcee Blitz", - "shortName": "Arcee Blitz", - "author": "arcee-ai", - "description": "Arcee Blitz is a 24 B‑parameter dense model distilled from DeepSeek and built on Mistral architecture for \"everyday\" chat. The distillation‑plus‑refinement pipeline trims compute while keeping DeepSeek‑style reasoning, so Blitz punches above its weight on MMLU, GSM‑8K and BBH compared with other mid‑size open models. With a default 128 k context window and competitive throughput, it serves as a cost‑efficient workhorse for summarization, brainstorming and light code help. Internally, Arcee uses Blitz as the default writer in Conductor pipelines when the heavier Virtuoso line is not required. Users therefore get near‑70 B quality at ~⅓ the latency and price. ", - "modelVersionGroupId": null, - "contextLength": 32768, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" + "repetition_penalty", + "logit_bias" ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "arcee-ai/arcee-blitz", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "arcee-ai/arcee-blitz", - "modelVariantPermaslug": "arcee-ai/arcee-blitz", - "providerName": "Together", - "providerInfo": { - "name": "Together", - "displayName": "Together", - "slug": "together", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", - "paidModels": { - "training": false, - "retainsPrompts": false - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "Together", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" - } - }, - "providerDisplayName": "Together", - "providerModelId": "arcee-ai/arcee-blitz", - "providerGroup": "Together", - "quantization": null, - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", - "paidModels": { - "training": false, - "retainsPrompts": false - }, - "training": false, - "retainsPrompts": false - }, - "pricing": { - "prompt": "0.00000045", - "completion": "0.00000075", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null + "inputCost": 0.06, + "outputCost": 0.06, + "throughput": 60.813, + "latency": 701 }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 236, - "newest": 9, - "throughputHighToLow": 87, - "latencyLowToHigh": 54, - "pricingLowToHigh": 177, - "pricingHighToLow": 141 - }, - "providers": [ { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", - "quantization": null, - "context": 32768, + "quantization": "fp8", + "context": 131072, "maxCompletionTokens": null, - "providerModelId": "arcee-ai/arcee-blitz", + "providerModelId": "meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo", "pricing": { - "prompt": "0.00000045", - "completion": "0.00000075", - "image": "0", + "prompt": "0.00000018", + "completion": "0.00000018", + "image": "0.001156", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -64588,23 +67284,25 @@ export default { "min_p", "response_format" ], - "inputCost": 0.45, - "outputCost": 0.75 + "inputCost": 0.18, + "outputCost": 0.18, + "throughput": 140.829, + "latency": 405 } ] }, { - "slug": "microsoft/phi-4-reasoning", - "hfSlug": "microsoft/Phi-4-reasoning", - "updatedAt": "2025-05-01T20:20:23.543108+00:00", - "createdAt": "2025-05-01T17:41:15.957157+00:00", + "slug": "alpindale/goliath-120b", + "hfSlug": "alpindale/goliath-120b", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2023-11-10T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Microsoft: Phi 4 Reasoning (free)", - "shortName": "Phi 4 Reasoning (free)", - "author": "microsoft", - "description": "Phi-4-reasoning is a 14B parameter dense decoder-only transformer developed by Microsoft, fine-tuned from Phi-4 to enhance complex reasoning capabilities. It uses a combination of supervised fine-tuning on chain-of-thought traces and reinforcement learning, targeting math, science, and code reasoning tasks. With a 32k context window and high inference efficiency, it is optimized for structured responses in a two-part format: reasoning trace followed by a final solution.\n\nThe model achieves strong results on specialized benchmarks such as AIME, OmniMath, and LiveCodeBench, outperforming many larger models in structured reasoning tasks. It is released under the MIT license and intended for use in latency-constrained, English-only environments requiring reliable step-by-step logic. Recommended usage includes ChatML prompts and structured reasoning format for best results.", + "name": "Goliath 120B", + "shortName": "Goliath 120B", + "author": "alpindale", + "description": "A large LLM created by combining two fine-tuned Llama 70B models into one 120B model. Combines Xwin and Euryale.\n\nCredits to\n- [@chargoddard](https://huggingface.co/chargoddard) for developing the framework used to merge the model - [mergekit](https://github.com/cg123/mergekit).\n- [@Undi95](https://huggingface.co/Undi95) for helping with the merge ratios.\n\n#merge", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 6144, "inputModalities": [ "text" ], @@ -64612,32 +67310,35 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Llama2", + "instructType": "airoboros", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "USER:", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "microsoft/phi-4-reasoning-04-30", + "permaslug": "alpindale/goliath-120b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "2cf5cbff-0bdf-44a4-bedb-722ca6e2d536", - "name": "Chutes | microsoft/phi-4-reasoning-04-30:free", - "contextLength": 32768, + "id": "73654edf-bbfe-43bd-9c4c-13b858f00dc8", + "name": "Mancer | alpindale/goliath-120b", + "contextLength": 6144, "model": { - "slug": "microsoft/phi-4-reasoning", - "hfSlug": "microsoft/Phi-4-reasoning", - "updatedAt": "2025-05-01T20:20:23.543108+00:00", - "createdAt": "2025-05-01T17:41:15.957157+00:00", + "slug": "alpindale/goliath-120b", + "hfSlug": "alpindale/goliath-120b", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2023-11-10T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Microsoft: Phi 4 Reasoning", - "shortName": "Phi 4 Reasoning", - "author": "microsoft", - "description": "Phi-4-reasoning is a 14B parameter dense decoder-only transformer developed by Microsoft, fine-tuned from Phi-4 to enhance complex reasoning capabilities. It uses a combination of supervised fine-tuning on chain-of-thought traces and reinforcement learning, targeting math, science, and code reasoning tasks. With a 32k context window and high inference efficiency, it is optimized for structured responses in a two-part format: reasoning trace followed by a final solution.\n\nThe model achieves strong results on specialized benchmarks such as AIME, OmniMath, and LiveCodeBench, outperforming many larger models in structured reasoning tasks. It is released under the MIT license and intended for use in latency-constrained, English-only environments requiring reliable step-by-step logic. Recommended usage includes ChatML prompts and structured reasoning format for best results.", + "name": "Goliath 120B", + "shortName": "Goliath 120B", + "author": "alpindale", + "description": "A large LLM created by combining two fine-tuned Llama 70B models into one 120B model. Combines Xwin and Euryale.\n\nCredits to\n- [@chargoddard](https://huggingface.co/chargoddard) for developing the framework used to merge the model - [mergekit](https://github.com/cg123/mergekit).\n- [@Undi95](https://huggingface.co/Undi95) for helping with the merge ratios.\n\n#merge", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 6144, "inputModalities": [ "text" ], @@ -64645,27 +67346,31 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Llama2", + "instructType": "airoboros", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "USER:", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "microsoft/phi-4-reasoning-04-30", + "permaslug": "alpindale/goliath-120b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "microsoft/phi-4-reasoning:free", - "modelVariantPermaslug": "microsoft/phi-4-reasoning-04-30:free", - "providerName": "Chutes", + "modelVariantSlug": "alpindale/goliath-120b", + "modelVariantPermaslug": "alpindale/goliath-120b", + "providerName": "Mancer", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "Mancer", + "displayName": "Mancer", + "slug": "mancer", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://mancer.tech/terms", + "privacyPolicyUrl": "https://mancer.tech/privacy", "paidModels": { "training": true, "retainsPrompts": true @@ -64675,7 +67380,7 @@ export default { "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Chutes", + "group": "Mancer", "editors": [], "owners": [], "isMultipartSupported": true, @@ -64683,41 +67388,39 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" - } - }, - "providerDisplayName": "Chutes", - "providerModelId": "microsoft/Phi-4-reasoning", - "providerGroup": "Chutes", - "quantization": "bf16", - "variant": "free", - "isFree": true, + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256" + } + }, + "providerDisplayName": "Mancer", + "providerModelId": "goliath-120b", + "providerGroup": "Mancer", + "quantization": "int4", + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 512, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "seed", - "top_k", - "min_p", "repetition_penalty", - "logprobs", "logit_bias", - "top_logprobs" + "top_k", + "min_p", + "seed", + "top_a" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://mancer.tech/terms", + "privacyPolicyUrl": "https://mancer.tech/privacy", "paidModels": { "training": true, "retainsPrompts": true @@ -64726,51 +67429,124 @@ export default { "retainsPrompts": true }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.0000075", + "completion": "0.000009375", "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", - "discount": 0 + "discount": 0.25 }, "variablePricings": [], "isHidden": false, "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 237, - "newest": 12, - "throughputHighToLow": 31, - "latencyLowToHigh": 177, - "pricingLowToHigh": 2, - "pricingHighToLow": 250 + "topWeekly": 236, + "newest": 299, + "throughputHighToLow": 286, + "latencyLowToHigh": 258, + "pricingLowToHigh": 295, + "pricingHighToLow": 23 }, - "providers": [] + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [ + { + "name": "Mancer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", + "slug": "mancer", + "quantization": "int4", + "context": 6144, + "maxCompletionTokens": 512, + "providerModelId": "goliath-120b", + "pricing": { + "prompt": "0.0000075", + "completion": "0.000009375", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0.25 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "logit_bias", + "top_k", + "min_p", + "seed", + "top_a" + ], + "inputCost": 7.5, + "outputCost": 9.38, + "throughput": 16.992, + "latency": 2116 + }, + { + "name": "Mancer (private)", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", + "slug": "mancer (private)", + "quantization": "int4", + "context": 6144, + "maxCompletionTokens": 512, + "providerModelId": "goliath-120b", + "pricing": { + "prompt": "0.00001", + "completion": "0.0000125", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "logit_bias", + "top_k", + "min_p", + "seed", + "top_a" + ], + "inputCost": 10, + "outputCost": 12.5, + "throughput": 20.0485, + "latency": 2045 + } + ] }, { - "slug": "open-r1/olympiccoder-32b", - "hfSlug": "open-r1/OlympicCoder-32B", + "slug": "cognitivecomputations/dolphin3.0-r1-mistral-24b", + "hfSlug": "cognitivecomputations/Dolphin3.0-R1-Mistral-24B", "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-03-15T22:20:28.942272+00:00", + "createdAt": "2025-02-13T16:01:38.017763+00:00", "hfUpdatedAt": null, - "name": "OlympicCoder 32B (free)", - "shortName": "OlympicCoder 32B (free)", - "author": "open-r1", - "description": "OlympicCoder-32B is a high-performing open-source model fine-tuned using the CodeForces-CoTs dataset, containing approximately 100,000 chain-of-thought programming samples. It excels at complex competitive programming benchmarks, such as IOI 2024 and Codeforces-style challenges, frequently surpassing state-of-the-art closed-source models. OlympicCoder-32B provides advanced reasoning, coherent multi-step problem-solving, and robust code generation capabilities, demonstrating significant potential for olympiad-level competitive programming applications.", + "name": "Dolphin3.0 R1 Mistral 24B (free)", + "shortName": "Dolphin3.0 R1 Mistral 24B (free)", + "author": "cognitivecomputations", + "description": "Dolphin 3.0 R1 is the next generation of the Dolphin series of instruct-tuned models. Designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.\n\nThe R1 version has been trained for 3 epochs to reason using 800k reasoning traces from the Dolphin-R1 dataset.\n\nDolphin aims to be a general purpose reasoning instruct model, similar to the models behind ChatGPT, Claude, Gemini.\n\nPart of the [Dolphin 3.0 Collection](https://huggingface.co/collections/cognitivecomputations/dolphin-30-677ab47f73d7ff66743979a3) Curated and trained by [Eric Hartford](https://huggingface.co/ehartford), [Ben Gitter](https://huggingface.co/bigstorm), [BlouseJury](https://huggingface.co/BlouseJury) and [Cognitive Computations](https://huggingface.co/cognitivecomputations)", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ @@ -64790,7 +67566,7 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "open-r1/olympiccoder-32b", + "permaslug": "cognitivecomputations/dolphin3.0-r1-mistral-24b", "reasoningConfig": { "startToken": "", "endToken": "" @@ -64802,19 +67578,19 @@ export default { } }, "endpoint": { - "id": "cf95d8ad-af21-42a1-9fa4-c68d7df33805", - "name": "Chutes | open-r1/olympiccoder-32b:free", + "id": "5ac910b5-4caf-40bc-a45c-f506089228b1", + "name": "Chutes | cognitivecomputations/dolphin3.0-r1-mistral-24b:free", "contextLength": 32768, "model": { - "slug": "open-r1/olympiccoder-32b", - "hfSlug": "open-r1/OlympicCoder-32B", + "slug": "cognitivecomputations/dolphin3.0-r1-mistral-24b", + "hfSlug": "cognitivecomputations/Dolphin3.0-R1-Mistral-24B", "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-03-15T22:20:28.942272+00:00", + "createdAt": "2025-02-13T16:01:38.017763+00:00", "hfUpdatedAt": null, - "name": "OlympicCoder 32B", - "shortName": "OlympicCoder 32B", - "author": "open-r1", - "description": "OlympicCoder-32B is a high-performing open-source model fine-tuned using the CodeForces-CoTs dataset, containing approximately 100,000 chain-of-thought programming samples. It excels at complex competitive programming benchmarks, such as IOI 2024 and Codeforces-style challenges, frequently surpassing state-of-the-art closed-source models. OlympicCoder-32B provides advanced reasoning, coherent multi-step problem-solving, and robust code generation capabilities, demonstrating significant potential for olympiad-level competitive programming applications.", + "name": "Dolphin3.0 R1 Mistral 24B", + "shortName": "Dolphin3.0 R1 Mistral 24B", + "author": "cognitivecomputations", + "description": "Dolphin 3.0 R1 is the next generation of the Dolphin series of instruct-tuned models. Designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.\n\nThe R1 version has been trained for 3 epochs to reason using 800k reasoning traces from the Dolphin-R1 dataset.\n\nDolphin aims to be a general purpose reasoning instruct model, similar to the models behind ChatGPT, Claude, Gemini.\n\nPart of the [Dolphin 3.0 Collection](https://huggingface.co/collections/cognitivecomputations/dolphin-30-677ab47f73d7ff66743979a3) Curated and trained by [Eric Hartford](https://huggingface.co/ehartford), [Ben Gitter](https://huggingface.co/bigstorm), [BlouseJury](https://huggingface.co/BlouseJury) and [Cognitive Computations](https://huggingface.co/cognitivecomputations)", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ @@ -64834,7 +67610,7 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "open-r1/olympiccoder-32b", + "permaslug": "cognitivecomputations/dolphin3.0-r1-mistral-24b", "reasoningConfig": { "startToken": "", "endToken": "" @@ -64846,8 +67622,8 @@ export default { } } }, - "modelVariantSlug": "open-r1/olympiccoder-32b:free", - "modelVariantPermaslug": "open-r1/olympiccoder-32b:free", + "modelVariantSlug": "cognitivecomputations/dolphin3.0-r1-mistral-24b:free", + "modelVariantPermaslug": "cognitivecomputations/dolphin3.0-r1-mistral-24b:free", "providerName": "Chutes", "providerInfo": { "name": "Chutes", @@ -64877,9 +67653,9 @@ export default { } }, "providerDisplayName": "Chutes", - "providerModelId": "open-r1/OlympicCoder-32B", + "providerModelId": "cognitivecomputations/Dolphin3.0-R1-Mistral-24B", "providerGroup": "Chutes", - "quantization": "bf16", + "quantization": null, "variant": "free", "isFree": true, "canAbort": true, @@ -64941,27 +67717,28 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 238, - "newest": 81, - "throughputHighToLow": 145, - "latencyLowToHigh": 269, - "pricingLowToHigh": 36, - "pricingHighToLow": 284 + "topWeekly": 237, + "newest": 113, + "throughputHighToLow": 112, + "latencyLowToHigh": 249, + "pricingLowToHigh": 46, + "pricingHighToLow": 294 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { - "slug": "openai/o1-preview", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-12T00:00:00+00:00", + "slug": "thudm/glm-z1-rumination-32b", + "hfSlug": "THUDM/GLM-Z1-Rumination-32B-0414", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-04-25T17:18:15.30585+00:00", "hfUpdatedAt": null, - "name": "OpenAI: o1-preview", - "shortName": "o1-preview", - "author": "openai", - "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", + "name": "THUDM: GLM Z1 Rumination 32B ", + "shortName": "GLM Z1 Rumination 32B ", + "author": "thudm", + "description": "THUDM: GLM Z1 Rumination 32B is a 32B-parameter deep reasoning model from the GLM-4-Z1 series, optimized for complex, open-ended tasks requiring prolonged deliberation. It builds upon glm-4-32b-0414 with additional reinforcement learning phases and multi-stage alignment strategies, introducing “rumination” capabilities designed to emulate extended cognitive processing. This includes iterative reasoning, multi-hop analysis, and tool-augmented workflows such as search, retrieval, and citation-aware synthesis.\n\nThe model excels in research-style writing, comparative analysis, and intricate question answering. It supports function calling for search and navigation primitives (`search`, `click`, `open`, `finish`), enabling use in agent-style pipelines. Rumination behavior is governed by multi-turn loops with rule-based reward shaping and delayed decision mechanisms, benchmarked against Deep Research frameworks such as OpenAI’s internal alignment stacks. This variant is suitable for scenarios requiring depth over speed.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 32000, "inputModalities": [ "text" ], @@ -64969,32 +67746,43 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, + "group": "Other", + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/o1-preview", - "reasoningConfig": null, - "features": {}, + "permaslug": "thudm/glm-z1-rumination-32b-0414", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "dde17d0c-3752-4967-84d5-a2b8d68a08d1", - "name": "OpenAI | openai/o1-preview", - "contextLength": 128000, + "id": "c61d1c8b-1e6d-4f60-97d8-893f51320223", + "name": "Novita | thudm/glm-z1-rumination-32b-0414", + "contextLength": 32000, "model": { - "slug": "openai/o1-preview", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-12T00:00:00+00:00", + "slug": "thudm/glm-z1-rumination-32b", + "hfSlug": "THUDM/GLM-Z1-Rumination-32B-0414", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-04-25T17:18:15.30585+00:00", "hfUpdatedAt": null, - "name": "OpenAI: o1-preview", - "shortName": "o1-preview", - "author": "openai", - "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", + "name": "THUDM: GLM Z1 Rumination 32B ", + "shortName": "GLM Z1 Rumination 32B ", + "author": "thudm", + "description": "THUDM: GLM Z1 Rumination 32B is a 32B-parameter deep reasoning model from the GLM-4-Z1 series, optimized for complex, open-ended tasks requiring prolonged deliberation. It builds upon glm-4-32b-0414 with additional reinforcement learning phases and multi-stage alignment strategies, introducing “rumination” capabilities designed to emulate extended cognitive processing. This includes iterative reasoning, multi-hop analysis, and tool-augmented workflows such as search, retrieval, and citation-aware synthesis.\n\nThe model excels in research-style writing, comparative analysis, and intricate question answering. It supports function calling for search and navigation primitives (`search`, `click`, `open`, `finish`), enabling use in agent-style pipelines. Rumination behavior is governed by multi-turn loops with rule-based reward shaping and delayed decision mechanisms, benchmarked against Deep Research frameworks such as OpenAI’s internal alignment stacks. This variant is suitable for scenarios requiring depth over speed.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 32000, "inputModalities": [ "text" ], @@ -65002,90 +67790,101 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, + "group": "Other", + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/o1-preview", - "reasoningConfig": null, - "features": {} + "permaslug": "thudm/glm-z1-rumination-32b-0414", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "openai/o1-preview", - "modelVariantPermaslug": "openai/o1-preview", - "providerName": "OpenAI", + "modelVariantSlug": "thudm/glm-z1-rumination-32b", + "modelVariantPermaslug": "thudm/glm-z1-rumination-32b-0414", + "providerName": "Novita", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Novita", + "displayName": "NovitaAI", + "slug": "novita", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "training": false + } }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", + "moderationRequired": false, + "group": "Novita", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": "https://status.novita.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "invertRequired": true } }, - "providerDisplayName": "OpenAI", - "providerModelId": "o1-preview", - "providerGroup": "OpenAI", - "quantization": "unknown", + "providerDisplayName": "NovitaAI", + "providerModelId": "thudm/glm-z1-rumination-32b-0414", + "providerGroup": "Novita", + "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 32768, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", "seed", - "max_tokens" + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, "pricing": { - "prompt": "0.000015", - "completion": "0.00006", + "prompt": "0.00000024", + "completion": "0.00000024", "image": "0", "request": "0", - "inputCacheRead": "0.0000075", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -65095,139 +67894,134 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 239, - "newest": 210, - "throughputHighToLow": 318, - "latencyLowToHigh": 306, - "pricingLowToHigh": 308, - "pricingHighToLow": 11 + "topWeekly": 238, + "newest": 33, + "throughputHighToLow": 263, + "latencyLowToHigh": 239, + "pricingLowToHigh": 156, + "pricingHighToLow": 160 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { - "name": "OpenAI", - "slug": "openAi", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 32768, - "providerModelId": "o1-preview", + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": null, + "context": 32000, + "maxCompletionTokens": null, + "providerModelId": "thudm/glm-z1-rumination-32b-0414", "pricing": { - "prompt": "0.000015", - "completion": "0.00006", + "prompt": "0.00000024", + "completion": "0.00000024", "image": "0", "request": "0", - "inputCacheRead": "0.0000075", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", "seed", - "max_tokens" + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" ], - "inputCost": 15, - "outputCost": 60 + "inputCost": 0.24, + "outputCost": 0.24, + "throughput": 20.503, + "latency": 1572 } ] }, { - "slug": "cognitivecomputations/dolphin3.0-r1-mistral-24b", - "hfSlug": "cognitivecomputations/Dolphin3.0-R1-Mistral-24B", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-02-13T16:01:38.017763+00:00", + "slug": "qwen/qwen-2.5-vl-7b-instruct", + "hfSlug": "Qwen/Qwen2.5-VL-7B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-08-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Dolphin3.0 R1 Mistral 24B (free)", - "shortName": "Dolphin3.0 R1 Mistral 24B (free)", - "author": "cognitivecomputations", - "description": "Dolphin 3.0 R1 is the next generation of the Dolphin series of instruct-tuned models. Designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.\n\nThe R1 version has been trained for 3 epochs to reason using 800k reasoning traces from the Dolphin-R1 dataset.\n\nDolphin aims to be a general purpose reasoning instruct model, similar to the models behind ChatGPT, Claude, Gemini.\n\nPart of the [Dolphin 3.0 Collection](https://huggingface.co/collections/cognitivecomputations/dolphin-30-677ab47f73d7ff66743979a3) Curated and trained by [Eric Hartford](https://huggingface.co/ehartford), [Ben Gitter](https://huggingface.co/bigstorm), [BlouseJury](https://huggingface.co/BlouseJury) and [Cognitive Computations](https://huggingface.co/cognitivecomputations)", + "name": "Qwen: Qwen2.5-VL 7B Instruct (free)", + "shortName": "Qwen2.5-VL 7B Instruct (free)", + "author": "qwen", + "description": "Qwen2.5 VL 7B is a multimodal LLM from the Qwen Team with the following key enhancements:\n\n- SoTA understanding of images of various resolution & ratio: Qwen2.5-VL achieves state-of-the-art performance on visual understanding benchmarks, including MathVista, DocVQA, RealWorldQA, MTVQA, etc.\n\n- Understanding videos of 20min+: Qwen2.5-VL can understand videos over 20 minutes for high-quality video-based question answering, dialog, content creation, etc.\n\n- Agent that can operate your mobiles, robots, etc.: with the abilities of complex reasoning and decision making, Qwen2.5-VL can be integrated with devices like mobile phones, robots, etc., for automatic operation based on visual environment and text instructions.\n\n- Multilingual Support: to serve global users, besides English and Chinese, Qwen2.5-VL now supports the understanding of texts in different languages inside images, including most European languages, Japanese, Korean, Arabic, Vietnamese, etc.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2-vl/) and [GitHub repo](https://github.com/QwenLM/Qwen2-VL).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 64000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "deepseek-r1", + "group": "Qwen", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cognitivecomputations/dolphin3.0-r1-mistral-24b", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "permaslug": "qwen/qwen-2-vl-7b-instruct", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "5ac910b5-4caf-40bc-a45c-f506089228b1", - "name": "Chutes | cognitivecomputations/dolphin3.0-r1-mistral-24b:free", - "contextLength": 32768, + "id": "4f1593d5-297e-433e-af46-3ebc8c84446a", + "name": "Chutes | qwen/qwen-2-vl-7b-instruct:free", + "contextLength": 64000, "model": { - "slug": "cognitivecomputations/dolphin3.0-r1-mistral-24b", - "hfSlug": "cognitivecomputations/Dolphin3.0-R1-Mistral-24B", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-02-13T16:01:38.017763+00:00", + "slug": "qwen/qwen-2.5-vl-7b-instruct", + "hfSlug": "Qwen/Qwen2.5-VL-7B-Instruct", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-08-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Dolphin3.0 R1 Mistral 24B", - "shortName": "Dolphin3.0 R1 Mistral 24B", - "author": "cognitivecomputations", - "description": "Dolphin 3.0 R1 is the next generation of the Dolphin series of instruct-tuned models. Designed to be the ultimate general purpose local model, enabling coding, math, agentic, function calling, and general use cases.\n\nThe R1 version has been trained for 3 epochs to reason using 800k reasoning traces from the Dolphin-R1 dataset.\n\nDolphin aims to be a general purpose reasoning instruct model, similar to the models behind ChatGPT, Claude, Gemini.\n\nPart of the [Dolphin 3.0 Collection](https://huggingface.co/collections/cognitivecomputations/dolphin-30-677ab47f73d7ff66743979a3) Curated and trained by [Eric Hartford](https://huggingface.co/ehartford), [Ben Gitter](https://huggingface.co/bigstorm), [BlouseJury](https://huggingface.co/BlouseJury) and [Cognitive Computations](https://huggingface.co/cognitivecomputations)", + "name": "Qwen: Qwen2.5-VL 7B Instruct", + "shortName": "Qwen2.5-VL 7B Instruct", + "author": "qwen", + "description": "Qwen2.5 VL 7B is a multimodal LLM from the Qwen Team with the following key enhancements:\n\n- SoTA understanding of images of various resolution & ratio: Qwen2.5-VL achieves state-of-the-art performance on visual understanding benchmarks, including MathVista, DocVQA, RealWorldQA, MTVQA, etc.\n\n- Understanding videos of 20min+: Qwen2.5-VL can understand videos over 20 minutes for high-quality video-based question answering, dialog, content creation, etc.\n\n- Agent that can operate your mobiles, robots, etc.: with the abilities of complex reasoning and decision making, Qwen2.5-VL can be integrated with devices like mobile phones, robots, etc., for automatic operation based on visual environment and text instructions.\n\n- Multilingual Support: to serve global users, besides English and Chinese, Qwen2.5-VL now supports the understanding of texts in different languages inside images, including most European languages, Japanese, Korean, Arabic, Vietnamese, etc.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2-vl/) and [GitHub repo](https://github.com/QwenLM/Qwen2-VL).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "deepseek-r1", + "group": "Qwen", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cognitivecomputations/dolphin3.0-r1-mistral-24b", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "permaslug": "qwen/qwen-2-vl-7b-instruct", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "cognitivecomputations/dolphin3.0-r1-mistral-24b:free", - "modelVariantPermaslug": "cognitivecomputations/dolphin3.0-r1-mistral-24b:free", + "modelVariantSlug": "qwen/qwen-2.5-vl-7b-instruct:free", + "modelVariantPermaslug": "qwen/qwen-2-vl-7b-instruct:free", "providerName": "Chutes", "providerInfo": { "name": "Chutes", @@ -65257,22 +68051,20 @@ export default { } }, "providerDisplayName": "Chutes", - "providerModelId": "cognitivecomputations/Dolphin3.0-R1-Mistral-24B", + "providerModelId": "Qwen/Qwen2.5-VL-7B-Instruct", "providerGroup": "Chutes", "quantization": null, "variant": "free", "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 64000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -65304,42 +68096,154 @@ export default { "internalReasoning": "0", "discount": 0 }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": true, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 115, + "newest": 214, + "throughputHighToLow": 80, + "latencyLowToHigh": 153, + "pricingLowToHigh": 65, + "pricingHighToLow": 171 + }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", + "providers": [ + { + "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", + "slug": "hyperbolic", + "quantization": "bf16", + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "Qwen/Qwen2.5-VL-7B-Instruct", + "pricing": { + "prompt": "0.0000002", + "completion": "0.0000002", + "image": "0.0001445", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "logprobs", + "top_logprobs", + "seed", + "logit_bias", + "top_k", + "min_p", + "repetition_penalty" + ], + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 49.225, + "latency": 2066 + }, + { + "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", + "slug": "inferenceNet", + "quantization": "bf16", + "context": 128000, + "maxCompletionTokens": 8192, + "providerModelId": "qwen/qwen2.5-7b-instruct/bf-16", + "pricing": { + "prompt": "0.0000002", + "completion": "0.0000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "logit_bias", + "top_logprobs" + ], + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 58.101, + "latency": 3841 }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 240, - "newest": 113, - "throughputHighToLow": 118, - "latencyLowToHigh": 249, - "pricingLowToHigh": 46, - "pricingHighToLow": 294 - }, - "providers": [] + { + "name": "kluster.ai", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.kluster.ai/&size=256", + "slug": "klusterAi", + "quantization": null, + "context": 32768, + "maxCompletionTokens": 32768, + "providerModelId": "Qwen/Qwen2.5-VL-7B-Instruct", + "pricing": { + "prompt": "0.0000003", + "completion": "0.0000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "logit_bias", + "logprobs", + "top_logprobs", + "min_p", + "seed" + ], + "inputCost": 0.3, + "outputCost": 0.3, + "throughput": 39.5905, + "latency": 2843.5 + } + ] }, { - "slug": "thudm/glm-z1-32b", - "hfSlug": "THUDM/GLM-Z1-32B-0414", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-04-17T21:09:08.005794+00:00", + "slug": "microsoft/phi-4-reasoning", + "hfSlug": "microsoft/Phi-4-reasoning", + "updatedAt": "2025-05-01T20:20:23.543108+00:00", + "createdAt": "2025-05-01T17:41:15.957157+00:00", "hfUpdatedAt": null, - "name": "THUDM: GLM Z1 32B (free)", - "shortName": "GLM Z1 32B (free)", - "author": "thudm", - "description": "GLM-Z1-32B-0414 is an enhanced reasoning variant of GLM-4-32B, built for deep mathematical, logical, and code-oriented problem solving. It applies extended reinforcement learning—both task-specific and general pairwise preference-based—to improve performance on complex multi-step tasks. Compared to the base GLM-4-32B model, Z1 significantly boosts capabilities in structured reasoning and formal domains.\n\nThe model supports enforced “thinking” steps via prompt engineering and offers improved coherence for long-form outputs. It’s optimized for use in agentic workflows, and includes support for long context (via YaRN), JSON tool calling, and fine-grained sampling configuration for stable inference. Ideal for use cases requiring deliberate, multi-step reasoning or formal derivations.", + "name": "Microsoft: Phi 4 Reasoning (free)", + "shortName": "Phi 4 Reasoning (free)", + "author": "microsoft", + "description": "Phi-4-reasoning is a 14B parameter dense decoder-only transformer developed by Microsoft, fine-tuned from Phi-4 to enhance complex reasoning capabilities. It uses a combination of supervised fine-tuning on chain-of-thought traces and reinforcement learning, targeting math, science, and code reasoning tasks. With a 32k context window and high inference efficiency, it is optimized for structured responses in a two-part format: reasoning trace followed by a final solution.\n\nThe model achieves strong results on specialized benchmarks such as AIME, OmniMath, and LiveCodeBench, outperforming many larger models in structured reasoning tasks. It is released under the MIT license and intended for use in latency-constrained, English-only environments requiring reliable step-by-step logic. Recommended usage includes ChatML prompts and structured reasoning format for best results.", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ @@ -65350,40 +68254,29 @@ export default { ], "hasTextOutput": true, "group": "Other", - "instructType": "deepseek-r1", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "thudm/glm-z1-32b-0414", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "permaslug": "microsoft/phi-4-reasoning-04-30", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "e22b972c-ea09-46e6-8028-89e654397b03", - "name": "Chutes | thudm/glm-z1-32b-0414:free", + "id": "2cf5cbff-0bdf-44a4-bedb-722ca6e2d536", + "name": "Chutes | microsoft/phi-4-reasoning-04-30:free", "contextLength": 32768, "model": { - "slug": "thudm/glm-z1-32b", - "hfSlug": "THUDM/GLM-Z1-32B-0414", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-04-17T21:09:08.005794+00:00", + "slug": "microsoft/phi-4-reasoning", + "hfSlug": "microsoft/Phi-4-reasoning", + "updatedAt": "2025-05-01T20:20:23.543108+00:00", + "createdAt": "2025-05-01T17:41:15.957157+00:00", "hfUpdatedAt": null, - "name": "THUDM: GLM Z1 32B", - "shortName": "GLM Z1 32B", - "author": "thudm", - "description": "GLM-Z1-32B-0414 is an enhanced reasoning variant of GLM-4-32B, built for deep mathematical, logical, and code-oriented problem solving. It applies extended reinforcement learning—both task-specific and general pairwise preference-based—to improve performance on complex multi-step tasks. Compared to the base GLM-4-32B model, Z1 significantly boosts capabilities in structured reasoning and formal domains.\n\nThe model supports enforced “thinking” steps via prompt engineering and offers improved coherence for long-form outputs. It’s optimized for use in agentic workflows, and includes support for long context (via YaRN), JSON tool calling, and fine-grained sampling configuration for stable inference. Ideal for use cases requiring deliberate, multi-step reasoning or formal derivations.", + "name": "Microsoft: Phi 4 Reasoning", + "shortName": "Phi 4 Reasoning", + "author": "microsoft", + "description": "Phi-4-reasoning is a 14B parameter dense decoder-only transformer developed by Microsoft, fine-tuned from Phi-4 to enhance complex reasoning capabilities. It uses a combination of supervised fine-tuning on chain-of-thought traces and reinforcement learning, targeting math, science, and code reasoning tasks. With a 32k context window and high inference efficiency, it is optimized for structured responses in a two-part format: reasoning trace followed by a final solution.\n\nThe model achieves strong results on specialized benchmarks such as AIME, OmniMath, and LiveCodeBench, outperforming many larger models in structured reasoning tasks. It is released under the MIT license and intended for use in latency-constrained, English-only environments requiring reliable step-by-step logic. Recommended usage includes ChatML prompts and structured reasoning format for best results.", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ @@ -65394,29 +68287,18 @@ export default { ], "hasTextOutput": true, "group": "Other", - "instructType": "deepseek-r1", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "thudm/glm-z1-32b-0414", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "permaslug": "microsoft/phi-4-reasoning-04-30", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "thudm/glm-z1-32b:free", - "modelVariantPermaslug": "thudm/glm-z1-32b-0414:free", + "modelVariantSlug": "microsoft/phi-4-reasoning:free", + "modelVariantPermaslug": "microsoft/phi-4-reasoning-04-30:free", "providerName": "Chutes", "providerInfo": { "name": "Chutes", @@ -65446,9 +68328,9 @@ export default { } }, "providerDisplayName": "Chutes", - "providerModelId": "THUDM/GLM-Z1-32B-0414", + "providerModelId": "microsoft/Phi-4-reasoning", "providerGroup": "Chutes", - "quantization": null, + "quantization": "bf16", "variant": "free", "isFree": true, "canAbort": true, @@ -65511,62 +68393,28 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 201, - "newest": 37, - "throughputHighToLow": 179, - "latencyLowToHigh": 220, - "pricingLowToHigh": 18, - "pricingHighToLow": 161 + "topWeekly": 240, + "newest": 12, + "throughputHighToLow": 32, + "latencyLowToHigh": 173, + "pricingLowToHigh": 2, + "pricingHighToLow": 250 }, - "providers": [ - { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": null, - "context": 32000, - "maxCompletionTokens": null, - "providerModelId": "thudm/glm-z1-32b-0414", - "pricing": { - "prompt": "0.00000024", - "completion": "0.00000024", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" - ], - "inputCost": 0.24, - "outputCost": 0.24 - } - ] + "authorIcon": "https://openrouter.ai/images/icons/Microsoft.svg", + "providers": [] }, { - "slug": "meta-llama/llama-guard-3-8b", - "hfSlug": "meta-llama/Llama-Guard-3-8B", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-12T23:01:58.468577+00:00", + "slug": "arcee-ai/arcee-blitz", + "hfSlug": "arcee-ai/arcee-blitz", + "updatedAt": "2025-05-05T20:32:10.090817+00:00", + "createdAt": "2025-05-05T18:35:00.177097+00:00", "hfUpdatedAt": null, - "name": "Llama Guard 3 8B", - "shortName": "Llama Guard 3 8B", - "author": "meta-llama", - "description": "Llama Guard 3 is a Llama-3.1-8B pretrained model, fine-tuned for content safety classification. Similar to previous versions, it can be used to classify content in both LLM inputs (prompt classification) and in LLM responses (response classification). It acts as an LLM – it generates text in its output that indicates whether a given prompt or response is safe or unsafe, and if unsafe, it also lists the content categories violated.\n\nLlama Guard 3 was aligned to safeguard against the MLCommons standardized hazards taxonomy and designed to support Llama 3.1 capabilities. Specifically, it provides content moderation in 8 languages, and was optimized to support safety and security for search and code interpreter tool calls.\n", + "name": "Arcee AI: Arcee Blitz", + "shortName": "Arcee Blitz", + "author": "arcee-ai", + "description": "Arcee Blitz is a 24 B‑parameter dense model distilled from DeepSeek and built on Mistral architecture for \"everyday\" chat. The distillation‑plus‑refinement pipeline trims compute while keeping DeepSeek‑style reasoning, so Blitz punches above its weight on MMLU, GSM‑8K and BBH compared with other mid‑size open models. With a default 128 k context window and competitive throughput, it serves as a cost‑efficient workhorse for summarization, brainstorming and light code help. Internally, Arcee uses Blitz as the default writer in Conductor pipelines when the heavier Virtuoso line is not required. Users therefore get near‑70 B quality at ~⅓ the latency and price. ", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -65574,32 +68422,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "none", + "group": "Other", + "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-guard-3-8b", + "permaslug": "arcee-ai/arcee-blitz", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "b2fbbc15-67f2-4b72-94a0-112f1591f295", - "name": "Nebius | meta-llama/llama-guard-3-8b", - "contextLength": 131072, + "id": "b9e4eb0d-b61e-4d25-a26d-71178306bed3", + "name": "Together | arcee-ai/arcee-blitz", + "contextLength": 32768, "model": { - "slug": "meta-llama/llama-guard-3-8b", - "hfSlug": "meta-llama/Llama-Guard-3-8B", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-12T23:01:58.468577+00:00", + "slug": "arcee-ai/arcee-blitz", + "hfSlug": "arcee-ai/arcee-blitz", + "updatedAt": "2025-05-05T20:32:10.090817+00:00", + "createdAt": "2025-05-05T18:35:00.177097+00:00", "hfUpdatedAt": null, - "name": "Llama Guard 3 8B", - "shortName": "Llama Guard 3 8B", - "author": "meta-llama", - "description": "Llama Guard 3 is a Llama-3.1-8B pretrained model, fine-tuned for content safety classification. Similar to previous versions, it can be used to classify content in both LLM inputs (prompt classification) and in LLM responses (response classification). It acts as an LLM – it generates text in its output that indicates whether a given prompt or response is safe or unsafe, and if unsafe, it also lists the content categories violated.\n\nLlama Guard 3 was aligned to safeguard against the MLCommons standardized hazards taxonomy and designed to support Llama 3.1 capabilities. Specifically, it provides content moderation in 8 languages, and was optimized to support safety and security for search and code interpreter tool calls.\n", + "name": "Arcee AI: Arcee Blitz", + "shortName": "Arcee Blitz", + "author": "arcee-ai", + "description": "Arcee Blitz is a 24 B‑parameter dense model distilled from DeepSeek and built on Mistral architecture for \"everyday\" chat. The distillation‑plus‑refinement pipeline trims compute while keeping DeepSeek‑style reasoning, so Blitz punches above its weight on MMLU, GSM‑8K and BBH compared with other mid‑size open models. With a default 128 k context window and competitive throughput, it serves as a cost‑efficient workhorse for summarization, brainstorming and light code help. Internally, Arcee uses Blitz as the default writer in Conductor pipelines when the heavier Virtuoso line is not required. Users therefore get near‑70 B quality at ~⅓ the latency and price. ", "modelVersionGroupId": null, - "contextLength": 0, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -65607,39 +68455,39 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "none", + "group": "Other", + "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-guard-3-8b", + "permaslug": "arcee-ai/arcee-blitz", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "meta-llama/llama-guard-3-8b", - "modelVariantPermaslug": "meta-llama/llama-guard-3-8b", - "providerName": "Nebius", + "modelVariantSlug": "arcee-ai/arcee-blitz", + "modelVariantPermaslug": "arcee-ai/arcee-blitz", + "providerName": "Together", "providerInfo": { - "name": "Nebius", - "displayName": "Nebius AI Studio", - "slug": "nebius", + "name": "Together", + "displayName": "Together", + "slug": "together", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false } }, - "headquarters": "DE", + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, + "isAbortable": true, "moderationRequired": false, - "group": "Nebius", + "group": "Together", "editors": [], "owners": [], "isMultipartSupported": true, @@ -65647,17 +68495,16 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" } }, - "providerDisplayName": "Nebius AI Studio", - "providerModelId": "meta-llama/Llama-Guard-3-8B", - "providerGroup": "Nebius", + "providerDisplayName": "Together", + "providerModelId": "arcee-ai/arcee-blitz", + "providerGroup": "Together", "quantization": null, "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, "maxPromptImages": null, @@ -65669,17 +68516,17 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", "top_k", + "repetition_penalty", "logit_bias", - "logprobs", - "top_logprobs" + "min_p", + "response_format" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -65688,8 +68535,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000002", - "completion": "0.00000006", + "prompt": "0.00000045", + "completion": "0.00000075", "image": "0", "request": "0", "webSearch": "0", @@ -65714,88 +68561,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 242, - "newest": 115, - "throughputHighToLow": 289, - "latencyLowToHigh": 20, - "pricingLowToHigh": 79, - "pricingHighToLow": 238 + "topWeekly": 241, + "newest": 9, + "throughputHighToLow": 85, + "latencyLowToHigh": 75, + "pricingLowToHigh": 177, + "pricingHighToLow": 141 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://arcee.ai/\\u0026size=256", "providers": [ - { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": null, - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-Guard-3-8B", - "pricing": { - "prompt": "0.00000002", - "completion": "0.00000006", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "logit_bias", - "logprobs", - "top_logprobs" - ], - "inputCost": 0.02, - "outputCost": 0.06 - }, - { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-Guard-3-8B", - "pricing": { - "prompt": "0.000000055", - "completion": "0.000000055", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "inputCost": 0.06, - "outputCost": 0.06 - }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": null, - "context": 8192, + "context": 32768, "maxCompletionTokens": null, - "providerModelId": "meta-llama/Meta-Llama-Guard-3-8B", + "providerModelId": "arcee-ai/arcee-blitz", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", + "prompt": "0.00000045", + "completion": "0.00000075", "image": "0", "request": "0", "webSearch": "0", @@ -65815,95 +68600,10 @@ export default { "min_p", "response_format" ], - "inputCost": 0.2, - "outputCost": 0.2 - }, - { - "name": "Groq", - "slug": "groq", - "quantization": null, - "context": 8192, - "maxCompletionTokens": 8192, - "providerModelId": "llama-guard-3-8b", - "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "top_logprobs", - "logprobs", - "logit_bias", - "seed" - ], - "inputCost": 0.2, - "outputCost": 0.2 - }, - { - "name": "SambaNova", - "slug": "sambaNova", - "quantization": null, - "context": 16384, - "maxCompletionTokens": 4096, - "providerModelId": "Meta-Llama-Guard-3-8B", - "pricing": { - "prompt": "0.0000003", - "completion": "0.0000003", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "top_k", - "stop" - ], - "inputCost": 0.3, - "outputCost": 0.3 - }, - { - "name": "Cloudflare", - "slug": "cloudflare", - "quantization": null, - "context": 0, - "maxCompletionTokens": null, - "providerModelId": "@cf/meta/llama-guard-3-8b", - "pricing": { - "prompt": "0.00000048", - "completion": "0.00000003", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "top_k", - "seed", - "repetition_penalty", - "frequency_penalty", - "presence_penalty" - ], - "inputCost": 0.48, - "outputCost": 0.03 + "inputCost": 0.45, + "outputCost": 0.75, + "throughput": 88.759, + "latency": 403 } ] }, @@ -66071,16 +68771,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 243, + "topWeekly": 242, "newest": 165, - "throughputHighToLow": 301, + "throughputHighToLow": 297, "latencyLowToHigh": 308, "pricingLowToHigh": 281, "pricingHighToLow": 36 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", "slug": "featherless", "quantization": "fp8", "context": 16384, @@ -66108,7 +68810,238 @@ export default { "seed" ], "inputCost": 4, - "outputCost": 6 + "outputCost": 6, + "throughput": 13.789, + "latency": 9296 + } + ] + }, + { + "slug": "thudm/glm-z1-32b", + "hfSlug": "THUDM/GLM-Z1-32B-0414", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-04-17T21:09:08.005794+00:00", + "hfUpdatedAt": null, + "name": "THUDM: GLM Z1 32B (free)", + "shortName": "GLM Z1 32B (free)", + "author": "thudm", + "description": "GLM-Z1-32B-0414 is an enhanced reasoning variant of GLM-4-32B, built for deep mathematical, logical, and code-oriented problem solving. It applies extended reinforcement learning—both task-specific and general pairwise preference-based—to improve performance on complex multi-step tasks. Compared to the base GLM-4-32B model, Z1 significantly boosts capabilities in structured reasoning and formal domains.\n\nThe model supports enforced “thinking” steps via prompt engineering and offers improved coherence for long-form outputs. It’s optimized for use in agentic workflows, and includes support for long context (via YaRN), JSON tool calling, and fine-grained sampling configuration for stable inference. Ideal for use cases requiring deliberate, multi-step reasoning or formal derivations.", + "modelVersionGroupId": null, + "contextLength": 32768, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Other", + "instructType": "deepseek-r1", + "defaultSystem": null, + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "thudm/glm-z1-32b-0414", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, + "endpoint": { + "id": "e22b972c-ea09-46e6-8028-89e654397b03", + "name": "Chutes | thudm/glm-z1-32b-0414:free", + "contextLength": 32768, + "model": { + "slug": "thudm/glm-z1-32b", + "hfSlug": "THUDM/GLM-Z1-32B-0414", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-04-17T21:09:08.005794+00:00", + "hfUpdatedAt": null, + "name": "THUDM: GLM Z1 32B", + "shortName": "GLM Z1 32B", + "author": "thudm", + "description": "GLM-Z1-32B-0414 is an enhanced reasoning variant of GLM-4-32B, built for deep mathematical, logical, and code-oriented problem solving. It applies extended reinforcement learning—both task-specific and general pairwise preference-based—to improve performance on complex multi-step tasks. Compared to the base GLM-4-32B model, Z1 significantly boosts capabilities in structured reasoning and formal domains.\n\nThe model supports enforced “thinking” steps via prompt engineering and offers improved coherence for long-form outputs. It’s optimized for use in agentic workflows, and includes support for long context (via YaRN), JSON tool calling, and fine-grained sampling configuration for stable inference. Ideal for use cases requiring deliberate, multi-step reasoning or formal derivations.", + "modelVersionGroupId": null, + "contextLength": 32768, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Other", + "instructType": "deepseek-r1", + "defaultSystem": null, + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "thudm/glm-z1-32b-0414", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } + }, + "modelVariantSlug": "thudm/glm-z1-32b:free", + "modelVariantPermaslug": "thudm/glm-z1-32b-0414:free", + "providerName": "Chutes", + "providerInfo": { + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://chutes.ai/tos", + "paidModels": { + "training": true, + "retainsPrompts": true + } + }, + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "Chutes", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + } + }, + "providerDisplayName": "Chutes", + "providerModelId": "THUDM/GLM-Z1-32B-0414", + "providerGroup": "Chutes", + "quantization": null, + "variant": "free", + "isFree": true, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://chutes.ai/tos", + "paidModels": { + "training": true, + "retainsPrompts": true + }, + "training": true, + "retainsPrompts": true + }, + "pricing": { + "prompt": "0", + "completion": "0", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": true, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 214, + "newest": 37, + "throughputHighToLow": 181, + "latencyLowToHigh": 215, + "pricingLowToHigh": 18, + "pricingHighToLow": 161 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [ + { + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": null, + "context": 32000, + "maxCompletionTokens": null, + "providerModelId": "thudm/glm-z1-32b-0414", + "pricing": { + "prompt": "0.00000024", + "completion": "0.00000024", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" + ], + "inputCost": 0.24, + "outputCost": 0.24, + "throughput": 21.073, + "latency": 1632 } ] }, @@ -66263,14 +69196,16 @@ export default { "sorting": { "topWeekly": 45, "newest": 197, - "throughputHighToLow": 3, - "latencyLowToHigh": 17, + "throughputHighToLow": 1, + "latencyLowToHigh": 11, "pricingLowToHigh": 61, "pricingHighToLow": 245 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "bf16", "context": 131072, @@ -66299,10 +69234,13 @@ export default { "min_p" ], "inputCost": 0.01, - "outputCost": 0.02 + "outputCost": 0.02, + "throughput": 111.9425, + "latency": 225 }, { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 131072, @@ -66331,10 +69269,13 @@ export default { "top_logprobs" ], "inputCost": 0.01, - "outputCost": 0.02 + "outputCost": 0.02, + "throughput": 111.2985, + "latency": 245 }, { "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", "slug": "lambda", "quantization": "bf16", "context": 131072, @@ -66363,10 +69304,13 @@ export default { "response_format" ], "inputCost": 0.01, - "outputCost": 0.02 + "outputCost": 0.02, + "throughput": 304.932, + "latency": 690 }, { "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", "slug": "inferenceNet", "quantization": "fp16", "context": 16384, @@ -66395,10 +69339,13 @@ export default { "top_logprobs" ], "inputCost": 0.02, - "outputCost": 0.02 + "outputCost": 0.02, + "throughput": 85.349, + "latency": 928 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "bf16", "context": 32768, @@ -66429,10 +69376,13 @@ export default { "logit_bias" ], "inputCost": 0.03, - "outputCost": 0.05 + "outputCost": 0.05, + "throughput": 108.3015, + "latency": 561 }, { "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", "slug": "cloudflare", "quantization": null, "context": 128000, @@ -66458,10 +69408,13 @@ export default { "presence_penalty" ], "inputCost": 0.05, - "outputCost": 0.34 + "outputCost": 0.34, + "throughput": 157.464, + "latency": 575 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "fp8", "context": 131072, @@ -66490,10 +69443,13 @@ export default { "response_format" ], "inputCost": 0.06, - "outputCost": 0.06 + "outputCost": 0.06, + "throughput": 162.995, + "latency": 551 }, { "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", "slug": "sambaNova", "quantization": "bf16", "context": 4096, @@ -66516,10 +69472,13 @@ export default { "stop" ], "inputCost": 0.08, - "outputCost": 0.16 + "outputCost": 0.16, + "throughput": 3051.282, + "latency": 653.5 }, { "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", "slug": "hyperbolic", "quantization": "fp8", "context": 131072, @@ -66550,7 +69509,9 @@ export default { "repetition_penalty" ], "inputCost": 0.1, - "outputCost": 0.1 + "outputCost": 0.1, + "throughput": 113.333, + "latency": 1108 } ] }, @@ -66701,14 +69662,16 @@ export default { "sorting": { "topWeekly": 245, "newest": 122, - "throughputHighToLow": 191, - "latencyLowToHigh": 202, - "pricingLowToHigh": 149, - "pricingHighToLow": 167 + "throughputHighToLow": 197, + "latencyLowToHigh": 207, + "pricingLowToHigh": 146, + "pricingHighToLow": 170 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://www.aionlabs.ai/\\u0026size=256", "providers": [ { "name": "AionLabs", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.aionlabs.ai/&size=256", "slug": "aionLabs", "quantization": "unknown", "context": 32768, @@ -66729,7 +69692,204 @@ export default { "top_p" ], "inputCost": 0.2, - "outputCost": 0.2 + "outputCost": 0.2, + "throughput": 43.069, + "latency": 1431.5 + } + ] + }, + { + "slug": "openai/o1-preview", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-09-12T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "OpenAI: o1-preview", + "shortName": "o1-preview", + "author": "openai", + "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", + "modelVersionGroupId": null, + "contextLength": 128000, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "GPT", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "openai/o1-preview", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "dde17d0c-3752-4967-84d5-a2b8d68a08d1", + "name": "OpenAI | openai/o1-preview", + "contextLength": 128000, + "model": { + "slug": "openai/o1-preview", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-09-12T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "OpenAI: o1-preview", + "shortName": "o1-preview", + "author": "openai", + "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", + "modelVersionGroupId": null, + "contextLength": 128000, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "GPT", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "openai/o1-preview", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "openai/o1-preview", + "modelVariantPermaslug": "openai/o1-preview", + "providerName": "OpenAI", + "providerInfo": { + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": true, + "group": "OpenAI", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": "https://status.openai.com/", + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/OpenAI.svg", + "invertRequired": true + } + }, + "providerDisplayName": "OpenAI", + "providerModelId": "o1-preview", + "providerGroup": "OpenAI", + "quantization": "unknown", + "variant": "standard", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": 32768, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "seed", + "max_tokens" + ], + "isByok": false, + "moderationRequired": true, + "dataPolicy": { + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "pricing": { + "prompt": "0.000015", + "completion": "0.00006", + "image": "0", + "request": "0", + "inputCacheRead": "0.0000075", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 246, + "newest": 207, + "throughputHighToLow": 318, + "latencyLowToHigh": 182, + "pricingLowToHigh": 308, + "pricingHighToLow": 11 + }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", + "providers": [ + { + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": "unknown", + "context": 128000, + "maxCompletionTokens": 32768, + "providerModelId": "o1-preview", + "pricing": { + "prompt": "0.000015", + "completion": "0.00006", + "image": "0", + "request": "0", + "inputCacheRead": "0.0000075", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "seed", + "max_tokens" + ], + "inputCost": 15, + "outputCost": 60, + "throughput": 109.215, + "latency": 6211 } ] }, @@ -66871,7 +70031,7 @@ export default { "retainsPrompts": true }, "pricing": { - "prompt": "0.0000015", + "prompt": "0.000001875", "completion": "0.00000225", "image": "0", "request": "0", @@ -66895,24 +70055,26 @@ export default { }, "providerRegion": null }, - "sorting": { - "topWeekly": 246, - "newest": 182, + "sorting": { + "topWeekly": 247, + "newest": 181, "throughputHighToLow": 288, - "latencyLowToHigh": 113, - "pricingLowToHigh": 238, - "pricingHighToLow": 80 + "latencyLowToHigh": 122, + "pricingLowToHigh": 239, + "pricingHighToLow": 79 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "Mancer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", "slug": "mancer", "quantization": "fp16", "context": 16384, "maxCompletionTokens": 2048, "providerModelId": "lumi-70b-v2", "pricing": { - "prompt": "0.0000015", + "prompt": "0.000001875", "completion": "0.00000225", "image": "0", "request": "0", @@ -66934,18 +70096,21 @@ export default { "seed", "top_a" ], - "inputCost": 1.5, - "outputCost": 2.25 + "inputCost": 1.88, + "outputCost": 2.25, + "throughput": 16.484, + "latency": 730 }, { "name": "Mancer (private)", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", "slug": "mancer (private)", "quantization": "fp16", "context": 16384, "maxCompletionTokens": 2048, "providerModelId": "lumi-70b-v2", "pricing": { - "prompt": "0.000002", + "prompt": "0.0000025", "completion": "0.000003", "image": "0", "request": "0", @@ -66967,11 +70132,14 @@ export default { "seed", "top_a" ], - "inputCost": 2, - "outputCost": 3 + "inputCost": 2.5, + "outputCost": 3, + "throughput": 16.399, + "latency": 1076 }, { "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", "slug": "featherless", "quantization": "fp8", "context": 16384, @@ -66999,7 +70167,9 @@ export default { "seed" ], "inputCost": 4, - "outputCost": 6 + "outputCost": 6, + "throughput": 13.7185, + "latency": 8308 } ] }, @@ -67160,16 +70330,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 247, - "newest": 249, - "throughputHighToLow": 25, - "latencyLowToHigh": 291, - "pricingLowToHigh": 82, - "pricingHighToLow": 236 + "topWeekly": 248, + "newest": 252, + "throughputHighToLow": 67, + "latencyLowToHigh": 293, + "pricingLowToHigh": 83, + "pricingHighToLow": 235 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { "name": "Enfer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://enfer.ai/&size=256", "slug": "enfer", "quantization": null, "context": 32768, @@ -67195,10 +70367,13 @@ export default { "logprobs" ], "inputCost": 0.03, - "outputCost": 0.05 + "outputCost": 0.05, + "throughput": 114.69, + "latency": 4570 }, { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "bf16", "context": 32768, @@ -67229,10 +70404,13 @@ export default { "min_p" ], "inputCost": 0.03, - "outputCost": 0.06 + "outputCost": 0.06, + "throughput": 109.947, + "latency": 584 }, { "name": "NextBit", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nextbit256.com/&size=256", "slug": "nextBit", "quantization": "bf16", "context": 32768, @@ -67258,10 +70436,13 @@ export default { "structured_outputs" ], "inputCost": 0.03, - "outputCost": 0.06 + "outputCost": 0.06, + "throughput": 62.5, + "latency": 1850 }, { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "unknown", "context": 32768, @@ -67290,10 +70471,13 @@ export default { "logit_bias" ], "inputCost": 0.03, - "outputCost": 0.06 + "outputCost": 0.06, + "throughput": 158.709, + "latency": 1081 }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": "unknown", "context": 32768, @@ -67322,533 +70506,9 @@ export default { "response_format" ], "inputCost": 0.2, - "outputCost": 0.2 - } - ] - }, - { - "slug": "qwen/qwen-2.5-vl-7b-instruct", - "hfSlug": "Qwen/Qwen2.5-VL-7B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-28T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Qwen: Qwen2.5-VL 7B Instruct (free)", - "shortName": "Qwen2.5-VL 7B Instruct (free)", - "author": "qwen", - "description": "Qwen2.5 VL 7B is a multimodal LLM from the Qwen Team with the following key enhancements:\n\n- SoTA understanding of images of various resolution & ratio: Qwen2.5-VL achieves state-of-the-art performance on visual understanding benchmarks, including MathVista, DocVQA, RealWorldQA, MTVQA, etc.\n\n- Understanding videos of 20min+: Qwen2.5-VL can understand videos over 20 minutes for high-quality video-based question answering, dialog, content creation, etc.\n\n- Agent that can operate your mobiles, robots, etc.: with the abilities of complex reasoning and decision making, Qwen2.5-VL can be integrated with devices like mobile phones, robots, etc., for automatic operation based on visual environment and text instructions.\n\n- Multilingual Support: to serve global users, besides English and Chinese, Qwen2.5-VL now supports the understanding of texts in different languages inside images, including most European languages, Japanese, Korean, Arabic, Vietnamese, etc.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2-vl/) and [GitHub repo](https://github.com/QwenLM/Qwen2-VL).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", - "modelVersionGroupId": null, - "contextLength": 64000, - "inputModalities": [ - "text", - "image" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Qwen", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "qwen/qwen-2-vl-7b-instruct", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "4f1593d5-297e-433e-af46-3ebc8c84446a", - "name": "Chutes | qwen/qwen-2-vl-7b-instruct:free", - "contextLength": 64000, - "model": { - "slug": "qwen/qwen-2.5-vl-7b-instruct", - "hfSlug": "Qwen/Qwen2.5-VL-7B-Instruct", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-28T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Qwen: Qwen2.5-VL 7B Instruct", - "shortName": "Qwen2.5-VL 7B Instruct", - "author": "qwen", - "description": "Qwen2.5 VL 7B is a multimodal LLM from the Qwen Team with the following key enhancements:\n\n- SoTA understanding of images of various resolution & ratio: Qwen2.5-VL achieves state-of-the-art performance on visual understanding benchmarks, including MathVista, DocVQA, RealWorldQA, MTVQA, etc.\n\n- Understanding videos of 20min+: Qwen2.5-VL can understand videos over 20 minutes for high-quality video-based question answering, dialog, content creation, etc.\n\n- Agent that can operate your mobiles, robots, etc.: with the abilities of complex reasoning and decision making, Qwen2.5-VL can be integrated with devices like mobile phones, robots, etc., for automatic operation based on visual environment and text instructions.\n\n- Multilingual Support: to serve global users, besides English and Chinese, Qwen2.5-VL now supports the understanding of texts in different languages inside images, including most European languages, Japanese, Korean, Arabic, Vietnamese, etc.\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen2-vl/) and [GitHub repo](https://github.com/QwenLM/Qwen2-VL).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", - "modelVersionGroupId": null, - "contextLength": 32768, - "inputModalities": [ - "text", - "image" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Qwen", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "qwen/qwen-2-vl-7b-instruct", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "qwen/qwen-2.5-vl-7b-instruct:free", - "modelVariantPermaslug": "qwen/qwen-2-vl-7b-instruct:free", - "providerName": "Chutes", - "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", - "paidModels": { - "training": true, - "retainsPrompts": true - } - }, - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "Chutes", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" - } - }, - "providerDisplayName": "Chutes", - "providerModelId": "Qwen/Qwen2.5-VL-7B-Instruct", - "providerGroup": "Chutes", - "quantization": null, - "variant": "free", - "isFree": true, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 64000, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", - "paidModels": { - "training": true, - "retainsPrompts": true - }, - "training": true, - "retainsPrompts": true - }, - "pricing": { - "prompt": "0", - "completion": "0", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 118, - "newest": 214, - "throughputHighToLow": 74, - "latencyLowToHigh": 126, - "pricingLowToHigh": 65, - "pricingHighToLow": 168 - }, - "providers": [ - { - "name": "Hyperbolic", - "slug": "hyperbolic", - "quantization": "bf16", - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "Qwen/Qwen2.5-VL-7B-Instruct", - "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", - "image": "0.0001445", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "logprobs", - "top_logprobs", - "seed", - "logit_bias", - "top_k", - "min_p", - "repetition_penalty" - ], - "inputCost": 0.2, - "outputCost": 0.2 - }, - { - "name": "inference.net", - "slug": "inferenceNet", - "quantization": "bf16", - "context": 128000, - "maxCompletionTokens": 8192, - "providerModelId": "qwen/qwen2.5-7b-instruct/bf-16", - "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "logit_bias", - "top_logprobs" - ], - "inputCost": 0.2, - "outputCost": 0.2 - }, - { - "name": "kluster.ai", - "slug": "klusterAi", - "quantization": null, - "context": 32768, - "maxCompletionTokens": 32768, - "providerModelId": "Qwen/Qwen2.5-VL-7B-Instruct", - "pricing": { - "prompt": "0.0000003", - "completion": "0.0000003", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature" - ], - "inputCost": 0.3, - "outputCost": 0.3 - } - ] - }, - { - "slug": "meta-llama/llama-guard-4-12b", - "hfSlug": "meta-llama/Llama-Guard-4-12B", - "updatedAt": "2025-04-30T14:10:44.640461+00:00", - "createdAt": "2025-04-30T01:06:33.531556+00:00", - "hfUpdatedAt": null, - "name": "Meta: Llama Guard 4 12B", - "shortName": "Llama Guard 4 12B", - "author": "meta-llama", - "description": "Llama Guard 4 is a Llama 4 Scout-derived multimodal pretrained model, fine-tuned for content safety classification. Similar to previous versions, it can be used to classify content in both LLM inputs (prompt classification) and in LLM responses (response classification). It acts as an LLM—generating text in its output that indicates whether a given prompt or response is safe or unsafe, and if unsafe, it also lists the content categories violated.\n\nLlama Guard 4 was aligned to safeguard against the standardized MLCommons hazards taxonomy and designed to support multimodal Llama 4 capabilities. Specifically, it combines features from previous Llama Guard models, providing content moderation for English and multiple supported languages, along with enhanced capabilities to handle mixed text-and-image prompts, including multiple images. Additionally, Llama Guard 4 is integrated into the Llama Moderations API, extending robust safety classification to text and images.", - "modelVersionGroupId": null, - "contextLength": 163840, - "inputModalities": [ - "image", - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "meta-llama/llama-guard-4-12b", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "850b84c3-42a7-4cec-99c0-b5582d0da66b", - "name": "DeepInfra | meta-llama/llama-guard-4-12b", - "contextLength": 163840, - "model": { - "slug": "meta-llama/llama-guard-4-12b", - "hfSlug": "meta-llama/Llama-Guard-4-12B", - "updatedAt": "2025-04-30T14:10:44.640461+00:00", - "createdAt": "2025-04-30T01:06:33.531556+00:00", - "hfUpdatedAt": null, - "name": "Meta: Llama Guard 4 12B", - "shortName": "Llama Guard 4 12B", - "author": "meta-llama", - "description": "Llama Guard 4 is a Llama 4 Scout-derived multimodal pretrained model, fine-tuned for content safety classification. Similar to previous versions, it can be used to classify content in both LLM inputs (prompt classification) and in LLM responses (response classification). It acts as an LLM—generating text in its output that indicates whether a given prompt or response is safe or unsafe, and if unsafe, it also lists the content categories violated.\n\nLlama Guard 4 was aligned to safeguard against the standardized MLCommons hazards taxonomy and designed to support multimodal Llama 4 capabilities. Specifically, it combines features from previous Llama Guard models, providing content moderation for English and multiple supported languages, along with enhanced capabilities to handle mixed text-and-image prompts, including multiple images. Additionally, Llama Guard 4 is integrated into the Llama Moderations API, extending robust safety classification to text and images.", - "modelVersionGroupId": null, - "contextLength": 163840, - "inputModalities": [ - "image", - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "meta-llama/llama-guard-4-12b", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "meta-llama/llama-guard-4-12b", - "modelVariantPermaslug": "meta-llama/llama-guard-4-12b", - "providerName": "DeepInfra", - "providerInfo": { - "name": "DeepInfra", - "displayName": "DeepInfra", - "slug": "deepinfra", - "baseUrl": "url", - "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", - "paidModels": { - "training": false, - "retainsPrompts": false - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "DeepInfra", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/DeepInfra.webp" - } - }, - "providerDisplayName": "DeepInfra", - "providerModelId": "meta-llama/Llama-Guard-4-12B", - "providerGroup": "DeepInfra", - "quantization": "bf16", - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "privacyPolicyUrl": "https://deepinfra.com/privacy", - "termsOfServiceUrl": "https://deepinfra.com/terms", - "dataPolicyUrl": "https://deepinfra.com/docs/data", - "paidModels": { - "training": false, - "retainsPrompts": false - }, - "training": false, - "retainsPrompts": false - }, - "pricing": { - "prompt": "0.00000005", - "completion": "0.00000005", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 249, - "newest": 21, - "throughputHighToLow": 79, - "latencyLowToHigh": 142, - "pricingLowToHigh": 95, - "pricingHighToLow": 223 - }, - "providers": [ - { - "name": "DeepInfra", - "slug": "deepInfra", - "quantization": "bf16", - "context": 163840, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-Guard-4-12B", - "pricing": { - "prompt": "0.00000005", - "completion": "0.00000005", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "response_format", - "top_k", - "seed", - "min_p" - ], - "inputCost": 0.05, - "outputCost": 0.05 - }, - { - "name": "Together", - "slug": "together", - "quantization": null, - "context": 1048576, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/Llama-Guard-4-12B", - "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" - ], - "inputCost": 0.2, - "outputCost": 0.2 - }, - { - "name": "Groq", - "slug": "groq", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 1024, - "providerModelId": "meta-llama/llama-guard-4-12b", - "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "top_logprobs", - "logprobs", - "logit_bias", - "seed" - ], - "inputCost": 0.2, - "outputCost": 0.2 + "outputCost": 0.2, + "throughput": 215.931, + "latency": 1034 } ] }, @@ -68016,16 +70676,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 250, + "topWeekly": 249, "newest": 69, "throughputHighToLow": 224, - "latencyLowToHigh": 119, - "pricingLowToHigh": 212, - "pricingHighToLow": 106 + "latencyLowToHigh": 146, + "pricingLowToHigh": 211, + "pricingHighToLow": 107 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": null, "context": 8192, @@ -68054,7 +70716,9 @@ export default { "response_format" ], "inputCost": 0.88, - "outputCost": 0.88 + "outputCost": 0.88, + "throughput": 36.294, + "latency": 836 } ] }, @@ -68219,16 +70883,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 251, + "topWeekly": 250, "newest": 274, - "throughputHighToLow": 287, - "latencyLowToHigh": 260, + "throughputHighToLow": 289, + "latencyLowToHigh": 262, "pricingLowToHigh": 202, "pricingHighToLow": 118 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "slug": "novitaAi", "quantization": "unknown", "context": 4096, @@ -68257,7 +70923,9 @@ export default { "logit_bias" ], "inputCost": 0.8, - "outputCost": 0.8 + "outputCost": 0.8, + "throughput": 16.3675, + "latency": 2213 } ] }, @@ -68421,16 +71089,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 252, + "topWeekly": 251, "newest": 233, - "throughputHighToLow": 67, - "latencyLowToHigh": 30, + "throughputHighToLow": 66, + "latencyLowToHigh": 29, "pricingLowToHigh": 160, "pricingHighToLow": 157 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", "slug": "mistral", "quantization": "unknown", "context": 262144, @@ -68458,22 +71128,266 @@ export default { "seed" ], "inputCost": 0.25, - "outputCost": 0.25 + "outputCost": 0.25, + "throughput": 109.535, + "latency": 311 } ] }, { - "slug": "arcee-ai/virtuoso-large", + "slug": "mistralai/mistral-saba", "hfSlug": "", - "updatedAt": "2025-05-05T21:48:18.128183+00:00", - "createdAt": "2025-05-05T21:01:25.294732+00:00", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-17T14:40:39.116446+00:00", + "hfUpdatedAt": null, + "name": "Mistral: Saba", + "shortName": "Saba", + "author": "mistralai", + "description": "Mistral Saba is a 24B-parameter language model specifically designed for the Middle East and South Asia, delivering accurate and contextually relevant responses while maintaining efficient performance. Trained on curated regional datasets, it supports multiple Indian-origin languages—including Tamil and Malayalam—alongside Arabic. This makes it a versatile option for a range of regional and multilingual applications. Read more at the blog post [here](https://mistral.ai/en/news/mistral-saba)", + "modelVersionGroupId": null, + "contextLength": 32768, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Mistral", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "mistralai/mistral-saba-2502", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "c5a8bebe-8564-47ed-9d05-4151aa3c6e3a", + "name": "Mistral | mistralai/mistral-saba-2502", + "contextLength": 32768, + "model": { + "slug": "mistralai/mistral-saba", + "hfSlug": "", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-17T14:40:39.116446+00:00", + "hfUpdatedAt": null, + "name": "Mistral: Saba", + "shortName": "Saba", + "author": "mistralai", + "description": "Mistral Saba is a 24B-parameter language model specifically designed for the Middle East and South Asia, delivering accurate and contextually relevant responses while maintaining efficient performance. Trained on curated regional datasets, it supports multiple Indian-origin languages—including Tamil and Malayalam—alongside Arabic. This makes it a versatile option for a range of regional and multilingual applications. Read more at the blog post [here](https://mistral.ai/en/news/mistral-saba)", + "modelVersionGroupId": null, + "contextLength": 32000, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Mistral", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "mistralai/mistral-saba-2502", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "mistralai/mistral-saba", + "modelVariantPermaslug": "mistralai/mistral-saba-2502", + "providerName": "Mistral", + "providerInfo": { + "name": "Mistral", + "displayName": "Mistral", + "slug": "mistral", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + } + }, + "headquarters": "FR", + "hasChatCompletions": true, + "hasCompletions": false, + "isAbortable": false, + "moderationRequired": false, + "group": "Mistral", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/Mistral.png" + } + }, + "providerDisplayName": "Mistral", + "providerModelId": "mistral-saba-latest", + "providerGroup": "Mistral", + "quantization": null, + "variant": "standard", + "isFree": false, + "canAbort": false, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "structured_outputs", + "seed" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", + "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "pricing": { + "prompt": "0.0000002", + "completion": "0.0000006", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": true, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": false, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 252, + "newest": 112, + "throughputHighToLow": 94, + "latencyLowToHigh": 17, + "pricingLowToHigh": 154, + "pricingHighToLow": 163 + }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", + "providers": [ + { + "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", + "slug": "mistral", + "quantization": null, + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "mistral-saba-latest", + "pricing": { + "prompt": "0.0000002", + "completion": "0.0000006", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "structured_outputs", + "seed" + ], + "inputCost": 0.2, + "outputCost": 0.6, + "throughput": 81.694, + "latency": 248 + }, + { + "name": "Groq", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://groq.com/&size=256", + "slug": "groq", + "quantization": null, + "context": 32768, + "maxCompletionTokens": 32768, + "providerModelId": "mistral-saba-24b", + "pricing": { + "prompt": "0.00000079", + "completion": "0.00000079", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "response_format", + "top_logprobs", + "logprobs", + "logit_bias", + "seed" + ], + "inputCost": 0.79, + "outputCost": 0.79, + "throughput": 423.2015, + "latency": 353 + } + ] + }, + { + "slug": "openai/gpt-3.5-turbo-0613", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-01-25T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Arcee AI: Virtuoso Large", - "shortName": "Virtuoso Large", - "author": "arcee-ai", - "description": "Virtuoso‑Large is Arcee's top‑tier general‑purpose LLM at 72 B parameters, tuned to tackle cross‑domain reasoning, creative writing and enterprise QA. Unlike many 70 B peers, it retains the 128 k context inherited from Qwen 2.5, letting it ingest books, codebases or financial filings wholesale. Training blended DeepSeek R1 distillation, multi‑epoch supervised fine‑tuning and a final DPO/RLHF alignment stage, yielding strong performance on BIG‑Bench‑Hard, GSM‑8K and long‑context Needle‑In‑Haystack tests. Enterprises use Virtuoso‑Large as the \"fallback\" brain in Conductor pipelines when other SLMs flag low confidence. Despite its size, aggressive KV‑cache optimizations keep first‑token latency in the low‑second range on 8× H100 nodes, making it a practical production‑grade powerhouse.", + "name": "OpenAI: GPT-3.5 Turbo (older v0613)", + "shortName": "GPT-3.5 Turbo (older v0613)", + "author": "openai", + "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 4095, "inputModalities": [ "text" ], @@ -68481,32 +71395,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "arcee-ai/virtuoso-large", + "permaslug": "openai/gpt-3.5-turbo-0613", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "6ea266fd-81e1-4ad2-ba62-2f09c5762435", - "name": "Together | arcee-ai/virtuoso-large", - "contextLength": 131072, + "id": "f29e5bba-56f3-408a-a036-185e3dc20be3", + "name": "Azure | openai/gpt-3.5-turbo-0613", + "contextLength": 4095, "model": { - "slug": "arcee-ai/virtuoso-large", - "hfSlug": "", - "updatedAt": "2025-05-05T21:48:18.128183+00:00", - "createdAt": "2025-05-05T21:01:25.294732+00:00", + "slug": "openai/gpt-3.5-turbo-0613", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-01-25T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Arcee AI: Virtuoso Large", - "shortName": "Virtuoso Large", - "author": "arcee-ai", - "description": "Virtuoso‑Large is Arcee's top‑tier general‑purpose LLM at 72 B parameters, tuned to tackle cross‑domain reasoning, creative writing and enterprise QA. Unlike many 70 B peers, it retains the 128 k context inherited from Qwen 2.5, letting it ingest books, codebases or financial filings wholesale. Training blended DeepSeek R1 distillation, multi‑epoch supervised fine‑tuning and a final DPO/RLHF alignment stage, yielding strong performance on BIG‑Bench‑Hard, GSM‑8K and long‑context Needle‑In‑Haystack tests. Enterprises use Virtuoso‑Large as the \"fallback\" brain in Conductor pipelines when other SLMs flag low confidence. Despite its size, aggressive KV‑cache optimizations keep first‑token latency in the low‑second range on 8× H100 nodes, making it a practical production‑grade powerhouse.", + "name": "OpenAI: GPT-3.5 Turbo (older v0613)", + "shortName": "GPT-3.5 Turbo (older v0613)", + "author": "openai", + "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 4095, "inputModalities": [ "text" ], @@ -68514,28 +71428,29 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "GPT", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "arcee-ai/virtuoso-large", + "permaslug": "openai/gpt-3.5-turbo-0613", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "arcee-ai/virtuoso-large", - "modelVariantPermaslug": "arcee-ai/virtuoso-large", - "providerName": "Together", + "modelVariantSlug": "openai/gpt-3.5-turbo-0613", + "modelVariantPermaslug": "openai/gpt-3.5-turbo-0613", + "providerName": "Azure", "providerInfo": { - "name": "Together", - "displayName": "Together", - "slug": "together", + "name": "Azure", + "displayName": "Azure", + "slug": "azure", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://www.microsoft.com/en-us/legal/terms-of-use?oneroute=true", + "privacyPolicyUrl": "https://www.microsoft.com/en-us/privacy/privacystatement", + "dataPolicyUrl": "https://learn.microsoft.com/en-us/legal/cognitive-services/openai/data-privacy?tabs=azure-portal", "paidModels": { "training": false, "retainsPrompts": false @@ -68543,49 +71458,53 @@ export default { }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, + "hasCompletions": false, "isAbortable": true, "moderationRequired": false, - "group": "Together", + "group": "Azure", "editors": [], "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, + "isMultipartSupported": false, + "statusPageUrl": "https://status.azure.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" + "url": "/images/icons/Azure.svg" } }, - "providerDisplayName": "Together", - "providerModelId": "arcee-ai/virtuoso-large", - "providerGroup": "Together", - "quantization": null, + "providerDisplayName": "Azure", + "providerModelId": "gpt-35-turbo", + "providerGroup": "Azure", + "quantization": "unknown", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 64000, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "top_k", - "repetition_penalty", + "seed", "logit_bias", - "min_p", - "response_format" + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://www.microsoft.com/en-us/legal/terms-of-use?oneroute=true", + "privacyPolicyUrl": "https://www.microsoft.com/en-us/privacy/privacystatement", + "dataPolicyUrl": "https://learn.microsoft.com/en-us/legal/cognitive-services/openai/data-privacy?tabs=azure-portal", "paidModels": { "training": false, "retainsPrompts": false @@ -68594,8 +71513,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000075", - "completion": "0.0000012", + "prompt": "0.000001", + "completion": "0.000002", "image": "0", "request": "0", "webSearch": "0", @@ -68606,38 +71525,39 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, - "supportsMultipart": true, + "supportsMultipart": false, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 253, - "newest": 6, - "throughputHighToLow": 106, - "latencyLowToHigh": 67, - "pricingLowToHigh": 199, - "pricingHighToLow": 119 + "newest": 285, + "throughputHighToLow": 22, + "latencyLowToHigh": 156, + "pricingLowToHigh": 224, + "pricingHighToLow": 98 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "Together", - "slug": "together", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 64000, - "providerModelId": "arcee-ai/virtuoso-large", + "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", + "slug": "azure", + "quantization": "unknown", + "context": 4095, + "maxCompletionTokens": 4096, + "providerModelId": "gpt-35-turbo", "pricing": { - "prompt": "0.00000075", - "completion": "0.0000012", + "prompt": "0.000001", + "completion": "0.000002", "image": "0", "request": "0", "webSearch": "0", @@ -68645,113 +71565,113 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "top_k", - "repetition_penalty", + "seed", "logit_bias", - "min_p", - "response_format" + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" ], - "inputCost": 0.75, - "outputCost": 1.2 + "inputCost": 1, + "outputCost": 2, + "throughput": 167.32, + "latency": 918 } ] }, { - "slug": "raifle/sorcererlm-8x22b", - "hfSlug": "rAIfle/SorcererLM-8x22b-bf16", + "slug": "bytedance-research/ui-tars-72b", + "hfSlug": "bytedance-research/UI-TARS-72B-DPO", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-08T22:31:23.953049+00:00", + "createdAt": "2025-03-26T20:14:25.673407+00:00", "hfUpdatedAt": null, - "name": "SorcererLM 8x22B", - "shortName": "SorcererLM 8x22B", - "author": "raifle", - "description": "SorcererLM is an advanced RP and storytelling model, built as a Low-rank 16-bit LoRA fine-tuned on [WizardLM-2 8x22B](/microsoft/wizardlm-2-8x22b).\n\n- Advanced reasoning and emotional intelligence for engaging and immersive interactions\n- Vivid writing capabilities enriched with spatial and contextual awareness\n- Enhanced narrative depth, promoting creative and dynamic storytelling", + "name": "Bytedance: UI-TARS 72B (free)", + "shortName": "UI-TARS 72B (free)", + "author": "bytedance-research", + "description": "UI-TARS 72B is an open-source multimodal AI model designed specifically for automating browser and desktop tasks through visual interaction and control. The model is built with a specialized vision architecture enabling accurate interpretation and manipulation of on-screen visual data. It supports automation tasks within web browsers as well as desktop applications, including Microsoft Office and VS Code.\n\nCore capabilities include intelligent screen detection, predictive action modeling, and efficient handling of repetitive interactions. UI-TARS employs supervised fine-tuning (SFT) tailored explicitly for computer control scenarios. It can be deployed locally or accessed via Hugging Face for demonstration purposes. Intended use cases encompass workflow automation, task scripting, and interactive desktop control applications.", "modelVersionGroupId": null, - "contextLength": 16000, + "contextLength": 32768, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "vicuna", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "USER:", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "raifle/sorcererlm-8x22b", + "permaslug": "bytedance-research/ui-tars-72b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "65206957-c772-4743-93ff-45900a190ddd", - "name": "Infermatic | raifle/sorcererlm-8x22b", - "contextLength": 16000, + "id": "e480b8f8-44f2-42cb-bcec-5488d3341b07", + "name": "Chutes | bytedance-research/ui-tars-72b:free", + "contextLength": 32768, "model": { - "slug": "raifle/sorcererlm-8x22b", - "hfSlug": "rAIfle/SorcererLM-8x22b-bf16", + "slug": "bytedance-research/ui-tars-72b", + "hfSlug": "bytedance-research/UI-TARS-72B-DPO", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-08T22:31:23.953049+00:00", + "createdAt": "2025-03-26T20:14:25.673407+00:00", "hfUpdatedAt": null, - "name": "SorcererLM 8x22B", - "shortName": "SorcererLM 8x22B", - "author": "raifle", - "description": "SorcererLM is an advanced RP and storytelling model, built as a Low-rank 16-bit LoRA fine-tuned on [WizardLM-2 8x22B](/microsoft/wizardlm-2-8x22b).\n\n- Advanced reasoning and emotional intelligence for engaging and immersive interactions\n- Vivid writing capabilities enriched with spatial and contextual awareness\n- Enhanced narrative depth, promoting creative and dynamic storytelling", + "name": "Bytedance: UI-TARS 72B ", + "shortName": "UI-TARS 72B ", + "author": "bytedance-research", + "description": "UI-TARS 72B is an open-source multimodal AI model designed specifically for automating browser and desktop tasks through visual interaction and control. The model is built with a specialized vision architecture enabling accurate interpretation and manipulation of on-screen visual data. It supports automation tasks within web browsers as well as desktop applications, including Microsoft Office and VS Code.\n\nCore capabilities include intelligent screen detection, predictive action modeling, and efficient handling of repetitive interactions. UI-TARS employs supervised fine-tuning (SFT) tailored explicitly for computer control scenarios. It can be deployed locally or accessed via Hugging Face for demonstration purposes. Intended use cases encompass workflow automation, task scripting, and interactive desktop control applications.", "modelVersionGroupId": null, - "contextLength": 16000, + "contextLength": 32768, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "vicuna", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "USER:", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "raifle/sorcererlm-8x22b", + "permaslug": "bytedance-research/ui-tars-72b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "raifle/sorcererlm-8x22b", - "modelVariantPermaslug": "raifle/sorcererlm-8x22b", - "providerName": "Infermatic", + "modelVariantSlug": "bytedance-research/ui-tars-72b:free", + "modelVariantPermaslug": "bytedance-research/ui-tars-72b:free", + "providerName": "Chutes", "providerInfo": { - "name": "Infermatic", - "displayName": "Infermatic", - "slug": "infermatic", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://infermatic.ai/privacy-policy/", - "termsOfServiceUrl": "https://infermatic.ai/terms-and-conditions/", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true } }, "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Infermatic", + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, @@ -68759,15 +71679,15 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://infermatic.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "Infermatic", - "providerModelId": "rAIfle-SorcererLM-8x22b-bf16", - "providerGroup": "Infermatic", + "providerDisplayName": "Chutes", + "providerModelId": "bytedance-research/UI-TARS-72B-DPO", + "providerGroup": "Chutes", "quantization": null, - "variant": "standard", - "isFree": false, + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, @@ -68780,27 +71700,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "logit_bias", + "seed", "top_k", "min_p", - "seed" + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://infermatic.ai/privacy-policy/", - "termsOfServiceUrl": "https://infermatic.ai/terms-and-conditions/", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true }, - "training": false, - "retainsPrompts": false + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.0000045", - "completion": "0.0000045", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -68819,65 +71740,34 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 254, - "newest": 174, - "throughputHighToLow": 169, - "latencyLowToHigh": 104, - "pricingLowToHigh": 285, - "pricingHighToLow": 33 + "newest": 70, + "throughputHighToLow": 285, + "latencyLowToHigh": 296, + "pricingLowToHigh": 29, + "pricingHighToLow": 277 }, - "providers": [ - { - "name": "Infermatic", - "slug": "infermatic", - "quantization": null, - "context": 16000, - "maxCompletionTokens": null, - "providerModelId": "rAIfle-SorcererLM-8x22b-bf16", - "pricing": { - "prompt": "0.0000045", - "completion": "0.0000045", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "logit_bias", - "top_k", - "min_p", - "seed" - ], - "inputCost": 4.5, - "outputCost": 4.5 - } - ] + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [] }, { - "slug": "mistralai/mistral-saba", - "hfSlug": "", + "slug": "ai21/jamba-1.6-large", + "hfSlug": "ai21labs/AI21-Jamba-Large-1.6", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-17T14:40:39.116446+00:00", + "createdAt": "2025-03-13T22:32:53+00:00", "hfUpdatedAt": null, - "name": "Mistral: Saba", - "shortName": "Saba", - "author": "mistralai", - "description": "Mistral Saba is a 24B-parameter language model specifically designed for the Middle East and South Asia, delivering accurate and contextually relevant responses while maintaining efficient performance. Trained on curated regional datasets, it supports multiple Indian-origin languages—including Tamil and Malayalam—alongside Arabic. This makes it a versatile option for a range of regional and multilingual applications. Read more at the blog post [here](https://mistral.ai/en/news/mistral-saba)", - "modelVersionGroupId": null, - "contextLength": 32768, + "name": "AI21: Jamba 1.6 Large", + "shortName": "Jamba 1.6 Large", + "author": "ai21", + "description": "AI21 Jamba Large 1.6 is a high-performance hybrid foundation model combining State Space Models (Mamba) with Transformer attention mechanisms. Developed by AI21, it excels in extremely long-context handling (256K tokens), demonstrates superior inference efficiency (up to 2.5x faster than comparable models), and supports structured JSON output and tool-use capabilities. It has 94 billion active parameters (398 billion total), optimized quantization support (ExpertsInt8), and multilingual proficiency in languages such as English, Spanish, French, Portuguese, Italian, Dutch, German, Arabic, and Hebrew.\n\nUsage of this model is subject to the [Jamba Open Model License](https://www.ai21.com/licenses/jamba-open-model-license).", + "modelVersionGroupId": "cd1eb031-30bc-4e2e-aa06-3c20f986e5c7", + "contextLength": 256000, "inputModalities": [ "text" ], @@ -68885,32 +71775,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-saba-2502", + "permaslug": "ai21/jamba-1.6-large", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "c5a8bebe-8564-47ed-9d05-4151aa3c6e3a", - "name": "Mistral | mistralai/mistral-saba-2502", - "contextLength": 32768, + "id": "f4dfbac5-0169-4bc8-9861-27dfef791169", + "name": "AI21 | ai21/jamba-1.6-large", + "contextLength": 256000, "model": { - "slug": "mistralai/mistral-saba", - "hfSlug": "", + "slug": "ai21/jamba-1.6-large", + "hfSlug": "ai21labs/AI21-Jamba-Large-1.6", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-17T14:40:39.116446+00:00", + "createdAt": "2025-03-13T22:32:53+00:00", "hfUpdatedAt": null, - "name": "Mistral: Saba", - "shortName": "Saba", - "author": "mistralai", - "description": "Mistral Saba is a 24B-parameter language model specifically designed for the Middle East and South Asia, delivering accurate and contextually relevant responses while maintaining efficient performance. Trained on curated regional datasets, it supports multiple Indian-origin languages—including Tamil and Malayalam—alongside Arabic. This makes it a versatile option for a range of regional and multilingual applications. Read more at the blog post [here](https://mistral.ai/en/news/mistral-saba)", - "modelVersionGroupId": null, - "contextLength": 32000, + "name": "AI21: Jamba 1.6 Large", + "shortName": "Jamba 1.6 Large", + "author": "ai21", + "description": "AI21 Jamba Large 1.6 is a high-performance hybrid foundation model combining State Space Models (Mamba) with Transformer attention mechanisms. Developed by AI21, it excels in extremely long-context handling (256K tokens), demonstrates superior inference efficiency (up to 2.5x faster than comparable models), and supports structured JSON output and tool-use capabilities. It has 94 billion active parameters (398 billion total), optimized quantization support (ExpertsInt8), and multilingual proficiency in languages such as English, Spanish, French, Portuguese, Italian, Dutch, German, Arabic, and Hebrew.\n\nUsage of this model is subject to the [Jamba Open Model License](https://www.ai21.com/licenses/jamba-open-model-license).", + "modelVersionGroupId": "cd1eb031-30bc-4e2e-aa06-3c20f986e5c7", + "contextLength": 256000, "inputModalities": [ "text" ], @@ -68918,59 +71808,57 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-saba-2502", + "permaslug": "ai21/jamba-1.6-large", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "mistralai/mistral-saba", - "modelVariantPermaslug": "mistralai/mistral-saba-2502", - "providerName": "Mistral", + "modelVariantSlug": "ai21/jamba-1.6-large", + "modelVariantPermaslug": "ai21/jamba-1.6-large", + "providerName": "AI21", "providerInfo": { - "name": "Mistral", - "displayName": "Mistral", - "slug": "mistral", + "name": "AI21", + "displayName": "AI21", + "slug": "ai21", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "privacyPolicyUrl": "https://www.ai21.com/privacy-policy/", + "termsOfServiceUrl": "https://www.ai21.com/terms-of-service/", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false } }, - "headquarters": "FR", + "headquarters": "IL", "hasChatCompletions": true, "hasCompletions": false, "isAbortable": false, "moderationRequired": false, - "group": "Mistral", + "group": "AI21", "editors": [], "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, + "isMultipartSupported": false, + "statusPageUrl": "https://status.ai21.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Mistral.png" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://ai21.com/&size=256" } }, - "providerDisplayName": "Mistral", - "providerModelId": "mistral-saba-latest", - "providerGroup": "Mistral", - "quantization": null, + "providerDisplayName": "AI21", + "providerModelId": "jamba-large-1.6", + "providerGroup": "AI21", + "quantization": "bf16", "variant": "standard", "isFree": false, "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -68979,30 +71867,21 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "structured_outputs", - "seed" + "stop" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://mistral.ai/terms/#terms-of-use", - "privacyPolicyUrl": "https://mistral.ai/terms/#privacy-policy", + "privacyPolicyUrl": "https://www.ai21.com/privacy-policy/", + "termsOfServiceUrl": "https://www.ai21.com/terms-of-service/", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, "pricing": { - "prompt": "0.0000002", - "completion": "0.0000006", + "prompt": "0.000002", + "completion": "0.000008", "image": "0", "request": "0", "webSearch": "0", @@ -69015,36 +71894,37 @@ export default { "isDisabled": false, "supportsToolParameters": true, "supportsReasoning": false, - "supportsMultipart": true, + "supportsMultipart": false, "limitRpm": null, "limitRpd": null, "hasCompletions": false, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 255, - "newest": 112, - "throughputHighToLow": 98, - "latencyLowToHigh": 33, - "pricingLowToHigh": 154, - "pricingHighToLow": 163 + "newest": 85, + "throughputHighToLow": 146, + "latencyLowToHigh": 159, + "pricingLowToHigh": 248, + "pricingHighToLow": 68 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai21.com/\\u0026size=256", "providers": [ { - "name": "Mistral", - "slug": "mistral", - "quantization": null, - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "mistral-saba-latest", + "name": "AI21", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://ai21.com/&size=256", + "slug": "ai21", + "quantization": "bf16", + "context": 256000, + "maxCompletionTokens": 4096, + "providerModelId": "jamba-large-1.6", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000006", + "prompt": "0.000002", + "completion": "0.000008", "image": "0", "request": "0", "webSearch": "0", @@ -69057,151 +71937,122 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "structured_outputs", - "seed" - ], - "inputCost": 0.2, - "outputCost": 0.6 - }, - { - "name": "Groq", - "slug": "groq", - "quantization": null, - "context": 32768, - "maxCompletionTokens": 32768, - "providerModelId": "mistral-saba-24b", - "pricing": { - "prompt": "0.00000079", - "completion": "0.00000079", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "response_format", - "top_logprobs", - "logprobs", - "logit_bias", - "seed" + "stop" ], - "inputCost": 0.79, - "outputCost": 0.79 + "inputCost": 2, + "outputCost": 8, + "throughput": 56.652, + "latency": 969 } ] }, { - "slug": "bytedance-research/ui-tars-72b", - "hfSlug": "bytedance-research/UI-TARS-72B-DPO", + "slug": "jondurbin/airoboros-l2-70b", + "hfSlug": "jondurbin/airoboros-l2-70b-2.2.1", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-26T20:14:25.673407+00:00", + "createdAt": "2023-10-29T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Bytedance: UI-TARS 72B (free)", - "shortName": "UI-TARS 72B (free)", - "author": "bytedance-research", - "description": "UI-TARS 72B is an open-source multimodal AI model designed specifically for automating browser and desktop tasks through visual interaction and control. The model is built with a specialized vision architecture enabling accurate interpretation and manipulation of on-screen visual data. It supports automation tasks within web browsers as well as desktop applications, including Microsoft Office and VS Code.\n\nCore capabilities include intelligent screen detection, predictive action modeling, and efficient handling of repetitive interactions. UI-TARS employs supervised fine-tuning (SFT) tailored explicitly for computer control scenarios. It can be deployed locally or accessed via Hugging Face for demonstration purposes. Intended use cases encompass workflow automation, task scripting, and interactive desktop control applications.", + "name": "Airoboros 70B", + "shortName": "Airoboros 70B", + "author": "jondurbin", + "description": "A Llama 2 70B fine-tune using synthetic data (the Airoboros dataset).\n\nCurrently based on [jondurbin/airoboros-l2-70b](https://huggingface.co/jondurbin/airoboros-l2-70b-2.2.1), but might get updated in the future.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 4096, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Llama2", + "instructType": "airoboros", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "USER:", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "bytedance-research/ui-tars-72b", + "permaslug": "jondurbin/airoboros-l2-70b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "e480b8f8-44f2-42cb-bcec-5488d3341b07", - "name": "Chutes | bytedance-research/ui-tars-72b:free", - "contextLength": 32768, + "id": "08737538-335d-4639-b6e9-cc34a8134e30", + "name": "Novita | jondurbin/airoboros-l2-70b", + "contextLength": 4096, "model": { - "slug": "bytedance-research/ui-tars-72b", - "hfSlug": "bytedance-research/UI-TARS-72B-DPO", + "slug": "jondurbin/airoboros-l2-70b", + "hfSlug": "jondurbin/airoboros-l2-70b-2.2.1", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-26T20:14:25.673407+00:00", + "createdAt": "2023-10-29T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Bytedance: UI-TARS 72B ", - "shortName": "UI-TARS 72B ", - "author": "bytedance-research", - "description": "UI-TARS 72B is an open-source multimodal AI model designed specifically for automating browser and desktop tasks through visual interaction and control. The model is built with a specialized vision architecture enabling accurate interpretation and manipulation of on-screen visual data. It supports automation tasks within web browsers as well as desktop applications, including Microsoft Office and VS Code.\n\nCore capabilities include intelligent screen detection, predictive action modeling, and efficient handling of repetitive interactions. UI-TARS employs supervised fine-tuning (SFT) tailored explicitly for computer control scenarios. It can be deployed locally or accessed via Hugging Face for demonstration purposes. Intended use cases encompass workflow automation, task scripting, and interactive desktop control applications.", + "name": "Airoboros 70B", + "shortName": "Airoboros 70B", + "author": "jondurbin", + "description": "A Llama 2 70B fine-tune using synthetic data (the Airoboros dataset).\n\nCurrently based on [jondurbin/airoboros-l2-70b](https://huggingface.co/jondurbin/airoboros-l2-70b-2.2.1), but might get updated in the future.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 4096, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Llama2", + "instructType": "airoboros", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "USER:", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "bytedance-research/ui-tars-72b", + "permaslug": "jondurbin/airoboros-l2-70b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "bytedance-research/ui-tars-72b:free", - "modelVariantPermaslug": "bytedance-research/ui-tars-72b:free", - "providerName": "Chutes", + "modelVariantSlug": "jondurbin/airoboros-l2-70b", + "modelVariantPermaslug": "jondurbin/airoboros-l2-70b", + "providerName": "Novita", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "Novita", + "displayName": "NovitaAI", + "slug": "novita", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false } }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Chutes", + "group": "Novita", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.novita.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "invertRequired": true } }, - "providerDisplayName": "Chutes", - "providerModelId": "bytedance-research/UI-TARS-72B-DPO", - "providerGroup": "Chutes", - "quantization": null, - "variant": "free", - "isFree": true, + "providerDisplayName": "NovitaAI", + "providerModelId": "jondurbin/airoboros-l2-70b", + "providerGroup": "Novita", + "quantization": "unknown", + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, @@ -69218,24 +72069,21 @@ export default { "top_k", "min_p", "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" + "logit_bias" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false }, - "training": true, - "retainsPrompts": true + "training": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.0000005", + "completion": "0.0000005", "image": "0", "request": "0", "webSearch": "0", @@ -69254,33 +72102,69 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 256, - "newest": 70, - "throughputHighToLow": 285, - "latencyLowToHigh": 302, - "pricingLowToHigh": 29, - "pricingHighToLow": 277 + "newest": 302, + "throughputHighToLow": 178, + "latencyLowToHigh": 279, + "pricingLowToHigh": 179, + "pricingHighToLow": 140 }, - "providers": [] + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [ + { + "name": "NovitaAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", + "slug": "novitaAi", + "quantization": "unknown", + "context": 4096, + "maxCompletionTokens": null, + "providerModelId": "jondurbin/airoboros-l2-70b", + "pricing": { + "prompt": "0.0000005", + "completion": "0.0000005", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logit_bias" + ], + "inputCost": 0.5, + "outputCost": 0.5, + "throughput": 46.624, + "latency": 2904.5 + } + ] }, { - "slug": "rekaai/reka-flash-3", - "hfSlug": "RekaAI/reka-flash-3", + "slug": "openai/gpt-3.5-turbo-instruct", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-12T20:53:33.296745+00:00", + "createdAt": "2023-09-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Reka: Flash 3 (free)", - "shortName": "Flash 3 (free)", - "author": "rekaai", - "description": "Reka Flash 3 is a general-purpose, instruction-tuned large language model with 21 billion parameters, developed by Reka. It excels at general chat, coding tasks, instruction-following, and function calling. Featuring a 32K context length and optimized through reinforcement learning (RLOO), it provides competitive performance comparable to proprietary models within a smaller parameter footprint. Ideal for low-latency, local, or on-device deployments, Reka Flash 3 is compact, supports efficient quantization (down to 11GB at 4-bit precision), and employs explicit reasoning tags (\"\") to indicate its internal thought process.\n\nReka Flash 3 is primarily an English model with limited multilingual understanding capabilities. The model weights are released under the Apache 2.0 license.", + "name": "OpenAI: GPT-3.5 Turbo Instruct", + "shortName": "GPT-3.5 Turbo Instruct", + "author": "openai", + "description": "This model is a variant of GPT-3.5 Turbo tuned for instructional prompts and omitting chat-related optimizations. Training data: up to Sep 2021.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 4095, "inputModalities": [ "text" ], @@ -69288,34 +72172,36 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "GPT", + "instructType": "chatml", "defaultSystem": null, "defaultStops": [ - "" + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "rekaai/reka-flash-3", + "permaslug": "openai/gpt-3.5-turbo-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "4f75b624-7616-44ba-8152-fee614887754", - "name": "Chutes | rekaai/reka-flash-3:free", - "contextLength": 32768, + "id": "39594b4d-6c7e-4ccd-aeed-79f9b3ca5819", + "name": "OpenAI | openai/gpt-3.5-turbo-instruct", + "contextLength": 4095, "model": { - "slug": "rekaai/reka-flash-3", - "hfSlug": "RekaAI/reka-flash-3", + "slug": "openai/gpt-3.5-turbo-instruct", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-12T20:53:33.296745+00:00", + "createdAt": "2023-09-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Reka: Flash 3", - "shortName": "Flash 3", - "author": "rekaai", - "description": "Reka Flash 3 is a general-purpose, instruction-tuned large language model with 21 billion parameters, developed by Reka. It excels at general chat, coding tasks, instruction-following, and function calling. Featuring a 32K context length and optimized through reinforcement learning (RLOO), it provides competitive performance comparable to proprietary models within a smaller parameter footprint. Ideal for low-latency, local, or on-device deployments, Reka Flash 3 is compact, supports efficient quantization (down to 11GB at 4-bit precision), and employs explicit reasoning tags (\"\") to indicate its internal thought process.\n\nReka Flash 3 is primarily an English model with limited multilingual understanding capabilities. The model weights are released under the Apache 2.0 license.", + "name": "OpenAI: GPT-3.5 Turbo Instruct", + "shortName": "GPT-3.5 Turbo Instruct", + "author": "openai", + "description": "This model is a variant of GPT-3.5 Turbo tuned for instructional prompts and omitting chat-related optimizations. Training data: up to Sep 2021.", "modelVersionGroupId": null, - "contextLength": 32000, + "contextLength": 4095, "inputModalities": [ "text" ], @@ -69323,91 +72209,100 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "GPT", + "instructType": "chatml", "defaultSystem": null, "defaultStops": [ - "" + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "rekaai/reka-flash-3", + "permaslug": "openai/gpt-3.5-turbo-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "rekaai/reka-flash-3:free", - "modelVariantPermaslug": "rekaai/reka-flash-3:free", - "providerName": "Chutes", + "modelVariantSlug": "openai/gpt-3.5-turbo-instruct", + "modelVariantPermaslug": "openai/gpt-3.5-turbo-instruct", + "providerName": "OpenAI", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { - "training": true, - "retainsPrompts": true - } + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": false, - "group": "Chutes", + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Chutes", - "providerModelId": "RekaAI/reka-flash-3", - "providerGroup": "Chutes", - "quantization": "bf16", - "variant": "free", - "isFree": true, + "providerDisplayName": "OpenAI", + "providerModelId": "gpt-3.5-turbo-instruct", + "providerGroup": "OpenAI", + "quantization": "unknown", + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", "seed", - "top_k", - "min_p", - "repetition_penalty", - "logprobs", "logit_bias", - "top_logprobs" + "logprobs", + "top_logprobs", + "response_format" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "training": true, - "retainsPrompts": true + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.0000015", + "completion": "0.000002", "image": "0", "request": "0", "webSearch": "0", @@ -69419,7 +72314,7 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, @@ -69432,26 +72327,63 @@ export default { }, "sorting": { "topWeekly": 257, - "newest": 92, - "throughputHighToLow": 120, - "latencyLowToHigh": 254, - "pricingLowToHigh": 40, - "pricingHighToLow": 288 + "newest": 303, + "throughputHighToLow": 39, + "latencyLowToHigh": 100, + "pricingLowToHigh": 236, + "pricingHighToLow": 82 }, - "providers": [] + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", + "providers": [ + { + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": "unknown", + "context": 4095, + "maxCompletionTokens": 4096, + "providerModelId": "gpt-3.5-turbo-instruct", + "pricing": { + "prompt": "0.0000015", + "completion": "0.000002", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" + ], + "inputCost": 1.5, + "outputCost": 2, + "throughput": 144.168, + "latency": 615.5 + } + ] }, { - "slug": "openai/gpt-3.5-turbo-0613", + "slug": "openai/gpt-4-turbo-preview", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-01-25T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-3.5 Turbo (older v0613)", - "shortName": "GPT-3.5 Turbo (older v0613)", + "name": "OpenAI: GPT-4 Turbo Preview", + "shortName": "GPT-4 Turbo Preview", "author": "openai", - "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.", + "description": "The preview GPT-4 model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more. Training data: up to Dec 2023.\n\n**Note:** heavily rate limited by OpenAI while in preview.", "modelVersionGroupId": null, - "contextLength": 4095, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -69466,25 +72398,25 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-3.5-turbo-0613", + "permaslug": "openai/gpt-4-turbo-preview", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "f29e5bba-56f3-408a-a036-185e3dc20be3", - "name": "Azure | openai/gpt-3.5-turbo-0613", - "contextLength": 4095, + "id": "003933da-395a-48eb-86a3-0c4ec486d67f", + "name": "OpenAI | openai/gpt-4-turbo-preview", + "contextLength": 128000, "model": { - "slug": "openai/gpt-3.5-turbo-0613", + "slug": "openai/gpt-4-turbo-preview", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-01-25T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-3.5 Turbo (older v0613)", - "shortName": "GPT-3.5 Turbo (older v0613)", + "name": "OpenAI: GPT-4 Turbo Preview", + "shortName": "GPT-4 Turbo Preview", "author": "openai", - "description": "GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.\n\nTraining data up to Sep 2021.", + "description": "The preview GPT-4 model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more. Training data: up to Dec 2023.\n\n**Note:** heavily rate limited by OpenAI while in preview.", "modelVersionGroupId": null, - "contextLength": 4095, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -69499,46 +72431,49 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-3.5-turbo-0613", + "permaslug": "openai/gpt-4-turbo-preview", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/gpt-3.5-turbo-0613", - "modelVariantPermaslug": "openai/gpt-3.5-turbo-0613", - "providerName": "Azure", + "modelVariantSlug": "openai/gpt-4-turbo-preview", + "modelVariantPermaslug": "openai/gpt-4-turbo-preview", + "providerName": "OpenAI", "providerInfo": { - "name": "Azure", - "displayName": "Azure", - "slug": "azure", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.microsoft.com/en-us/legal/terms-of-use?oneroute=true", - "privacyPolicyUrl": "https://www.microsoft.com/en-us/privacy/privacystatement", - "dataPolicyUrl": "https://learn.microsoft.com/en-us/legal/cognitive-services/openai/data-privacy?tabs=azure-portal", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false - } + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, + "hasCompletions": true, "isAbortable": true, - "moderationRequired": false, - "group": "Azure", + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], - "isMultipartSupported": false, - "statusPageUrl": "https://status.azure.com/", + "isMultipartSupported": true, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Azure.svg" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Azure", - "providerModelId": "gpt-35-turbo", - "providerGroup": "Azure", + "providerDisplayName": "OpenAI", + "providerModelId": "gpt-4-turbo-preview", + "providerGroup": "OpenAI", "quantization": "unknown", "variant": "standard", "isFree": false, @@ -69564,21 +72499,24 @@ export default { "structured_outputs" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://www.microsoft.com/en-us/legal/terms-of-use?oneroute=true", - "privacyPolicyUrl": "https://www.microsoft.com/en-us/privacy/privacystatement", - "dataPolicyUrl": "https://learn.microsoft.com/en-us/legal/cognitive-services/openai/data-privacy?tabs=azure-portal", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, + "requiresUserIds": true, "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.000001", - "completion": "0.000002", + "prompt": "0.00001", + "completion": "0.00003", "image": "0", "request": "0", "webSearch": "0", @@ -69591,10 +72529,10 @@ export default { "isDisabled": false, "supportsToolParameters": true, "supportsReasoning": false, - "supportsMultipart": false, + "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -69604,22 +72542,24 @@ export default { "sorting": { "topWeekly": 258, "newest": 286, - "throughputHighToLow": 35, - "latencyLowToHigh": 149, - "pricingLowToHigh": 224, - "pricingHighToLow": 98 + "throughputHighToLow": 243, + "latencyLowToHigh": 167, + "pricingLowToHigh": 303, + "pricingHighToLow": 15 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "Azure", - "slug": "azure", + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", "quantization": "unknown", - "context": 4095, + "context": 128000, "maxCompletionTokens": 4096, - "providerModelId": "gpt-35-turbo", + "providerModelId": "gpt-4-turbo-preview", "pricing": { - "prompt": "0.000001", - "completion": "0.000002", + "prompt": "0.00001", + "completion": "0.00003", "image": "0", "request": "0", "webSearch": "0", @@ -69642,121 +72582,125 @@ export default { "response_format", "structured_outputs" ], - "inputCost": 1, - "outputCost": 2 + "inputCost": 10, + "outputCost": 30, + "throughput": 31.919, + "latency": 1061 } ] }, { - "slug": "jondurbin/airoboros-l2-70b", - "hfSlug": "jondurbin/airoboros-l2-70b-2.2.1", + "slug": "google/gemma-3-1b-it", + "hfSlug": "google/gemma-3-1b-it", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-10-29T00:00:00+00:00", + "createdAt": "2025-03-14T14:45:56.842499+00:00", "hfUpdatedAt": null, - "name": "Airoboros 70B", - "shortName": "Airoboros 70B", - "author": "jondurbin", - "description": "A Llama 2 70B fine-tune using synthetic data (the Airoboros dataset).\n\nCurrently based on [jondurbin/airoboros-l2-70b](https://huggingface.co/jondurbin/airoboros-l2-70b-2.2.1), but might get updated in the future.", - "modelVersionGroupId": null, - "contextLength": 4096, + "name": "Google: Gemma 3 1B (free)", + "shortName": "Gemma 3 1B (free)", + "author": "google", + "description": "Gemma 3 1B is the smallest of the new Gemma 3 family. It handles context windows up to 32k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Note: Gemma 3 1B is not multimodal. For the smallest multimodal Gemma 3 model, please see [Gemma 3 4B](google/gemma-3-4b-it)", + "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", + "contextLength": 32768, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Llama2", - "instructType": "airoboros", + "group": "Gemini", + "instructType": "gemma", "defaultSystem": null, "defaultStops": [ - "USER:", - "" + "", + "", + "" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "jondurbin/airoboros-l2-70b", + "permaslug": "google/gemma-3-1b-it", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "08737538-335d-4639-b6e9-cc34a8134e30", - "name": "Novita | jondurbin/airoboros-l2-70b", - "contextLength": 4096, + "id": "19c49025-8f42-48eb-ae62-5b03cfb165ee", + "name": "Chutes | google/gemma-3-1b-it:free", + "contextLength": 32768, "model": { - "slug": "jondurbin/airoboros-l2-70b", - "hfSlug": "jondurbin/airoboros-l2-70b-2.2.1", + "slug": "google/gemma-3-1b-it", + "hfSlug": "google/gemma-3-1b-it", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-10-29T00:00:00+00:00", + "createdAt": "2025-03-14T14:45:56.842499+00:00", "hfUpdatedAt": null, - "name": "Airoboros 70B", - "shortName": "Airoboros 70B", - "author": "jondurbin", - "description": "A Llama 2 70B fine-tune using synthetic data (the Airoboros dataset).\n\nCurrently based on [jondurbin/airoboros-l2-70b](https://huggingface.co/jondurbin/airoboros-l2-70b-2.2.1), but might get updated in the future.", - "modelVersionGroupId": null, - "contextLength": 4096, + "name": "Google: Gemma 3 1B", + "shortName": "Gemma 3 1B", + "author": "google", + "description": "Gemma 3 1B is the smallest of the new Gemma 3 family. It handles context windows up to 32k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Note: Gemma 3 1B is not multimodal. For the smallest multimodal Gemma 3 model, please see [Gemma 3 4B](google/gemma-3-4b-it)", + "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", + "contextLength": 32000, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Llama2", - "instructType": "airoboros", + "group": "Gemini", + "instructType": "gemma", "defaultSystem": null, "defaultStops": [ - "USER:", - "" + "", + "", + "" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "jondurbin/airoboros-l2-70b", + "permaslug": "google/gemma-3-1b-it", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "jondurbin/airoboros-l2-70b", - "modelVariantPermaslug": "jondurbin/airoboros-l2-70b", - "providerName": "Novita", + "modelVariantSlug": "google/gemma-3-1b-it:free", + "modelVariantPermaslug": "google/gemma-3-1b-it:free", + "providerName": "Chutes", "providerInfo": { - "name": "Novita", - "displayName": "NovitaAI", - "slug": "novita", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Novita", + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.novita.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "NovitaAI", - "providerModelId": "jondurbin/airoboros-l2-70b", - "providerGroup": "Novita", - "quantization": "unknown", - "variant": "standard", - "isFree": false, + "providerDisplayName": "Chutes", + "providerModelId": "unsloth/gemma-3-1b-it", + "providerGroup": "Chutes", + "quantization": "bf16", + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 8192, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -69770,21 +72714,24 @@ export default { "top_k", "min_p", "repetition_penalty", - "logit_bias" + "logprobs", + "logit_bias", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false + "training": true, + "retainsPrompts": true }, - "training": false + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.0000005", - "completion": "0.0000005", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -69803,52 +72750,21 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 259, - "newest": 302, - "throughputHighToLow": 185, - "latencyLowToHigh": 275, - "pricingLowToHigh": 179, - "pricingHighToLow": 140 + "newest": 82, + "throughputHighToLow": 9, + "latencyLowToHigh": 149, + "pricingLowToHigh": 37, + "pricingHighToLow": 285 }, - "providers": [ - { - "name": "NovitaAI", - "slug": "novitaAi", - "quantization": "unknown", - "context": 4096, - "maxCompletionTokens": null, - "providerModelId": "jondurbin/airoboros-l2-70b", - "pricing": { - "prompt": "0.0000005", - "completion": "0.0000005", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logit_bias" - ], - "inputCost": 0.5, - "outputCost": 0.5 - } - ] + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", + "providers": [] }, { "slug": "mistralai/mistral-medium", @@ -70012,15 +72928,17 @@ export default { }, "sorting": { "topWeekly": 260, - "newest": 289, - "throughputHighToLow": 189, - "latencyLowToHigh": 47, + "newest": 288, + "throughputHighToLow": 177, + "latencyLowToHigh": 62, "pricingLowToHigh": 266, "pricingHighToLow": 52 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { "name": "Mistral", + "icon": "https://openrouter.ai/images/icons/Mistral.png", "slug": "mistral", "quantization": "unknown", "context": 32768, @@ -70049,22 +72967,24 @@ export default { "seed" ], "inputCost": 2.75, - "outputCost": 8.1 + "outputCost": 8.1, + "throughput": 46.868, + "latency": 458 } ] }, { - "slug": "nousresearch/nous-hermes-2-mixtral-8x7b-dpo", - "hfSlug": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO", + "slug": "openai/o1-mini-2024-09-12", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-01-16T00:00:00+00:00", + "createdAt": "2024-09-12T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Nous: Hermes 2 Mixtral 8x7B DPO", - "shortName": "Hermes 2 Mixtral 8x7B DPO", - "author": "nousresearch", - "description": "Nous Hermes 2 Mixtral 8x7B DPO is the new flagship Nous Research model trained over the [Mixtral 8x7B MoE LLM](/models/mistralai/mixtral-8x7b).\n\nThe model was trained on over 1,000,000 entries of primarily [GPT-4](/models/openai/gpt-4) generated data, as well as other high quality data from open datasets across the AI landscape, achieving state of the art performance on a variety of tasks.\n\n#moe", + "name": "OpenAI: o1-mini (2024-09-12)", + "shortName": "o1-mini (2024-09-12)", + "author": "openai", + "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -70072,36 +72992,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "chatml", + "group": "GPT", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "nousresearch/nous-hermes-2-mixtral-8x7b-dpo", + "permaslug": "openai/o1-mini-2024-09-12", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "5ec74ebb-b872-4d66-a93a-3d2696ecc664", - "name": "Together | nousresearch/nous-hermes-2-mixtral-8x7b-dpo", - "contextLength": 32768, + "id": "66ac30ec-dd07-4b8d-9dc0-5c7369bb2efe", + "name": "OpenAI | openai/o1-mini-2024-09-12", + "contextLength": 128000, "model": { - "slug": "nousresearch/nous-hermes-2-mixtral-8x7b-dpo", - "hfSlug": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO", + "slug": "openai/o1-mini-2024-09-12", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-01-16T00:00:00+00:00", + "createdAt": "2024-09-12T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Nous: Hermes 2 Mixtral 8x7B DPO", - "shortName": "Hermes 2 Mixtral 8x7B DPO", - "author": "nousresearch", - "description": "Nous Hermes 2 Mixtral 8x7B DPO is the new flagship Nous Research model trained over the [Mixtral 8x7B MoE LLM](/models/mistralai/mixtral-8x7b).\n\nThe model was trained on over 1,000,000 entries of primarily [GPT-4](/models/openai/gpt-4) generated data, as well as other high quality data from open datasets across the AI landscape, achieving state of the art performance on a variety of tasks.\n\n#moe", + "name": "OpenAI: o1-mini (2024-09-12)", + "shortName": "o1-mini (2024-09-12)", + "author": "openai", + "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -70109,94 +73025,90 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "chatml", + "group": "GPT", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "nousresearch/nous-hermes-2-mixtral-8x7b-dpo", + "permaslug": "openai/o1-mini-2024-09-12", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "nousresearch/nous-hermes-2-mixtral-8x7b-dpo", - "modelVariantPermaslug": "nousresearch/nous-hermes-2-mixtral-8x7b-dpo", - "providerName": "Together", + "modelVariantSlug": "openai/o1-mini-2024-09-12", + "modelVariantPermaslug": "openai/o1-mini-2024-09-12", + "providerName": "OpenAI", "providerInfo": { - "name": "Together", - "displayName": "Together", - "slug": "together", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false - } + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": false, - "group": "Together", + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Together", - "providerModelId": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO", - "providerGroup": "Together", + "providerDisplayName": "OpenAI", + "providerModelId": "o1-mini-2024-09-12", + "providerGroup": "OpenAI", "quantization": "unknown", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 2048, + "maxCompletionTokens": 65536, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "seed", + "max_tokens" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, + "requiresUserIds": true, "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.0000006", - "completion": "0.0000006", + "prompt": "0.0000011", + "completion": "0.0000044", "image": "0", "request": "0", + "inputCacheRead": "0.00000055", "webSearch": "0", "internalReasoning": "0", "discount": 0 @@ -70219,59 +73131,55 @@ export default { }, "sorting": { "topWeekly": 261, - "newest": 287, - "throughputHighToLow": 55, - "latencyLowToHigh": 75, - "pricingLowToHigh": 191, - "pricingHighToLow": 127 + "newest": 210, + "throughputHighToLow": 316, + "latencyLowToHigh": 298, + "pricingLowToHigh": 233, + "pricingHighToLow": 89 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "Together", - "slug": "together", + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", "quantization": "unknown", - "context": 32768, - "maxCompletionTokens": 2048, - "providerModelId": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO", + "context": 128000, + "maxCompletionTokens": 65536, + "providerModelId": "o1-mini-2024-09-12", "pricing": { - "prompt": "0.0000006", - "completion": "0.0000006", + "prompt": "0.0000011", + "completion": "0.0000044", "image": "0", "request": "0", + "inputCacheRead": "0.00000055", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "seed", + "max_tokens" ], - "inputCost": 0.6, - "outputCost": 0.6 + "inputCost": 1.1, + "outputCost": 4.4, + "throughput": 108.9225, + "latency": 5499 } ] }, { - "slug": "openai/gpt-3.5-turbo-instruct", - "hfSlug": null, + "slug": "rekaai/reka-flash-3", + "hfSlug": "RekaAI/reka-flash-3", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-09-28T00:00:00+00:00", + "createdAt": "2025-03-12T20:53:33.296745+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-3.5 Turbo Instruct", - "shortName": "GPT-3.5 Turbo Instruct", - "author": "openai", - "description": "This model is a variant of GPT-3.5 Turbo tuned for instructional prompts and omitting chat-related optimizations. Training data: up to Sep 2021.", + "name": "Reka: Flash 3 (free)", + "shortName": "Flash 3 (free)", + "author": "rekaai", + "description": "Reka Flash 3 is a general-purpose, instruction-tuned large language model with 21 billion parameters, developed by Reka. It excels at general chat, coding tasks, instruction-following, and function calling. Featuring a 32K context length and optimized through reinforcement learning (RLOO), it provides competitive performance comparable to proprietary models within a smaller parameter footprint. Ideal for low-latency, local, or on-device deployments, Reka Flash 3 is compact, supports efficient quantization (down to 11GB at 4-bit precision), and employs explicit reasoning tags (\"\") to indicate its internal thought process.\n\nReka Flash 3 is primarily an English model with limited multilingual understanding capabilities. The model weights are released under the Apache 2.0 license.", "modelVersionGroupId": null, - "contextLength": 4095, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -70279,36 +73187,34 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": "chatml", + "group": "Other", + "instructType": null, "defaultSystem": null, "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" + "" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-3.5-turbo-instruct", + "permaslug": "rekaai/reka-flash-3", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "39594b4d-6c7e-4ccd-aeed-79f9b3ca5819", - "name": "OpenAI | openai/gpt-3.5-turbo-instruct", - "contextLength": 4095, + "id": "4f75b624-7616-44ba-8152-fee614887754", + "name": "Chutes | rekaai/reka-flash-3:free", + "contextLength": 32768, "model": { - "slug": "openai/gpt-3.5-turbo-instruct", - "hfSlug": null, + "slug": "rekaai/reka-flash-3", + "hfSlug": "RekaAI/reka-flash-3", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-09-28T00:00:00+00:00", + "createdAt": "2025-03-12T20:53:33.296745+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-3.5 Turbo Instruct", - "shortName": "GPT-3.5 Turbo Instruct", - "author": "openai", - "description": "This model is a variant of GPT-3.5 Turbo tuned for instructional prompts and omitting chat-related optimizations. Training data: up to Sep 2021.", + "name": "Reka: Flash 3", + "shortName": "Flash 3", + "author": "rekaai", + "description": "Reka Flash 3 is a general-purpose, instruction-tuned large language model with 21 billion parameters, developed by Reka. It excels at general chat, coding tasks, instruction-following, and function calling. Featuring a 32K context length and optimized through reinforcement learning (RLOO), it provides competitive performance comparable to proprietary models within a smaller parameter footprint. Ideal for low-latency, local, or on-device deployments, Reka Flash 3 is compact, supports efficient quantization (down to 11GB at 4-bit precision), and employs explicit reasoning tags (\"\") to indicate its internal thought process.\n\nReka Flash 3 is primarily an English model with limited multilingual understanding capabilities. The model weights are released under the Apache 2.0 license.", "modelVersionGroupId": null, - "contextLength": 4095, + "contextLength": 32000, "inputModalities": [ "text" ], @@ -70316,100 +73222,258 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": "chatml", + "group": "Other", + "instructType": null, "defaultSystem": null, "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" + "" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "rekaai/reka-flash-3", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "rekaai/reka-flash-3:free", + "modelVariantPermaslug": "rekaai/reka-flash-3:free", + "providerName": "Chutes", + "providerInfo": { + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://chutes.ai/tos", + "paidModels": { + "training": true, + "retainsPrompts": true + } + }, + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "Chutes", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + } + }, + "providerDisplayName": "Chutes", + "providerModelId": "RekaAI/reka-flash-3", + "providerGroup": "Chutes", + "quantization": "bf16", + "variant": "free", + "isFree": true, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty", + "logprobs", + "logit_bias", + "top_logprobs" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://chutes.ai/tos", + "paidModels": { + "training": true, + "retainsPrompts": true + }, + "training": true, + "retainsPrompts": true + }, + "pricing": { + "prompt": "0", + "completion": "0", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": true, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 262, + "newest": 92, + "throughputHighToLow": 114, + "latencyLowToHigh": 255, + "pricingLowToHigh": 40, + "pricingHighToLow": 288 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [] + }, + { + "slug": "mistralai/mistral-7b-instruct-v0.1", + "hfSlug": "mistralai/Mistral-7B-Instruct-v0.1", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2023-09-28T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Mistral: Mistral 7B Instruct v0.1", + "shortName": "Mistral 7B Instruct v0.1", + "author": "mistralai", + "description": "A 7.3B parameter model that outperforms Llama 2 13B on all benchmarks, with optimizations for speed and context length.", + "modelVersionGroupId": "1d07cc56-c54d-4587-b785-5093496397a4", + "contextLength": 2824, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Mistral", + "instructType": "mistral", + "defaultSystem": null, + "defaultStops": [ + "[INST]", + "" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "mistralai/mistral-7b-instruct-v0.1", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "7f9cc99b-0c5c-4dc4-a662-07cebf628081", + "name": "Cloudflare | mistralai/mistral-7b-instruct-v0.1", + "contextLength": 2824, + "model": { + "slug": "mistralai/mistral-7b-instruct-v0.1", + "hfSlug": "mistralai/Mistral-7B-Instruct-v0.1", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2023-09-28T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Mistral: Mistral 7B Instruct v0.1", + "shortName": "Mistral 7B Instruct v0.1", + "author": "mistralai", + "description": "A 7.3B parameter model that outperforms Llama 2 13B on all benchmarks, with optimizations for speed and context length.", + "modelVersionGroupId": "1d07cc56-c54d-4587-b785-5093496397a4", + "contextLength": 4096, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Mistral", + "instructType": "mistral", + "defaultSystem": null, + "defaultStops": [ + "[INST]", + "" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-3.5-turbo-instruct", + "permaslug": "mistralai/mistral-7b-instruct-v0.1", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/gpt-3.5-turbo-instruct", - "modelVariantPermaslug": "openai/gpt-3.5-turbo-instruct", - "providerName": "OpenAI", + "modelVariantSlug": "mistralai/mistral-7b-instruct-v0.1", + "modelVariantPermaslug": "mistralai/mistral-7b-instruct-v0.1", + "providerName": "Cloudflare", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Cloudflare", + "displayName": "Cloudflare", + "slug": "cloudflare", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://www.cloudflare.com/service-specific-terms-developer-platform/#developer-platform-terms", + "privacyPolicyUrl": "https://developers.cloudflare.com/workers-ai/privacy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "training": false + } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, + "hasCompletions": false, "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", + "moderationRequired": false, + "group": "Cloudflare", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256" } }, - "providerDisplayName": "OpenAI", - "providerModelId": "gpt-3.5-turbo-instruct", - "providerGroup": "OpenAI", - "quantization": "unknown", + "providerDisplayName": "Cloudflare", + "providerModelId": "@cf/mistral/mistral-7b-instruct-v0.1", + "providerGroup": "Cloudflare", + "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", + "top_k", "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format" + "repetition_penalty", + "frequency_penalty", + "presence_penalty" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://www.cloudflare.com/service-specific-terms-developer-platform/#developer-platform-terms", + "privacyPolicyUrl": "https://developers.cloudflare.com/workers-ai/privacy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, "pricing": { - "prompt": "0.0000015", - "completion": "0.000002", + "prompt": "0.00000011", + "completion": "0.00000019", "image": "0", "request": "0", "webSearch": "0", @@ -70425,32 +73489,67 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 262, + "topWeekly": 263, "newest": 304, - "throughputHighToLow": 39, - "latencyLowToHigh": 80, - "pricingLowToHigh": 236, - "pricingHighToLow": 82 + "throughputHighToLow": 295, + "latencyLowToHigh": 152, + "pricingLowToHigh": 121, + "pricingHighToLow": 197 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [ { - "name": "OpenAI", - "slug": "openAi", + "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", + "slug": "cloudflare", + "quantization": null, + "context": 2824, + "maxCompletionTokens": null, + "providerModelId": "@cf/mistral/mistral-7b-instruct-v0.1", + "pricing": { + "prompt": "0.00000011", + "completion": "0.00000019", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "top_k", + "seed", + "repetition_penalty", + "frequency_penalty", + "presence_penalty" + ], + "inputCost": 0.11, + "outputCost": 0.19, + "throughput": 15.902, + "latency": 910 + }, + { + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", "quantization": "unknown", - "context": 4095, - "maxCompletionTokens": 4096, - "providerModelId": "gpt-3.5-turbo-instruct", + "context": 32768, + "maxCompletionTokens": 2048, + "providerModelId": "mistralai/Mistral-7B-Instruct-v0.1", "pricing": { - "prompt": "0.0000015", - "completion": "0.000002", + "prompt": "0.0000002", + "completion": "0.0000002", "image": "0", "request": "0", "webSearch": "0", @@ -70458,35 +73557,39 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "seed", + "top_k", + "repetition_penalty", "logit_bias", - "logprobs", - "top_logprobs", + "min_p", "response_format" ], - "inputCost": 1.5, - "outputCost": 2 + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 204.972, + "latency": 304 } ] }, { - "slug": "ai21/jamba-1.6-large", - "hfSlug": "ai21labs/AI21-Jamba-Large-1.6", + "slug": "microsoft/phi-3-medium-128k-instruct", + "hfSlug": "microsoft/Phi-3-medium-128k-instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-13T22:32:53+00:00", + "createdAt": "2024-05-24T00:00:00+00:00", "hfUpdatedAt": null, - "name": "AI21: Jamba 1.6 Large", - "shortName": "Jamba 1.6 Large", - "author": "ai21", - "description": "AI21 Jamba Large 1.6 is a high-performance hybrid foundation model combining State Space Models (Mamba) with Transformer attention mechanisms. Developed by AI21, it excels in extremely long-context handling (256K tokens), demonstrates superior inference efficiency (up to 2.5x faster than comparable models), and supports structured JSON output and tool-use capabilities. It has 94 billion active parameters (398 billion total), optimized quantization support (ExpertsInt8), and multilingual proficiency in languages such as English, Spanish, French, Portuguese, Italian, Dutch, German, Arabic, and Hebrew.\n\nUsage of this model is subject to the [Jamba Open Model License](https://www.ai21.com/licenses/jamba-open-model-license).", - "modelVersionGroupId": "cd1eb031-30bc-4e2e-aa06-3c20f986e5c7", - "contextLength": 256000, + "name": "Microsoft: Phi-3 Medium 128K Instruct", + "shortName": "Phi-3 Medium 128K Instruct", + "author": "microsoft", + "description": "Phi-3 128K Medium is a powerful 14-billion parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. In the MMLU-Pro eval, the model even comes close to a Llama3 70B level of performance.\n\nFor 4k context length, try [Phi-3 Medium 4K](/models/microsoft/phi-3-medium-4k-instruct).", + "modelVersionGroupId": null, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -70495,31 +73598,34 @@ export default { ], "hasTextOutput": true, "group": "Other", - "instructType": null, + "instructType": "phi3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|end|>", + "<|endoftext|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "ai21/jamba-1.6-large", + "permaslug": "microsoft/phi-3-medium-128k-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "f4dfbac5-0169-4bc8-9861-27dfef791169", - "name": "AI21 | ai21/jamba-1.6-large", - "contextLength": 256000, + "id": "1899ce7d-d477-442a-8528-50d8ec66df16", + "name": "Nebius | microsoft/phi-3-medium-128k-instruct", + "contextLength": 131072, "model": { - "slug": "ai21/jamba-1.6-large", - "hfSlug": "ai21labs/AI21-Jamba-Large-1.6", + "slug": "microsoft/phi-3-medium-128k-instruct", + "hfSlug": "microsoft/Phi-3-medium-128k-instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-13T22:32:53+00:00", + "createdAt": "2024-05-24T00:00:00+00:00", "hfUpdatedAt": null, - "name": "AI21: Jamba 1.6 Large", - "shortName": "Jamba 1.6 Large", - "author": "ai21", - "description": "AI21 Jamba Large 1.6 is a high-performance hybrid foundation model combining State Space Models (Mamba) with Transformer attention mechanisms. Developed by AI21, it excels in extremely long-context handling (256K tokens), demonstrates superior inference efficiency (up to 2.5x faster than comparable models), and supports structured JSON output and tool-use capabilities. It has 94 billion active parameters (398 billion total), optimized quantization support (ExpertsInt8), and multilingual proficiency in languages such as English, Spanish, French, Portuguese, Italian, Dutch, German, Arabic, and Hebrew.\n\nUsage of this model is subject to the [Jamba Open Model License](https://www.ai21.com/licenses/jamba-open-model-license).", - "modelVersionGroupId": "cd1eb031-30bc-4e2e-aa06-3c20f986e5c7", - "contextLength": 256000, + "name": "Microsoft: Phi-3 Medium 128K Instruct", + "shortName": "Phi-3 Medium 128K Instruct", + "author": "microsoft", + "description": "Phi-3 128K Medium is a powerful 14-billion parameter model designed for advanced language understanding, reasoning, and instruction following. Optimized through supervised fine-tuning and preference adjustments, it excels in tasks involving common sense, mathematics, logical reasoning, and code processing.\n\nAt time of release, Phi-3 Medium demonstrated state-of-the-art performance among lightweight models. In the MMLU-Pro eval, the model even comes close to a Llama3 70B level of performance.\n\nFor 4k context length, try [Phi-3 Medium 4K](/models/microsoft/phi-3-medium-4k-instruct).", + "modelVersionGroupId": null, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -70528,79 +73634,91 @@ export default { ], "hasTextOutput": true, "group": "Other", - "instructType": null, + "instructType": "phi3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|end|>", + "<|endoftext|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "ai21/jamba-1.6-large", + "permaslug": "microsoft/phi-3-medium-128k-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "ai21/jamba-1.6-large", - "modelVariantPermaslug": "ai21/jamba-1.6-large", - "providerName": "AI21", + "modelVariantSlug": "microsoft/phi-3-medium-128k-instruct", + "modelVariantPermaslug": "microsoft/phi-3-medium-128k-instruct", + "providerName": "Nebius", "providerInfo": { - "name": "AI21", - "displayName": "AI21", - "slug": "ai21", + "name": "Nebius", + "displayName": "Nebius AI Studio", + "slug": "nebius", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://www.ai21.com/privacy-policy/", - "termsOfServiceUrl": "https://www.ai21.com/terms-of-service/", + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false } }, - "headquarters": "IL", + "headquarters": "DE", "hasChatCompletions": true, - "hasCompletions": false, + "hasCompletions": true, "isAbortable": false, "moderationRequired": false, - "group": "AI21", + "group": "Nebius", "editors": [], "owners": [], - "isMultipartSupported": false, - "statusPageUrl": "https://status.ai21.com/", + "isMultipartSupported": true, + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://ai21.com/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "invertRequired": true } }, - "providerDisplayName": "AI21", - "providerModelId": "jamba-large-1.6", - "providerGroup": "AI21", - "quantization": "bf16", + "providerDisplayName": "Nebius AI Studio", + "providerModelId": "microsoft/Phi-3-medium-128k-instruct", + "providerGroup": "Nebius", + "quantization": "fp8", "variant": "standard", "isFree": false, "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", - "stop" + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://www.ai21.com/privacy-policy/", - "termsOfServiceUrl": "https://www.ai21.com/terms-of-service/", + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false }, - "training": false + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0.000002", - "completion": "0.000008", + "prompt": "0.0000001", + "completion": "0.0000003", "image": "0", "request": "0", "webSearch": "0", @@ -70611,37 +73729,75 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, - "supportsMultipart": false, + "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 263, - "newest": 85, - "throughputHighToLow": 161, - "latencyLowToHigh": 162, - "pricingLowToHigh": 248, - "pricingHighToLow": 68 + "topWeekly": 264, + "newest": 254, + "throughputHighToLow": 49, + "latencyLowToHigh": 7, + "pricingLowToHigh": 125, + "pricingHighToLow": 196 }, + "authorIcon": "https://openrouter.ai/images/icons/Microsoft.svg", "providers": [ { - "name": "AI21", - "slug": "ai21", - "quantization": "bf16", - "context": 256000, - "maxCompletionTokens": 4096, - "providerModelId": "jamba-large-1.6", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": "fp8", + "context": 131072, + "maxCompletionTokens": null, + "providerModelId": "microsoft/Phi-3-medium-128k-instruct", "pricing": { - "prompt": "0.000002", - "completion": "0.000008", + "prompt": "0.0000001", + "completion": "0.0000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 0.1, + "outputCost": 0.3, + "throughput": 127.3485, + "latency": 182.5 + }, + { + "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", + "slug": "azure", + "quantization": "unknown", + "context": 128000, + "maxCompletionTokens": null, + "providerModelId": "phi3-medium-128k", + "pricing": { + "prompt": "0.000001", + "completion": "0.000001", "image": "0", "request": "0", "webSearch": "0", @@ -70653,26 +73809,27 @@ export default { "tool_choice", "max_tokens", "temperature", - "top_p", - "stop" + "top_p" ], - "inputCost": 2, - "outputCost": 8 + "inputCost": 1, + "outputCost": 1, + "throughput": 57.342, + "latency": 596 } ] }, { - "slug": "mistralai/mistral-7b-instruct-v0.1", - "hfSlug": "mistralai/Mistral-7B-Instruct-v0.1", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-09-28T00:00:00+00:00", + "slug": "arcee-ai/virtuoso-large", + "hfSlug": "", + "updatedAt": "2025-05-05T21:48:18.128183+00:00", + "createdAt": "2025-05-05T21:01:25.294732+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral 7B Instruct v0.1", - "shortName": "Mistral 7B Instruct v0.1", - "author": "mistralai", - "description": "A 7.3B parameter model that outperforms Llama 2 13B on all benchmarks, with optimizations for speed and context length.", - "modelVersionGroupId": "1d07cc56-c54d-4587-b785-5093496397a4", - "contextLength": 2824, + "name": "Arcee AI: Virtuoso Large", + "shortName": "Virtuoso Large", + "author": "arcee-ai", + "description": "Virtuoso‑Large is Arcee's top‑tier general‑purpose LLM at 72 B parameters, tuned to tackle cross‑domain reasoning, creative writing and enterprise QA. Unlike many 70 B peers, it retains the 128 k context inherited from Qwen 2.5, letting it ingest books, codebases or financial filings wholesale. Training blended DeepSeek R1 distillation, multi‑epoch supervised fine‑tuning and a final DPO/RLHF alignment stage, yielding strong performance on BIG‑Bench‑Hard, GSM‑8K and long‑context Needle‑In‑Haystack tests. Enterprises use Virtuoso‑Large as the \"fallback\" brain in Conductor pipelines when other SLMs flag low confidence. Despite its size, aggressive KV‑cache optimizations keep first‑token latency in the low‑second range on 8× H100 nodes, making it a practical production‑grade powerhouse.", + "modelVersionGroupId": null, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -70680,35 +73837,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "mistral", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "[INST]", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-7b-instruct-v0.1", + "permaslug": "arcee-ai/virtuoso-large", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "7f9cc99b-0c5c-4dc4-a662-07cebf628081", - "name": "Cloudflare | mistralai/mistral-7b-instruct-v0.1", - "contextLength": 2824, + "id": "6ea266fd-81e1-4ad2-ba62-2f09c5762435", + "name": "Together | arcee-ai/virtuoso-large", + "contextLength": 131072, "model": { - "slug": "mistralai/mistral-7b-instruct-v0.1", - "hfSlug": "mistralai/Mistral-7B-Instruct-v0.1", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-09-28T00:00:00+00:00", + "slug": "arcee-ai/virtuoso-large", + "hfSlug": "", + "updatedAt": "2025-05-05T21:48:18.128183+00:00", + "createdAt": "2025-05-05T21:01:25.294732+00:00", "hfUpdatedAt": null, - "name": "Mistral: Mistral 7B Instruct v0.1", - "shortName": "Mistral 7B Instruct v0.1", - "author": "mistralai", - "description": "A 7.3B parameter model that outperforms Llama 2 13B on all benchmarks, with optimizations for speed and context length.", - "modelVersionGroupId": "1d07cc56-c54d-4587-b785-5093496397a4", - "contextLength": 4096, + "name": "Arcee AI: Virtuoso Large", + "shortName": "Virtuoso Large", + "author": "arcee-ai", + "description": "Virtuoso‑Large is Arcee's top‑tier general‑purpose LLM at 72 B parameters, tuned to tackle cross‑domain reasoning, creative writing and enterprise QA. Unlike many 70 B peers, it retains the 128 k context inherited from Qwen 2.5, letting it ingest books, codebases or financial filings wholesale. Training blended DeepSeek R1 distillation, multi‑epoch supervised fine‑tuning and a final DPO/RLHF alignment stage, yielding strong performance on BIG‑Bench‑Hard, GSM‑8K and long‑context Needle‑In‑Haystack tests. Enterprises use Virtuoso‑Large as the \"fallback\" brain in Conductor pipelines when other SLMs flag low confidence. Despite its size, aggressive KV‑cache optimizations keep first‑token latency in the low‑second range on 8× H100 nodes, making it a practical production‑grade powerhouse.", + "modelVersionGroupId": null, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -70716,41 +73870,39 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "mistral", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "[INST]", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "mistralai/mistral-7b-instruct-v0.1", + "permaslug": "arcee-ai/virtuoso-large", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "mistralai/mistral-7b-instruct-v0.1", - "modelVariantPermaslug": "mistralai/mistral-7b-instruct-v0.1", - "providerName": "Cloudflare", + "modelVariantSlug": "arcee-ai/virtuoso-large", + "modelVariantPermaslug": "arcee-ai/virtuoso-large", + "providerName": "Together", "providerInfo": { - "name": "Cloudflare", - "displayName": "Cloudflare", - "slug": "cloudflare", + "name": "Together", + "displayName": "Together", + "slug": "together", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.cloudflare.com/service-specific-terms-developer-platform/#developer-platform-terms", - "privacyPolicyUrl": "https://developers.cloudflare.com/workers-ai/privacy", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, + "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Cloudflare", + "group": "Together", "editors": [], "owners": [], "isMultipartSupported": true, @@ -70758,43 +73910,48 @@ export default { "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" } }, - "providerDisplayName": "Cloudflare", - "providerModelId": "@cf/mistral/mistral-7b-instruct-v0.1", - "providerGroup": "Cloudflare", + "providerDisplayName": "Together", + "providerModelId": "arcee-ai/virtuoso-large", + "providerGroup": "Together", "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 64000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "top_k", - "seed", "repetition_penalty", - "frequency_penalty", - "presence_penalty" + "logit_bias", + "min_p", + "response_format" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.cloudflare.com/service-specific-terms-developer-platform/#developer-platform-terms", - "privacyPolicyUrl": "https://developers.cloudflare.com/workers-ai/privacy", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false }, - "training": false + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0.00000011", - "completion": "0.00000019", + "prompt": "0.00000075", + "completion": "0.0000012", "image": "0", "request": "0", "webSearch": "0", @@ -70810,7 +73967,7 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { "supportedParameters": {}, @@ -70819,53 +73976,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 264, - "newest": 303, - "throughputHighToLow": 295, - "latencyLowToHigh": 133, - "pricingLowToHigh": 121, - "pricingHighToLow": 197 + "topWeekly": 265, + "newest": 6, + "throughputHighToLow": 101, + "latencyLowToHigh": 107, + "pricingLowToHigh": 199, + "pricingHighToLow": 119 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://arcee.ai/\\u0026size=256", "providers": [ - { - "name": "Cloudflare", - "slug": "cloudflare", - "quantization": null, - "context": 2824, - "maxCompletionTokens": null, - "providerModelId": "@cf/mistral/mistral-7b-instruct-v0.1", - "pricing": { - "prompt": "0.00000011", - "completion": "0.00000019", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "top_k", - "seed", - "repetition_penalty", - "frequency_penalty", - "presence_penalty" - ], - "inputCost": 0.11, - "outputCost": 0.19 - }, { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", - "quantization": "unknown", - "context": 32768, - "maxCompletionTokens": 2048, - "providerModelId": "mistralai/Mistral-7B-Instruct-v0.1", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 64000, + "providerModelId": "arcee-ai/virtuoso-large", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", + "prompt": "0.00000075", + "completion": "0.0000012", "image": "0", "request": "0", "webSearch": "0", @@ -70873,8 +74003,6 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -70887,8 +74015,10 @@ export default { "min_p", "response_format" ], - "inputCost": 0.2, - "outputCost": 0.2 + "inputCost": 0.75, + "outputCost": 1.2, + "throughput": 78.419, + "latency": 687.5 } ] }, @@ -71030,8 +74160,8 @@ export default { "retainsPrompts": true }, "pricing": { - "prompt": "0.00000009375", - "completion": "0.00000075", + "prompt": "0.00000015", + "completion": "0.0000009375", "image": "0", "request": "0", "webSearch": "0", @@ -71057,22 +74187,24 @@ export default { "sorting": { "topWeekly": 81, "newest": 263, - "throughputHighToLow": 148, - "latencyLowToHigh": 101, - "pricingLowToHigh": 136, - "pricingHighToLow": 182 + "throughputHighToLow": 144, + "latencyLowToHigh": 95, + "pricingLowToHigh": 152, + "pricingHighToLow": 166 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "Mancer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", "slug": "mancer", "quantization": "unknown", "context": 24576, "maxCompletionTokens": 2048, "providerModelId": "lumi-8b", "pricing": { - "prompt": "0.00000009375", - "completion": "0.00000075", + "prompt": "0.00000015", + "completion": "0.0000009375", "image": "0", "request": "0", "webSearch": "0", @@ -71093,19 +74225,22 @@ export default { "seed", "top_a" ], - "inputCost": 0.09, - "outputCost": 0.75 + "inputCost": 0.15, + "outputCost": 0.94, + "throughput": 58.175, + "latency": 642 }, { "name": "Mancer (private)", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", "slug": "mancer (private)", "quantization": "unknown", "context": 24576, "maxCompletionTokens": 2048, "providerModelId": "lumi-8b", "pricing": { - "prompt": "0.000000125", - "completion": "0.000001", + "prompt": "0.0000002", + "completion": "0.00000125", "image": "0", "request": "0", "webSearch": "0", @@ -71126,11 +74261,14 @@ export default { "seed", "top_a" ], - "inputCost": 0.13, - "outputCost": 1 + "inputCost": 0.2, + "outputCost": 1.25, + "throughput": 60.202, + "latency": 435 }, { "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", "slug": "featherless", "quantization": "fp8", "context": 8192, @@ -71158,191 +74296,9 @@ export default { "seed" ], "inputCost": 0.8, - "outputCost": 1.2 - } - ] - }, - { - "slug": "aion-labs/aion-1.0", - "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-04T19:32:37.000231+00:00", - "hfUpdatedAt": null, - "name": "AionLabs: Aion-1.0", - "shortName": "Aion-1.0", - "author": "aion-labs", - "description": "Aion-1.0 is a multi-model system designed for high performance across various tasks, including reasoning and coding. It is built on DeepSeek-R1, augmented with additional models and techniques such as Tree of Thoughts (ToT) and Mixture of Experts (MoE). It is Aion Lab's most powerful reasoning model.", - "modelVersionGroupId": null, - "contextLength": 131072, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "aion-labs/aion-1.0", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "68716168-5661-4ed3-a841-caa166d7913e", - "name": "AionLabs | aion-labs/aion-1.0", - "contextLength": 131072, - "model": { - "slug": "aion-labs/aion-1.0", - "hfSlug": "", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-04T19:32:37.000231+00:00", - "hfUpdatedAt": null, - "name": "AionLabs: Aion-1.0", - "shortName": "Aion-1.0", - "author": "aion-labs", - "description": "Aion-1.0 is a multi-model system designed for high performance across various tasks, including reasoning and coding. It is built on DeepSeek-R1, augmented with additional models and techniques such as Tree of Thoughts (ToT) and Mixture of Experts (MoE). It is Aion Lab's most powerful reasoning model.", - "modelVersionGroupId": null, - "contextLength": 32768, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "aion-labs/aion-1.0", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "aion-labs/aion-1.0", - "modelVariantPermaslug": "aion-labs/aion-1.0", - "providerName": "AionLabs", - "providerInfo": { - "name": "AionLabs", - "displayName": "AionLabs", - "slug": "aion-labs", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://www.aionlabs.ai/terms/", - "privacyPolicyUrl": "https://www.aionlabs.ai/privacy-policy/", - "paidModels": { - "training": false - } - }, - "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, - "moderationRequired": false, - "group": "AionLabs", - "editors": [], - "owners": [], - "isMultipartSupported": false, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.aionlabs.ai/&size=256" - } - }, - "providerDisplayName": "AionLabs", - "providerModelId": "aion-labs/aion-1.0", - "providerGroup": "AionLabs", - "quantization": null, - "variant": "standard", - "isFree": false, - "canAbort": false, - "maxPromptTokens": null, - "maxCompletionTokens": 32768, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://www.aionlabs.ai/terms/", - "privacyPolicyUrl": "https://www.aionlabs.ai/privacy-policy/", - "paidModels": { - "training": false - }, - "training": false - }, - "pricing": { - "prompt": "0.000004", - "completion": "0.000008", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": true, - "supportsMultipart": false, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": false, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 266, - "newest": 120, - "throughputHighToLow": 196, - "latencyLowToHigh": 205, - "pricingLowToHigh": 284, - "pricingHighToLow": 34 - }, - "providers": [ - { - "name": "AionLabs", - "slug": "aionLabs", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 32768, - "providerModelId": "aion-labs/aion-1.0", - "pricing": { - "prompt": "0.000004", - "completion": "0.000008", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning" - ], - "inputCost": 4, - "outputCost": 8 + "outputCost": 1.2, + "throughput": 25.6735, + "latency": 1299.5 } ] }, @@ -71503,15 +74459,17 @@ export default { }, "sorting": { "topWeekly": 267, - "newest": 296, - "throughputHighToLow": 291, - "latencyLowToHigh": 199, + "newest": 294, + "throughputHighToLow": 293, + "latencyLowToHigh": 224, "pricingLowToHigh": 298, "pricingHighToLow": 17 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", "slug": "anthropic", "quantization": "unknown", "context": 200000, @@ -71534,22 +74492,24 @@ export default { "stop" ], "inputCost": 8, - "outputCost": 24 + "outputCost": 24, + "throughput": 16.038, + "latency": 1532 } ] }, { - "slug": "thudm/glm-z1-9b", - "hfSlug": "thudm/glm-z1-9b-0414", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-04-25T17:12:20.695158+00:00", + "slug": "raifle/sorcererlm-8x22b", + "hfSlug": "rAIfle/SorcererLM-8x22b-bf16", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-11-08T22:31:23.953049+00:00", "hfUpdatedAt": null, - "name": "THUDM: GLM Z1 9B (free)", - "shortName": "GLM Z1 9B (free)", - "author": "thudm", - "description": "GLM-Z1-9B-0414 is a 9B-parameter language model developed by THUDM as part of the GLM-4 family. It incorporates techniques originally applied to larger GLM-Z1 models, including extended reinforcement learning, pairwise ranking alignment, and training on reasoning-intensive tasks such as mathematics, code, and logic. Despite its smaller size, it demonstrates strong performance on general-purpose reasoning tasks and outperforms many open-source models in its weight class.", + "name": "SorcererLM 8x22B", + "shortName": "SorcererLM 8x22B", + "author": "raifle", + "description": "SorcererLM is an advanced RP and storytelling model, built as a Low-rank 16-bit LoRA fine-tuned on [WizardLM-2 8x22B](/microsoft/wizardlm-2-8x22b).\n\n- Advanced reasoning and emotional intelligence for engaging and immersive interactions\n- Vivid writing capabilities enriched with spatial and contextual awareness\n- Enhanced narrative depth, promoting creative and dynamic storytelling", "modelVersionGroupId": null, - "contextLength": 32000, + "contextLength": 16000, "inputModalities": [ "text" ], @@ -71557,43 +74517,35 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "deepseek-r1", + "group": "Mistral", + "instructType": "vicuna", "defaultSystem": null, "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" + "USER:", + "" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "thudm/glm-z1-9b-0414", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - }, + "permaslug": "raifle/sorcererlm-8x22b", + "reasoningConfig": null, + "features": {}, "endpoint": { - "id": "8ad6c714-038b-432e-8719-d2cb7feb7187", - "name": "Novita | thudm/glm-z1-9b-0414:free", - "contextLength": 32000, + "id": "65206957-c772-4743-93ff-45900a190ddd", + "name": "Infermatic | raifle/sorcererlm-8x22b", + "contextLength": 16000, "model": { - "slug": "thudm/glm-z1-9b", - "hfSlug": "thudm/glm-z1-9b-0414", - "updatedAt": "2025-05-02T21:14:16.355933+00:00", - "createdAt": "2025-04-25T17:12:20.695158+00:00", + "slug": "raifle/sorcererlm-8x22b", + "hfSlug": "rAIfle/SorcererLM-8x22b-bf16", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-11-08T22:31:23.953049+00:00", "hfUpdatedAt": null, - "name": "THUDM: GLM Z1 9B", - "shortName": "GLM Z1 9B", - "author": "thudm", - "description": "GLM-Z1-9B-0414 is a 9B-parameter language model developed by THUDM as part of the GLM-4 family. It incorporates techniques originally applied to larger GLM-Z1 models, including extended reinforcement learning, pairwise ranking alignment, and training on reasoning-intensive tasks such as mathematics, code, and logic. Despite its smaller size, it demonstrates strong performance on general-purpose reasoning tasks and outperforms many open-source models in its weight class.", + "name": "SorcererLM 8x22B", + "shortName": "SorcererLM 8x22B", + "author": "raifle", + "description": "SorcererLM is an advanced RP and storytelling model, built as a Low-rank 16-bit LoRA fine-tuned on [WizardLM-2 8x22B](/microsoft/wizardlm-2-8x22b).\n\n- Advanced reasoning and emotional intelligence for engaging and immersive interactions\n- Vivid writing capabilities enriched with spatial and contextual awareness\n- Enhanced narrative depth, promoting creative and dynamic storytelling", "modelVersionGroupId": null, - "contextLength": 32000, + "contextLength": 16000, "inputModalities": [ "text" ], @@ -71601,66 +74553,57 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "deepseek-r1", + "group": "Mistral", + "instructType": "vicuna", "defaultSystem": null, "defaultStops": [ - "<｜User｜>", - "<｜end▁of▁sentence｜>" + "USER:", + "" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "thudm/glm-z1-9b-0414", - "reasoningConfig": { - "startToken": "", - "endToken": "" - }, - "features": { - "reasoningConfig": { - "startToken": "", - "endToken": "" - } - } + "permaslug": "raifle/sorcererlm-8x22b", + "reasoningConfig": null, + "features": {} }, - "modelVariantSlug": "thudm/glm-z1-9b:free", - "modelVariantPermaslug": "thudm/glm-z1-9b-0414:free", - "providerName": "Novita", + "modelVariantSlug": "raifle/sorcererlm-8x22b", + "modelVariantPermaslug": "raifle/sorcererlm-8x22b", + "providerName": "Infermatic", "providerInfo": { - "name": "Novita", - "displayName": "NovitaAI", - "slug": "novita", + "name": "Infermatic", + "displayName": "Infermatic", + "slug": "infermatic", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "privacyPolicyUrl": "https://infermatic.ai/privacy-policy/", + "termsOfServiceUrl": "https://infermatic.ai/terms-and-conditions/", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Novita", + "group": "Infermatic", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.novita.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://infermatic.ai/&size=256" } }, - "providerDisplayName": "NovitaAI", - "providerModelId": "thudm/glm-z1-9b-0414", - "providerGroup": "Novita", + "providerDisplayName": "Infermatic", + "providerModelId": "rAIfle-SorcererLM-8x22b-bf16", + "providerGroup": "Infermatic", "quantization": null, - "variant": "free", - "isFree": true, + "variant": "standard", + "isFree": false, "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, @@ -71670,30 +74613,30 @@ export default { "max_tokens", "temperature", "top_p", - "reasoning", - "include_reasoning", "stop", "frequency_penalty", "presence_penalty", - "seed", + "repetition_penalty", + "logit_bias", "top_k", "min_p", - "repetition_penalty", - "logit_bias" + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", + "privacyPolicyUrl": "https://infermatic.ai/privacy-policy/", + "termsOfServiceUrl": "https://infermatic.ai/terms-and-conditions/", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false }, - "training": false + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.0000045", + "completion": "0.0000045", "image": "0", "request": "0", "webSearch": "0", @@ -71705,40 +74648,76 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": true, + "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 268, - "newest": 34, - "throughputHighToLow": 204, - "latencyLowToHigh": 195, - "pricingLowToHigh": 15, - "pricingHighToLow": 263 + "newest": 174, + "throughputHighToLow": 160, + "latencyLowToHigh": 141, + "pricingLowToHigh": 285, + "pricingHighToLow": 33 }, - "providers": [] + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [ + { + "name": "Infermatic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://infermatic.ai/&size=256", + "slug": "infermatic", + "quantization": null, + "context": 16000, + "maxCompletionTokens": null, + "providerModelId": "rAIfle-SorcererLM-8x22b-bf16", + "pricing": { + "prompt": "0.0000045", + "completion": "0.0000045", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "logit_bias", + "top_k", + "min_p", + "seed" + ], + "inputCost": 4.5, + "outputCost": 4.5, + "throughput": 52.655, + "latency": 799 + } + ] }, { - "slug": "liquid/lfm-40b", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-30T00:00:00+00:00", + "slug": "thudm/glm-4-9b", + "hfSlug": "THUDM/GLM-4-9B-0414", + "updatedAt": "2025-04-25T17:11:00.514702+00:00", + "createdAt": "2025-04-25T17:10:23.758806+00:00", "hfUpdatedAt": null, - "name": "Liquid: LFM 40B MoE", - "shortName": "LFM 40B MoE", - "author": "liquid", - "description": "Liquid's 40.3B Mixture of Experts (MoE) model. Liquid Foundation Models (LFMs) are large neural networks built with computational units rooted in dynamic systems.\n\nLFMs are general-purpose AI models that can be used to model any kind of sequential data, including video, audio, text, time series, and signals.\n\nSee the [launch announcement](https://www.liquid.ai/liquid-foundation-models) for benchmarks and more info.", + "name": "THUDM: GLM 4 9B (free)", + "shortName": "GLM 4 9B (free)", + "author": "thudm", + "description": "GLM-4-9B-0414 is a 9 billion parameter language model from the GLM-4 series developed by THUDM. Trained using the same reinforcement learning and alignment strategies as its larger 32B counterparts, GLM-4-9B-0414 achieves high performance relative to its size, making it suitable for resource-constrained deployments that still require robust language understanding and generation capabilities.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 32000, "inputModalities": [ "text" ], @@ -71747,35 +74726,31 @@ export default { ], "hasTextOutput": true, "group": "Other", - "instructType": "chatml", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "liquid/lfm-40b", + "permaslug": "thudm/glm-4-9b-0414", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "7901899b-4845-4089-979d-34a1ef065875", - "name": "Liquid | liquid/lfm-40b", - "contextLength": 32768, + "id": "9998778a-6cd2-479c-8e48-cb349270eac6", + "name": "Novita | thudm/glm-4-9b-0414:free", + "contextLength": 32000, "model": { - "slug": "liquid/lfm-40b", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-30T00:00:00+00:00", + "slug": "thudm/glm-4-9b", + "hfSlug": "THUDM/GLM-4-9B-0414", + "updatedAt": "2025-04-25T17:11:00.514702+00:00", + "createdAt": "2025-04-25T17:10:23.758806+00:00", "hfUpdatedAt": null, - "name": "Liquid: LFM 40B MoE", - "shortName": "LFM 40B MoE", - "author": "liquid", - "description": "Liquid's 40.3B Mixture of Experts (MoE) model. Liquid Foundation Models (LFMs) are large neural networks built with computational units rooted in dynamic systems.\n\nLFMs are general-purpose AI models that can be used to model any kind of sequential data, including video, audio, text, time series, and signals.\n\nSee the [launch announcement](https://www.liquid.ai/liquid-foundation-models) for benchmarks and more info.", + "name": "THUDM: GLM 4 9B", + "shortName": "GLM 4 9B", + "author": "thudm", + "description": "GLM-4-9B-0414 is a 9 billion parameter language model from the GLM-4 series developed by THUDM. Trained using the same reinforcement learning and alignment strategies as its larger 32B counterparts, GLM-4-9B-0414 achieves high performance relative to its size, making it suitable for resource-constrained deployments that still require robust language understanding and generation capabilities.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 32000, "inputModalities": [ "text" ], @@ -71784,57 +74759,54 @@ export default { ], "hasTextOutput": true, "group": "Other", - "instructType": "chatml", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "liquid/lfm-40b", + "permaslug": "thudm/glm-4-9b-0414", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "liquid/lfm-40b", - "modelVariantPermaslug": "liquid/lfm-40b", - "providerName": "Liquid", + "modelVariantSlug": "thudm/glm-4-9b:free", + "modelVariantPermaslug": "thudm/glm-4-9b-0414:free", + "providerName": "Novita", "providerInfo": { - "name": "Liquid", - "displayName": "Liquid", - "slug": "liquid", + "name": "Novita", + "displayName": "NovitaAI", + "slug": "novita", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.liquid.ai/terms-conditions", - "privacyPolicyUrl": "https://www.liquid.ai/privacy-policy", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { "training": false } }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Liquid", + "group": "Novita", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.novita.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.liquid.ai/&size=256", + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://novita.ai/&size=256", "invertRequired": true } }, - "providerDisplayName": "Liquid", - "providerModelId": "lfm-40b", - "providerGroup": "Liquid", + "providerDisplayName": "NovitaAI", + "providerModelId": "thudm/glm-4-9b-0414", + "providerGroup": "Novita", "quantization": null, - "variant": "standard", - "isFree": false, + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": null, @@ -71850,21 +74822,22 @@ export default { "seed", "top_k", "min_p", - "repetition_penalty" + "repetition_penalty", + "logit_bias" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.liquid.ai/terms-conditions", - "privacyPolicyUrl": "https://www.liquid.ai/privacy-policy", + "termsOfServiceUrl": "https://novita.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://novita.ai/legal/privacy-policy", "paidModels": { "training": false }, "training": false }, "pricing": { - "prompt": "0.00000015", - "completion": "0.00000015", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -71883,227 +74856,146 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 269, - "newest": 194, - "throughputHighToLow": 20, - "latencyLowToHigh": 72, - "pricingLowToHigh": 134, - "pricingHighToLow": 185 + "newest": 35, + "throughputHighToLow": 193, + "latencyLowToHigh": 191, + "pricingLowToHigh": 16, + "pricingHighToLow": 264 }, - "providers": [ - { - "name": "Liquid", - "slug": "liquid", - "quantization": null, - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "lfm-40b", - "pricing": { - "prompt": "0.00000015", - "completion": "0.00000015", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty" - ], - "inputCost": 0.15, - "outputCost": 0.15 - }, - { - "name": "Lambda", - "slug": "lambda", - "quantization": "bf16", - "context": 65536, - "maxCompletionTokens": 65536, - "providerModelId": "lfm-40b", - "pricing": { - "prompt": "0.00000015", - "completion": "0.00000015", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format" - ], - "inputCost": 0.15, - "outputCost": 0.15 - } - ] + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [] }, { - "slug": "google/gemma-3-1b-it", - "hfSlug": "google/gemma-3-1b-it", + "slug": "aion-labs/aion-1.0", + "hfSlug": "", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-14T14:45:56.842499+00:00", + "createdAt": "2025-02-04T19:32:37.000231+00:00", "hfUpdatedAt": null, - "name": "Google: Gemma 3 1B (free)", - "shortName": "Gemma 3 1B (free)", - "author": "google", - "description": "Gemma 3 1B is the smallest of the new Gemma 3 family. It handles context windows up to 32k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Note: Gemma 3 1B is not multimodal. For the smallest multimodal Gemma 3 model, please see [Gemma 3 4B](google/gemma-3-4b-it)", - "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", - "contextLength": 32768, + "name": "AionLabs: Aion-1.0", + "shortName": "Aion-1.0", + "author": "aion-labs", + "description": "Aion-1.0 is a multi-model system designed for high performance across various tasks, including reasoning and coding. It is built on DeepSeek-R1, augmented with additional models and techniques such as Tree of Thoughts (ToT) and Mixture of Experts (MoE). It is Aion Lab's most powerful reasoning model.", + "modelVersionGroupId": null, + "contextLength": 131072, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Gemini", - "instructType": "gemma", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "", - "", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "google/gemma-3-1b-it", + "permaslug": "aion-labs/aion-1.0", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "19c49025-8f42-48eb-ae62-5b03cfb165ee", - "name": "Chutes | google/gemma-3-1b-it:free", - "contextLength": 32768, + "id": "68716168-5661-4ed3-a841-caa166d7913e", + "name": "AionLabs | aion-labs/aion-1.0", + "contextLength": 131072, "model": { - "slug": "google/gemma-3-1b-it", - "hfSlug": "google/gemma-3-1b-it", + "slug": "aion-labs/aion-1.0", + "hfSlug": "", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-14T14:45:56.842499+00:00", + "createdAt": "2025-02-04T19:32:37.000231+00:00", "hfUpdatedAt": null, - "name": "Google: Gemma 3 1B", - "shortName": "Gemma 3 1B", - "author": "google", - "description": "Gemma 3 1B is the smallest of the new Gemma 3 family. It handles context windows up to 32k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities, including structured outputs and function calling. Note: Gemma 3 1B is not multimodal. For the smallest multimodal Gemma 3 model, please see [Gemma 3 4B](google/gemma-3-4b-it)", - "modelVersionGroupId": "c99b277a-cfaf-4e93-9360-8a79cfa2b2c4", - "contextLength": 32000, + "name": "AionLabs: Aion-1.0", + "shortName": "Aion-1.0", + "author": "aion-labs", + "description": "Aion-1.0 is a multi-model system designed for high performance across various tasks, including reasoning and coding. It is built on DeepSeek-R1, augmented with additional models and techniques such as Tree of Thoughts (ToT) and Mixture of Experts (MoE). It is Aion Lab's most powerful reasoning model.", + "modelVersionGroupId": null, + "contextLength": 32768, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Gemini", - "instructType": "gemma", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "", - "", - "" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "google/gemma-3-1b-it", + "permaslug": "aion-labs/aion-1.0", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "google/gemma-3-1b-it:free", - "modelVariantPermaslug": "google/gemma-3-1b-it:free", - "providerName": "Chutes", + "modelVariantSlug": "aion-labs/aion-1.0", + "modelVariantPermaslug": "aion-labs/aion-1.0", + "providerName": "AionLabs", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "AionLabs", + "displayName": "AionLabs", + "slug": "aion-labs", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://www.aionlabs.ai/terms/", + "privacyPolicyUrl": "https://www.aionlabs.ai/privacy-policy/", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false } }, "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, + "hasCompletions": false, + "isAbortable": false, "moderationRequired": false, - "group": "Chutes", + "group": "AionLabs", "editors": [], "owners": [], - "isMultipartSupported": true, + "isMultipartSupported": false, "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.aionlabs.ai/&size=256" } }, - "providerDisplayName": "Chutes", - "providerModelId": "unsloth/gemma-3-1b-it", - "providerGroup": "Chutes", - "quantization": "bf16", - "variant": "free", - "isFree": true, - "canAbort": true, + "providerDisplayName": "AionLabs", + "providerModelId": "aion-labs/aion-1.0", + "providerGroup": "AionLabs", + "quantization": null, + "variant": "standard", + "isFree": false, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 8192, + "maxCompletionTokens": 32768, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "top_k", - "min_p", - "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" + "reasoning", + "include_reasoning" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://www.aionlabs.ai/terms/", + "privacyPolicyUrl": "https://www.aionlabs.ai/privacy-policy/", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false }, - "training": true, - "retainsPrompts": true + "training": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.000004", + "completion": "0.000008", "image": "0", "request": "0", "webSearch": "0", @@ -72115,11 +75007,11 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, + "supportsReasoning": true, + "supportsMultipart": false, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { "supportedParameters": {}, @@ -72129,136 +75021,186 @@ export default { }, "sorting": { "topWeekly": 270, - "newest": 82, - "throughputHighToLow": 10, - "latencyLowToHigh": 160, - "pricingLowToHigh": 37, - "pricingHighToLow": 285 + "newest": 120, + "throughputHighToLow": 191, + "latencyLowToHigh": 213, + "pricingLowToHigh": 284, + "pricingHighToLow": 34 }, - "providers": [] + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://www.aionlabs.ai/\\u0026size=256", + "providers": [ + { + "name": "AionLabs", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.aionlabs.ai/&size=256", + "slug": "aionLabs", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 32768, + "providerModelId": "aion-labs/aion-1.0", + "pricing": { + "prompt": "0.000004", + "completion": "0.000008", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "reasoning", + "include_reasoning" + ], + "inputCost": 4, + "outputCost": 8, + "throughput": 44.229, + "latency": 1461 + } + ] }, { - "slug": "opengvlab/internvl3-2b", - "hfSlug": "OpenGVLab/InternVL3-2B", - "updatedAt": "2025-04-30T13:55:48.070004+00:00", - "createdAt": "2025-04-30T13:30:07.912688+00:00", + "slug": "aetherwiing/mn-starcannon-12b", + "hfSlug": "aetherwiing/MN-12B-Starcannon-v2", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-08-13T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenGVLab: InternVL3 2B (free)", - "shortName": "InternVL3 2B (free)", - "author": "opengvlab", - "description": "The 2b version of the InternVL3 series, for an even higher inference speed and very reasonable performance. An advanced multimodal large language model (MLLM) series that demonstrates superior overall performance. Compared to InternVL 2.5, InternVL3 exhibits superior multimodal perception and reasoning capabilities, while further extending its multimodal capabilities to encompass tool usage, GUI agents, industrial image analysis, 3D vision perception, and more.", + "name": "Aetherwiing: Starcannon 12B", + "shortName": "Starcannon 12B", + "author": "aetherwiing", + "description": "Starcannon 12B v2 is a creative roleplay and story writing model, based on Mistral Nemo, using [nothingiisreal/mn-celeste-12b](/nothingiisreal/mn-celeste-12b) as a base, with [intervitens/mini-magnum-12b-v1.1](https://huggingface.co/intervitens/mini-magnum-12b-v1.1) merged in using the [TIES](https://arxiv.org/abs/2306.01708) method.\n\nAlthough more similar to Magnum overall, the model remains very creative, with a pleasant writing style. It is recommended for people wanting more variety than Magnum, and yet more verbose prose than Celeste.", "modelVersionGroupId": null, - "contextLength": 32000, + "contextLength": 16384, "inputModalities": [ - "image", "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Mistral", + "instructType": "chatml", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "opengvlab/internvl3-2b", + "permaslug": "aetherwiing/mn-starcannon-12b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "ec1cddcf-1619-457b-9008-bbadc41eb2bb", - "name": "Nineteen | opengvlab/internvl3-2b:free", - "contextLength": 32000, + "id": "1d498993-f126-4279-8bd7-66c2cfc88423", + "name": "Featherless | aetherwiing/mn-starcannon-12b", + "contextLength": 16384, "model": { - "slug": "opengvlab/internvl3-2b", - "hfSlug": "OpenGVLab/InternVL3-2B", - "updatedAt": "2025-04-30T13:55:48.070004+00:00", - "createdAt": "2025-04-30T13:30:07.912688+00:00", + "slug": "aetherwiing/mn-starcannon-12b", + "hfSlug": "aetherwiing/MN-12B-Starcannon-v2", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-08-13T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenGVLab: InternVL3 2B", - "shortName": "InternVL3 2B", - "author": "opengvlab", - "description": "The 2b version of the InternVL3 series, for an even higher inference speed and very reasonable performance. An advanced multimodal large language model (MLLM) series that demonstrates superior overall performance. Compared to InternVL 2.5, InternVL3 exhibits superior multimodal perception and reasoning capabilities, while further extending its multimodal capabilities to encompass tool usage, GUI agents, industrial image analysis, 3D vision perception, and more.", + "name": "Aetherwiing: Starcannon 12B", + "shortName": "Starcannon 12B", + "author": "aetherwiing", + "description": "Starcannon 12B v2 is a creative roleplay and story writing model, based on Mistral Nemo, using [nothingiisreal/mn-celeste-12b](/nothingiisreal/mn-celeste-12b) as a base, with [intervitens/mini-magnum-12b-v1.1](https://huggingface.co/intervitens/mini-magnum-12b-v1.1) merged in using the [TIES](https://arxiv.org/abs/2306.01708) method.\n\nAlthough more similar to Magnum overall, the model remains very creative, with a pleasant writing style. It is recommended for people wanting more variety than Magnum, and yet more verbose prose than Celeste.", "modelVersionGroupId": null, - "contextLength": 32000, + "contextLength": 12000, "inputModalities": [ - "image", "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": null, + "group": "Mistral", + "instructType": "chatml", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "opengvlab/internvl3-2b", + "permaslug": "aetherwiing/mn-starcannon-12b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "opengvlab/internvl3-2b:free", - "modelVariantPermaslug": "opengvlab/internvl3-2b:free", - "providerName": "Nineteen", + "modelVariantSlug": "aetherwiing/mn-starcannon-12b", + "modelVariantPermaslug": "aetherwiing/mn-starcannon-12b", + "providerName": "Featherless", "providerInfo": { - "name": "Nineteen", - "displayName": "Nineteen", - "slug": "nineteen", + "name": "Featherless", + "displayName": "Featherless", + "slug": "featherless", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://nineteen.ai/tos", + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false } }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": true, + "isAbortable": false, "moderationRequired": false, - "group": "Nineteen", + "group": "Featherless", "editors": [], "owners": [], - "isMultipartSupported": true, + "isMultipartSupported": false, "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nineteen.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" } }, - "providerDisplayName": "Nineteen", - "providerModelId": "OpenGVLab/InternVL3-2B", - "providerGroup": "Nineteen", - "quantization": "bf16", - "variant": "free", - "isFree": true, - "canAbort": true, + "providerDisplayName": "Featherless", + "providerModelId": "AuriAetherwiing/MN-12B-Starcannon-v2", + "providerGroup": "Featherless", + "quantization": "fp8", + "variant": "standard", + "isFree": false, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", - "top_p" + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "top_k", + "min_p", + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://nineteen.ai/tos", + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": false }, - "training": false + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.0000008", + "completion": "0.0000012", "image": "0", "request": "0", "webSearch": "0", @@ -72271,37 +75213,72 @@ export default { "isDisabled": false, "supportsToolParameters": false, "supportsReasoning": false, - "supportsMultipart": true, + "supportsMultipart": false, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 271, - "newest": 18, - "throughputHighToLow": 1, - "latencyLowToHigh": 150, - "pricingLowToHigh": 7, - "pricingHighToLow": 255 + "newest": 222, + "throughputHighToLow": 281, + "latencyLowToHigh": 285, + "pricingLowToHigh": 207, + "pricingHighToLow": 111 }, - "providers": [] + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [ + { + "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", + "slug": "featherless", + "quantization": "fp8", + "context": 16384, + "maxCompletionTokens": 4096, + "providerModelId": "AuriAetherwiing/MN-12B-Starcannon-v2", + "pricing": { + "prompt": "0.0000008", + "completion": "0.0000012", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "top_k", + "min_p", + "seed" + ], + "inputCost": 0.8, + "outputCost": 1.2, + "throughput": 19.115, + "latency": 3381 + } + ] }, { - "slug": "thudm/glm-4-9b", - "hfSlug": "THUDM/GLM-4-9B-0414", - "updatedAt": "2025-04-25T17:11:00.514702+00:00", - "createdAt": "2025-04-25T17:10:23.758806+00:00", + "slug": "thudm/glm-z1-9b", + "hfSlug": "thudm/glm-z1-9b-0414", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-04-25T17:12:20.695158+00:00", "hfUpdatedAt": null, - "name": "THUDM: GLM 4 9B (free)", - "shortName": "GLM 4 9B (free)", + "name": "THUDM: GLM Z1 9B (free)", + "shortName": "GLM Z1 9B (free)", "author": "thudm", - "description": "GLM-4-9B-0414 is a 9 billion parameter language model from the GLM-4 series developed by THUDM. Trained using the same reinforcement learning and alignment strategies as its larger 32B counterparts, GLM-4-9B-0414 achieves high performance relative to its size, making it suitable for resource-constrained deployments that still require robust language understanding and generation capabilities.", + "description": "GLM-Z1-9B-0414 is a 9B-parameter language model developed by THUDM as part of the GLM-4 family. It incorporates techniques originally applied to larger GLM-Z1 models, including extended reinforcement learning, pairwise ranking alignment, and training on reasoning-intensive tasks such as mathematics, code, and logic. Despite its smaller size, it demonstrates strong performance on general-purpose reasoning tasks and outperforms many open-source models in its weight class.", "modelVersionGroupId": null, "contextLength": 32000, "inputModalities": [ @@ -72312,29 +75289,40 @@ export default { ], "hasTextOutput": true, "group": "Other", - "instructType": null, + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "thudm/glm-4-9b-0414", - "reasoningConfig": null, - "features": {}, + "permaslug": "thudm/glm-z1-9b-0414", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + }, "endpoint": { - "id": "9998778a-6cd2-479c-8e48-cb349270eac6", - "name": "Novita | thudm/glm-4-9b-0414:free", + "id": "8ad6c714-038b-432e-8719-d2cb7feb7187", + "name": "Novita | thudm/glm-z1-9b-0414:free", "contextLength": 32000, "model": { - "slug": "thudm/glm-4-9b", - "hfSlug": "THUDM/GLM-4-9B-0414", - "updatedAt": "2025-04-25T17:11:00.514702+00:00", - "createdAt": "2025-04-25T17:10:23.758806+00:00", + "slug": "thudm/glm-z1-9b", + "hfSlug": "thudm/glm-z1-9b-0414", + "updatedAt": "2025-05-02T21:14:16.355933+00:00", + "createdAt": "2025-04-25T17:12:20.695158+00:00", "hfUpdatedAt": null, - "name": "THUDM: GLM 4 9B", - "shortName": "GLM 4 9B", + "name": "THUDM: GLM Z1 9B", + "shortName": "GLM Z1 9B", "author": "thudm", - "description": "GLM-4-9B-0414 is a 9 billion parameter language model from the GLM-4 series developed by THUDM. Trained using the same reinforcement learning and alignment strategies as its larger 32B counterparts, GLM-4-9B-0414 achieves high performance relative to its size, making it suitable for resource-constrained deployments that still require robust language understanding and generation capabilities.", + "description": "GLM-Z1-9B-0414 is a 9B-parameter language model developed by THUDM as part of the GLM-4 family. It incorporates techniques originally applied to larger GLM-Z1 models, including extended reinforcement learning, pairwise ranking alignment, and training on reasoning-intensive tasks such as mathematics, code, and logic. Despite its smaller size, it demonstrates strong performance on general-purpose reasoning tasks and outperforms many open-source models in its weight class.", "modelVersionGroupId": null, "contextLength": 32000, "inputModalities": [ @@ -72345,18 +75333,29 @@ export default { ], "hasTextOutput": true, "group": "Other", - "instructType": null, + "instructType": "deepseek-r1", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<｜User｜>", + "<｜end▁of▁sentence｜>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "thudm/glm-4-9b-0414", - "reasoningConfig": null, - "features": {} + "permaslug": "thudm/glm-z1-9b-0414", + "reasoningConfig": { + "startToken": "", + "endToken": "" + }, + "features": { + "reasoningConfig": { + "startToken": "", + "endToken": "" + } + } }, - "modelVariantSlug": "thudm/glm-4-9b:free", - "modelVariantPermaslug": "thudm/glm-4-9b-0414:free", + "modelVariantSlug": "thudm/glm-z1-9b:free", + "modelVariantPermaslug": "thudm/glm-z1-9b-0414:free", "providerName": "Novita", "providerInfo": { "name": "Novita", @@ -72388,7 +75387,7 @@ export default { } }, "providerDisplayName": "NovitaAI", - "providerModelId": "thudm/glm-4-9b-0414", + "providerModelId": "thudm/glm-z1-9b-0414", "providerGroup": "Novita", "quantization": null, "variant": "free", @@ -72402,6 +75401,8 @@ export default { "max_tokens", "temperature", "top_p", + "reasoning", + "include_reasoning", "stop", "frequency_penalty", "presence_penalty", @@ -72435,7 +75436,7 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": false, + "supportsReasoning": true, "supportsMultipart": true, "limitRpm": null, "limitRpd": null, @@ -72449,26 +75450,27 @@ export default { }, "sorting": { "topWeekly": 272, - "newest": 35, - "throughputHighToLow": 198, - "latencyLowToHigh": 157, - "pricingLowToHigh": 16, - "pricingHighToLow": 264 + "newest": 34, + "throughputHighToLow": 184, + "latencyLowToHigh": 177, + "pricingLowToHigh": 15, + "pricingHighToLow": 263 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { - "slug": "openai/gpt-4-turbo-preview", + "slug": "liquid/lfm-40b", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-01-25T00:00:00+00:00", + "createdAt": "2024-09-30T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4 Turbo Preview", - "shortName": "GPT-4 Turbo Preview", - "author": "openai", - "description": "The preview GPT-4 model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more. Training data: up to Dec 2023.\n\n**Note:** heavily rate limited by OpenAI while in preview.", + "name": "Liquid: LFM 40B MoE", + "shortName": "LFM 40B MoE", + "author": "liquid", + "description": "Liquid's 40.3B Mixture of Experts (MoE) model. Liquid Foundation Models (LFMs) are large neural networks built with computational units rooted in dynamic systems.\n\nLFMs are general-purpose AI models that can be used to model any kind of sequential data, including video, audio, text, time series, and signals.\n\nSee the [launch announcement](https://www.liquid.ai/liquid-foundation-models) for benchmarks and more info.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -72476,32 +75478,36 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, + "group": "Other", + "instructType": "chatml", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-4-turbo-preview", + "permaslug": "liquid/lfm-40b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "003933da-395a-48eb-86a3-0c4ec486d67f", - "name": "OpenAI | openai/gpt-4-turbo-preview", - "contextLength": 128000, + "id": "7901899b-4845-4089-979d-34a1ef065875", + "name": "Liquid | liquid/lfm-40b", + "contextLength": 32768, "model": { - "slug": "openai/gpt-4-turbo-preview", + "slug": "liquid/lfm-40b", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-01-25T00:00:00+00:00", + "createdAt": "2024-09-30T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4 Turbo Preview", - "shortName": "GPT-4 Turbo Preview", - "author": "openai", - "description": "The preview GPT-4 model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more. Training data: up to Dec 2023.\n\n**Note:** heavily rate limited by OpenAI while in preview.", + "name": "Liquid: LFM 40B MoE", + "shortName": "LFM 40B MoE", + "author": "liquid", + "description": "Liquid's 40.3B Mixture of Experts (MoE) model. Liquid Foundation Models (LFMs) are large neural networks built with computational units rooted in dynamic systems.\n\nLFMs are general-purpose AI models that can be used to model any kind of sequential data, including video, audio, text, time series, and signals.\n\nSee the [launch announcement](https://www.liquid.ai/liquid-foundation-models) for benchmarks and more info.", "modelVersionGroupId": null, - "contextLength": 128000, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -72509,67 +75515,64 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, + "group": "Other", + "instructType": "chatml", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-4-turbo-preview", + "permaslug": "liquid/lfm-40b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/gpt-4-turbo-preview", - "modelVariantPermaslug": "openai/gpt-4-turbo-preview", - "providerName": "OpenAI", + "modelVariantSlug": "liquid/lfm-40b", + "modelVariantPermaslug": "liquid/lfm-40b", + "providerName": "Liquid", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Liquid", + "displayName": "Liquid", + "slug": "liquid", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://www.liquid.ai/terms-conditions", + "privacyPolicyUrl": "https://www.liquid.ai/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "training": false + } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", + "moderationRequired": false, + "group": "Liquid", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.liquid.ai/&size=256", "invertRequired": true } }, - "providerDisplayName": "OpenAI", - "providerModelId": "gpt-4-turbo-preview", - "providerGroup": "OpenAI", - "quantization": "unknown", + "providerDisplayName": "Liquid", + "providerModelId": "lfm-40b", + "providerGroup": "Liquid", + "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -72577,31 +75580,23 @@ export default { "frequency_penalty", "presence_penalty", "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" + "top_k", + "min_p", + "repetition_penalty" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://www.liquid.ai/terms-conditions", + "privacyPolicyUrl": "https://www.liquid.ai/privacy-policy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": false }, "pricing": { - "prompt": "0.00001", - "completion": "0.00003", + "prompt": "0.00000015", + "completion": "0.00000015", "image": "0", "request": "0", "webSearch": "0", @@ -72612,7 +75607,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -72626,23 +75621,59 @@ export default { }, "sorting": { "topWeekly": 273, - "newest": 285, - "throughputHighToLow": 233, - "latencyLowToHigh": 167, - "pricingLowToHigh": 303, - "pricingHighToLow": 15 + "newest": 196, + "throughputHighToLow": 18, + "latencyLowToHigh": 71, + "pricingLowToHigh": 134, + "pricingHighToLow": 185 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://www.liquid.ai/\\u0026size=256", "providers": [ { - "name": "OpenAI", - "slug": "openAi", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 4096, - "providerModelId": "gpt-4-turbo-preview", + "name": "Liquid", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.liquid.ai/&size=256", + "slug": "liquid", + "quantization": null, + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "lfm-40b", "pricing": { - "prompt": "0.00001", - "completion": "0.00003", + "prompt": "0.00000015", + "completion": "0.00000015", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "top_k", + "min_p", + "repetition_penalty" + ], + "inputCost": 0.15, + "outputCost": 0.15, + "throughput": 184.404, + "latency": 507 + }, + { + "name": "Lambda", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://lambdalabs.com/&size=256", + "slug": "lambda", + "quantization": "bf16", + "context": 65536, + "maxCompletionTokens": 65536, + "providerModelId": "lfm-40b", + "pricing": { + "prompt": "0.00000015", + "completion": "0.00000015", "image": "0", "request": "0", "webSearch": "0", @@ -72650,8 +75681,6 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -72662,24 +75691,25 @@ export default { "logit_bias", "logprobs", "top_logprobs", - "response_format", - "structured_outputs" + "response_format" ], - "inputCost": 10, - "outputCost": 30 + "inputCost": 0.15, + "outputCost": 0.15, + "throughput": 182.443, + "latency": 266 } ] }, { - "slug": "01-ai/yi-large", - "hfSlug": null, + "slug": "nousresearch/nous-hermes-2-mixtral-8x7b-dpo", + "hfSlug": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-06-25T00:00:00+00:00", + "createdAt": "2024-01-16T00:00:00+00:00", "hfUpdatedAt": null, - "name": "01.AI: Yi Large", - "shortName": "Yi Large", - "author": "01-ai", - "description": "The Yi Large model was designed by 01.AI with the following usecases in mind: knowledge search, data classification, human-like chat bots, and customer service.\n\nIt stands out for its multilingual proficiency, particularly in Spanish, Chinese, Japanese, German, and French.\n\nCheck out the [launch announcement](https://01-ai.github.io/blog/01.ai-yi-large-llm-launch) to learn more.", + "name": "Nous: Hermes 2 Mixtral 8x7B DPO", + "shortName": "Hermes 2 Mixtral 8x7B DPO", + "author": "nousresearch", + "description": "Nous Hermes 2 Mixtral 8x7B DPO is the new flagship Nous Research model trained over the [Mixtral 8x7B MoE LLM](/models/mistralai/mixtral-8x7b).\n\nThe model was trained on over 1,000,000 entries of primarily [GPT-4](/models/openai/gpt-4) generated data, as well as other high quality data from open datasets across the AI landscape, achieving state of the art performance on a variety of tasks.\n\n#moe", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ @@ -72689,30 +75719,34 @@ export default { "text" ], "hasTextOutput": true, - "group": "Yi", - "instructType": null, + "group": "Mistral", + "instructType": "chatml", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "01-ai/yi-large", + "permaslug": "nousresearch/nous-hermes-2-mixtral-8x7b-dpo", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "5880ca91-521e-4755-b316-1d887d67d582", - "name": "Fireworks | 01-ai/yi-large", + "id": "5ec74ebb-b872-4d66-a93a-3d2696ecc664", + "name": "Together | nousresearch/nous-hermes-2-mixtral-8x7b-dpo", "contextLength": 32768, "model": { - "slug": "01-ai/yi-large", - "hfSlug": null, + "slug": "nousresearch/nous-hermes-2-mixtral-8x7b-dpo", + "hfSlug": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-06-25T00:00:00+00:00", + "createdAt": "2024-01-16T00:00:00+00:00", "hfUpdatedAt": null, - "name": "01.AI: Yi Large", - "shortName": "Yi Large", - "author": "01-ai", - "description": "The Yi Large model was designed by 01.AI with the following usecases in mind: knowledge search, data classification, human-like chat bots, and customer service.\n\nIt stands out for its multilingual proficiency, particularly in Spanish, Chinese, Japanese, German, and French.\n\nCheck out the [launch announcement](https://01-ai.github.io/blog/01.ai-yi-large-llm-launch) to learn more.", + "name": "Nous: Hermes 2 Mixtral 8x7B DPO", + "shortName": "Hermes 2 Mixtral 8x7B DPO", + "author": "nousresearch", + "description": "Nous Hermes 2 Mixtral 8x7B DPO is the new flagship Nous Research model trained over the [Mixtral 8x7B MoE LLM](/models/mistralai/mixtral-8x7b).\n\nThe model was trained on over 1,000,000 entries of primarily [GPT-4](/models/openai/gpt-4) generated data, as well as other high quality data from open datasets across the AI landscape, achieving state of the art performance on a variety of tasks.\n\n#moe", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ @@ -72722,29 +75756,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Yi", - "instructType": null, + "group": "Mistral", + "instructType": "chatml", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "01-ai/yi-large", + "permaslug": "nousresearch/nous-hermes-2-mixtral-8x7b-dpo", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "01-ai/yi-large", - "modelVariantPermaslug": "01-ai/yi-large", - "providerName": "Fireworks", + "modelVariantSlug": "nousresearch/nous-hermes-2-mixtral-8x7b-dpo", + "modelVariantPermaslug": "nousresearch/nous-hermes-2-mixtral-8x7b-dpo", + "providerName": "Together", "providerInfo": { - "name": "Fireworks", - "displayName": "Fireworks", - "slug": "fireworks", + "name": "Together", + "displayName": "Together", + "slug": "together", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://fireworks.ai/terms-of-service", - "privacyPolicyUrl": "https://fireworks.ai/privacy-policy", - "dataPolicyUrl": "https://docs.fireworks.ai/guides/security_compliance/data_handling", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -72755,26 +75792,26 @@ export default { "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Fireworks", + "group": "Together", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.fireworks.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Fireworks.png" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" } }, - "providerDisplayName": "Fireworks", - "providerModelId": "accounts/yi-01-ai/models/yi-large", - "providerGroup": "Fireworks", - "quantization": null, + "providerDisplayName": "Together", + "providerModelId": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO", + "providerGroup": "Together", + "quantization": "unknown", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": 2048, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -72786,18 +75823,15 @@ export default { "presence_penalty", "top_k", "repetition_penalty", - "response_format", - "structured_outputs", "logit_bias", - "logprobs", - "top_logprobs" + "min_p", + "response_format" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://fireworks.ai/terms-of-service", - "privacyPolicyUrl": "https://fireworks.ai/privacy-policy", - "dataPolicyUrl": "https://docs.fireworks.ai/guides/security_compliance/data_handling", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -72806,8 +75840,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.000003", - "completion": "0.000003", + "prompt": "0.0000006", + "completion": "0.0000006", "image": "0", "request": "0", "webSearch": "0", @@ -72832,23 +75866,25 @@ export default { }, "sorting": { "topWeekly": 274, - "newest": 242, - "throughputHighToLow": 113, - "latencyLowToHigh": 94, - "pricingLowToHigh": 257, - "pricingHighToLow": 62 + "newest": 287, + "throughputHighToLow": 59, + "latencyLowToHigh": 86, + "pricingLowToHigh": 192, + "pricingHighToLow": 126 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nousresearch.com/\\u0026size=256", "providers": [ { - "name": "Fireworks", - "slug": "fireworks", - "quantization": null, + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": "unknown", "context": 32768, - "maxCompletionTokens": 4096, - "providerModelId": "accounts/yi-01-ai/models/yi-large", + "maxCompletionTokens": 2048, + "providerModelId": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO", "pricing": { - "prompt": "0.000003", - "completion": "0.000003", + "prompt": "0.0000006", + "completion": "0.0000006", "image": "0", "request": "0", "webSearch": "0", @@ -72864,171 +75900,352 @@ export default { "presence_penalty", "top_k", "repetition_penalty", - "response_format", - "structured_outputs", "logit_bias", - "logprobs", - "top_logprobs" + "min_p", + "response_format" ], - "inputCost": 3, - "outputCost": 3 + "inputCost": 0.6, + "outputCost": 0.6, + "throughput": 121.58, + "latency": 593 } ] }, { - "slug": "aetherwiing/mn-starcannon-12b", - "hfSlug": "aetherwiing/MN-12B-Starcannon-v2", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-13T00:00:00+00:00", + "slug": "opengvlab/internvl3-2b", + "hfSlug": "OpenGVLab/InternVL3-2B", + "updatedAt": "2025-04-30T13:55:48.070004+00:00", + "createdAt": "2025-04-30T13:30:07.912688+00:00", + "hfUpdatedAt": null, + "name": "OpenGVLab: InternVL3 2B (free)", + "shortName": "InternVL3 2B (free)", + "author": "opengvlab", + "description": "The 2b version of the InternVL3 series, for an even higher inference speed and very reasonable performance. An advanced multimodal large language model (MLLM) series that demonstrates superior overall performance. Compared to InternVL 2.5, InternVL3 exhibits superior multimodal perception and reasoning capabilities, while further extending its multimodal capabilities to encompass tool usage, GUI agents, industrial image analysis, 3D vision perception, and more.", + "modelVersionGroupId": null, + "contextLength": 32000, + "inputModalities": [ + "image", + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Other", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "opengvlab/internvl3-2b", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "ec1cddcf-1619-457b-9008-bbadc41eb2bb", + "name": "Nineteen | opengvlab/internvl3-2b:free", + "contextLength": 32000, + "model": { + "slug": "opengvlab/internvl3-2b", + "hfSlug": "OpenGVLab/InternVL3-2B", + "updatedAt": "2025-04-30T13:55:48.070004+00:00", + "createdAt": "2025-04-30T13:30:07.912688+00:00", + "hfUpdatedAt": null, + "name": "OpenGVLab: InternVL3 2B", + "shortName": "InternVL3 2B", + "author": "opengvlab", + "description": "The 2b version of the InternVL3 series, for an even higher inference speed and very reasonable performance. An advanced multimodal large language model (MLLM) series that demonstrates superior overall performance. Compared to InternVL 2.5, InternVL3 exhibits superior multimodal perception and reasoning capabilities, while further extending its multimodal capabilities to encompass tool usage, GUI agents, industrial image analysis, 3D vision perception, and more.", + "modelVersionGroupId": null, + "contextLength": 32000, + "inputModalities": [ + "image", + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Other", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "opengvlab/internvl3-2b", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "opengvlab/internvl3-2b:free", + "modelVariantPermaslug": "opengvlab/internvl3-2b:free", + "providerName": "Nineteen", + "providerInfo": { + "name": "Nineteen", + "displayName": "Nineteen", + "slug": "nineteen", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://nineteen.ai/tos", + "paidModels": { + "training": false + } + }, + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "Nineteen", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://nineteen.ai/&size=256" + } + }, + "providerDisplayName": "Nineteen", + "providerModelId": "OpenGVLab/InternVL3-2B", + "providerGroup": "Nineteen", + "quantization": "bf16", + "variant": "free", + "isFree": true, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://nineteen.ai/tos", + "paidModels": { + "training": false + }, + "training": false + }, + "pricing": { + "prompt": "0", + "completion": "0", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 275, + "newest": 18, + "throughputHighToLow": 2, + "latencyLowToHigh": 109, + "pricingLowToHigh": 7, + "pricingHighToLow": 255 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [] + }, + { + "slug": "openai/gpt-4o", + "hfSlug": null, + "updatedAt": "2025-04-23T22:07:26.226912+00:00", + "createdAt": "2024-05-13T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Aetherwiing: Starcannon 12B", - "shortName": "Starcannon 12B", - "author": "aetherwiing", - "description": "Starcannon 12B v2 is a creative roleplay and story writing model, based on Mistral Nemo, using [nothingiisreal/mn-celeste-12b](/nothingiisreal/mn-celeste-12b) as a base, with [intervitens/mini-magnum-12b-v1.1](https://huggingface.co/intervitens/mini-magnum-12b-v1.1) merged in using the [TIES](https://arxiv.org/abs/2306.01708) method.\n\nAlthough more similar to Magnum overall, the model remains very creative, with a pleasant writing style. It is recommended for people wanting more variety than Magnum, and yet more verbose prose than Celeste.", - "modelVersionGroupId": null, - "contextLength": 16384, + "name": "OpenAI: GPT-4o (extended)", + "shortName": "GPT-4o (extended)", + "author": "openai", + "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)\n\n#multimodal", + "modelVersionGroupId": "76e36b33-358e-477a-be24-09f954fcea74", + "contextLength": 128000, "inputModalities": [ - "text" + "text", + "image", + "file" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "chatml", + "group": "GPT", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "aetherwiing/mn-starcannon-12b", + "permaslug": "openai/gpt-4o", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "1d498993-f126-4279-8bd7-66c2cfc88423", - "name": "Featherless | aetherwiing/mn-starcannon-12b", - "contextLength": 16384, + "id": "3f4c883a-bd8b-4e01-ac1b-25cc9a17dd61", + "name": "OpenAI | openai/gpt-4o:extended", + "contextLength": 128000, "model": { - "slug": "aetherwiing/mn-starcannon-12b", - "hfSlug": "aetherwiing/MN-12B-Starcannon-v2", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-13T00:00:00+00:00", + "slug": "openai/gpt-4o", + "hfSlug": null, + "updatedAt": "2025-04-23T22:07:26.226912+00:00", + "createdAt": "2024-05-13T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Aetherwiing: Starcannon 12B", - "shortName": "Starcannon 12B", - "author": "aetherwiing", - "description": "Starcannon 12B v2 is a creative roleplay and story writing model, based on Mistral Nemo, using [nothingiisreal/mn-celeste-12b](/nothingiisreal/mn-celeste-12b) as a base, with [intervitens/mini-magnum-12b-v1.1](https://huggingface.co/intervitens/mini-magnum-12b-v1.1) merged in using the [TIES](https://arxiv.org/abs/2306.01708) method.\n\nAlthough more similar to Magnum overall, the model remains very creative, with a pleasant writing style. It is recommended for people wanting more variety than Magnum, and yet more verbose prose than Celeste.", - "modelVersionGroupId": null, - "contextLength": 12000, + "name": "OpenAI: GPT-4o", + "shortName": "GPT-4o", + "author": "openai", + "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)\n\n#multimodal", + "modelVersionGroupId": "76e36b33-358e-477a-be24-09f954fcea74", + "contextLength": 128000, "inputModalities": [ - "text" + "text", + "image", + "file" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "chatml", + "group": "GPT", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "aetherwiing/mn-starcannon-12b", + "permaslug": "openai/gpt-4o", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "aetherwiing/mn-starcannon-12b", - "modelVariantPermaslug": "aetherwiing/mn-starcannon-12b", - "providerName": "Featherless", + "modelVariantSlug": "openai/gpt-4o:extended", + "modelVariantPermaslug": "openai/gpt-4o:extended", + "providerName": "OpenAI", "providerInfo": { - "name": "Featherless", - "displayName": "Featherless", - "slug": "featherless", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false - } + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, - "moderationRequired": false, - "group": "Featherless", + "isAbortable": true, + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], - "isMultipartSupported": false, - "statusPageUrl": null, + "isMultipartSupported": true, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Featherless", - "providerModelId": "AuriAetherwiing/MN-12B-Starcannon-v2", - "providerGroup": "Featherless", - "quantization": "fp8", - "variant": "standard", + "providerDisplayName": "OpenAI", + "providerModelId": "gpt-4o-64k-output-alpha", + "providerGroup": "OpenAI", + "quantization": "unknown", + "variant": "extended", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": 64000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "top_k", - "min_p", - "seed" + "web_search_options", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, + "requiresUserIds": true, "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.0000008", - "completion": "0.0000012", - "image": "0", + "prompt": "0.000006", + "completion": "0.000018", + "image": "0.007225", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, - "variablePricings": [], + "variablePricings": [ + { + "type": "search-threshold", + "threshold": "high", + "request": "0.05" + }, + { + "type": "search-threshold", + "threshold": "medium", + "request": "0.035" + }, + { + "type": "search-threshold", + "threshold": "low", + "request": "0.03" + } + ], "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, - "supportsMultipart": false, + "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, @@ -73039,44 +76256,93 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 275, - "newest": 222, - "throughputHighToLow": 280, - "latencyLowToHigh": 282, - "pricingLowToHigh": 208, - "pricingHighToLow": 110 + "topWeekly": 35, + "newest": 258, + "throughputHighToLow": 43, + "latencyLowToHigh": 110, + "pricingLowToHigh": 265, + "pricingHighToLow": 24 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "Featherless", - "slug": "featherless", - "quantization": "fp8", - "context": 16384, - "maxCompletionTokens": 4096, - "providerModelId": "AuriAetherwiing/MN-12B-Starcannon-v2", + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": "unknown", + "context": 128000, + "maxCompletionTokens": 16384, + "providerModelId": "gpt-4o", "pricing": { - "prompt": "0.0000008", - "completion": "0.0000012", - "image": "0", + "prompt": "0.0000025", + "completion": "0.00001", + "image": "0.003613", "request": "0", + "inputCacheRead": "0.00000125", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "top_k", - "min_p", - "seed" + "web_search_options", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" ], - "inputCost": 0.8, - "outputCost": 1.2 + "inputCost": 2.5, + "outputCost": 10, + "throughput": 49.1885, + "latency": 672 + }, + { + "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", + "slug": "azure", + "quantization": "unknown", + "context": 128000, + "maxCompletionTokens": 16384, + "providerModelId": "gpt-4o", + "pricing": { + "prompt": "0.0000025", + "completion": "0.00001", + "image": "0.003613", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "web_search_options", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" + ], + "inputCost": 2.5, + "outputCost": 10, + "throughput": 111.027, + "latency": 2185 } ] }, @@ -73258,16 +76524,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 180, + "topWeekly": 177, "newest": 162, - "throughputHighToLow": 246, - "latencyLowToHigh": 46, + "throughputHighToLow": 255, + "latencyLowToHigh": 284, "pricingLowToHigh": 57, "pricingHighToLow": 200 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [ { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 32768, @@ -73296,10 +76564,13 @@ export default { "top_logprobs" ], "inputCost": 0.09, - "outputCost": 0.27 + "outputCost": 0.27, + "throughput": 29.426, + "latency": 1264.5 }, { "name": "Hyperbolic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://hyperbolic.xyz/&size=256", "slug": "hyperbolic", "quantization": "fp8", "context": 32768, @@ -73330,187 +76601,372 @@ export default { "repetition_penalty" ], "inputCost": 0.2, - "outputCost": 0.2 + "outputCost": 0.2, + "throughput": 24.9545, + "latency": 21851.5 } ] }, { - "slug": "openai/gpt-4o", + "slug": "01-ai/yi-large", "hfSlug": null, - "updatedAt": "2025-04-23T22:07:26.226912+00:00", - "createdAt": "2024-05-13T00:00:00+00:00", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-06-25T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4o (extended)", - "shortName": "GPT-4o (extended)", - "author": "openai", - "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)\n\n#multimodal", - "modelVersionGroupId": "76e36b33-358e-477a-be24-09f954fcea74", - "contextLength": 128000, + "name": "01.AI: Yi Large", + "shortName": "Yi Large", + "author": "01-ai", + "description": "The Yi Large model was designed by 01.AI with the following usecases in mind: knowledge search, data classification, human-like chat bots, and customer service.\n\nIt stands out for its multilingual proficiency, particularly in Spanish, Chinese, Japanese, German, and French.\n\nCheck out the [launch announcement](https://01-ai.github.io/blog/01.ai-yi-large-llm-launch) to learn more.", + "modelVersionGroupId": null, + "contextLength": 32768, "inputModalities": [ - "text", - "image", - "file" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "GPT", + "group": "Yi", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-4o", + "permaslug": "01-ai/yi-large", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "3f4c883a-bd8b-4e01-ac1b-25cc9a17dd61", - "name": "OpenAI | openai/gpt-4o:extended", - "contextLength": 128000, + "id": "5880ca91-521e-4755-b316-1d887d67d582", + "name": "Fireworks | 01-ai/yi-large", + "contextLength": 32768, "model": { - "slug": "openai/gpt-4o", + "slug": "01-ai/yi-large", "hfSlug": null, - "updatedAt": "2025-04-23T22:07:26.226912+00:00", - "createdAt": "2024-05-13T00:00:00+00:00", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-06-25T00:00:00+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4o", - "shortName": "GPT-4o", - "author": "openai", - "description": "GPT-4o (\"o\" for \"omni\") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of [GPT-4 Turbo](/models/openai/gpt-4-turbo) while being twice as fast and 50% more cost-effective. GPT-4o also offers improved performance in processing non-English languages and enhanced visual capabilities.\n\nFor benchmarking against other models, it was briefly called [\"im-also-a-good-gpt2-chatbot\"](https://twitter.com/LiamFedus/status/1790064963966370209)\n\n#multimodal", - "modelVersionGroupId": "76e36b33-358e-477a-be24-09f954fcea74", - "contextLength": 128000, + "name": "01.AI: Yi Large", + "shortName": "Yi Large", + "author": "01-ai", + "description": "The Yi Large model was designed by 01.AI with the following usecases in mind: knowledge search, data classification, human-like chat bots, and customer service.\n\nIt stands out for its multilingual proficiency, particularly in Spanish, Chinese, Japanese, German, and French.\n\nCheck out the [launch announcement](https://01-ai.github.io/blog/01.ai-yi-large-llm-launch) to learn more.", + "modelVersionGroupId": null, + "contextLength": 32768, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Yi", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "01-ai/yi-large", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "01-ai/yi-large", + "modelVariantPermaslug": "01-ai/yi-large", + "providerName": "Fireworks", + "providerInfo": { + "name": "Fireworks", + "displayName": "Fireworks", + "slug": "fireworks", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://fireworks.ai/terms-of-service", + "privacyPolicyUrl": "https://fireworks.ai/privacy-policy", + "dataPolicyUrl": "https://docs.fireworks.ai/guides/security_compliance/data_handling", + "paidModels": { + "training": false, + "retainsPrompts": false + } + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "Fireworks", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": "https://status.fireworks.ai/", + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/Fireworks.png" + } + }, + "providerDisplayName": "Fireworks", + "providerModelId": "accounts/yi-01-ai/models/yi-large", + "providerGroup": "Fireworks", + "quantization": null, + "variant": "standard", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": 4096, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "response_format", + "structured_outputs", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://fireworks.ai/terms-of-service", + "privacyPolicyUrl": "https://fireworks.ai/privacy-policy", + "dataPolicyUrl": "https://docs.fireworks.ai/guides/security_compliance/data_handling", + "paidModels": { + "training": false, + "retainsPrompts": false + }, + "training": false, + "retainsPrompts": false + }, + "pricing": { + "prompt": "0.000003", + "completion": "0.000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 278, + "newest": 242, + "throughputHighToLow": 105, + "latencyLowToHigh": 94, + "pricingLowToHigh": 257, + "pricingHighToLow": 62 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [ + { + "name": "Fireworks", + "icon": "https://openrouter.ai/images/icons/Fireworks.png", + "slug": "fireworks", + "quantization": null, + "context": 32768, + "maxCompletionTokens": 4096, + "providerModelId": "accounts/yi-01-ai/models/yi-large", + "pricing": { + "prompt": "0.000003", + "completion": "0.000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "top_k", + "repetition_penalty", + "response_format", + "structured_outputs", + "logit_bias", + "logprobs", + "top_logprobs" + ], + "inputCost": 3, + "outputCost": 3, + "throughput": 75.5795, + "latency": 616.5 + } + ] + }, + { + "slug": "pygmalionai/mythalion-13b", + "hfSlug": "PygmalionAI/mythalion-13b", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2023-09-02T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Pygmalion: Mythalion 13B", + "shortName": "Mythalion 13B", + "author": "pygmalionai", + "description": "A blend of the new Pygmalion-13b and MythoMax. #merge", + "modelVersionGroupId": null, + "contextLength": 8192, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Llama2", + "instructType": "alpaca", + "defaultSystem": null, + "defaultStops": [ + "###", + "" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "pygmalionai/mythalion-13b", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "81260a05-530c-42d2-b654-2e149377cbfd", + "name": "Mancer | pygmalionai/mythalion-13b", + "contextLength": 8192, + "model": { + "slug": "pygmalionai/mythalion-13b", + "hfSlug": "PygmalionAI/mythalion-13b", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2023-09-02T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Pygmalion: Mythalion 13B", + "shortName": "Mythalion 13B", + "author": "pygmalionai", + "description": "A blend of the new Pygmalion-13b and MythoMax. #merge", + "modelVersionGroupId": null, + "contextLength": 8192, "inputModalities": [ - "text", - "image", - "file" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, + "group": "Llama2", + "instructType": "alpaca", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "###", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-4o", + "permaslug": "pygmalionai/mythalion-13b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/gpt-4o:extended", - "modelVariantPermaslug": "openai/gpt-4o:extended", - "providerName": "OpenAI", + "modelVariantSlug": "pygmalionai/mythalion-13b", + "modelVariantPermaslug": "pygmalionai/mythalion-13b", + "providerName": "Mancer", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Mancer", + "displayName": "Mancer", + "slug": "mancer", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://mancer.tech/terms", + "privacyPolicyUrl": "https://mancer.tech/privacy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "training": true, + "retainsPrompts": true + } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", + "moderationRequired": false, + "group": "Mancer", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256" } }, - "providerDisplayName": "OpenAI", - "providerModelId": "gpt-4o-64k-output-alpha", - "providerGroup": "OpenAI", + "providerDisplayName": "Mancer", + "providerModelId": "mythalion", + "providerGroup": "Mancer", "quantization": "unknown", - "variant": "extended", + "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 64000, + "maxCompletionTokens": 1024, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "web_search_options", - "seed", + "repetition_penalty", "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" + "top_k", + "min_p", + "seed", + "top_a" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://mancer.tech/terms", + "privacyPolicyUrl": "https://mancer.tech/privacy", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.000006", - "completion": "0.000018", - "image": "0.007225", + "prompt": "0.0000005625", + "completion": "0.000001125", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", - "discount": 0 + "discount": 0.25 }, - "variablePricings": [ - { - "type": "search-threshold", - "threshold": "high", - "request": "0.05" - }, - { - "type": "search-threshold", - "threshold": "medium", - "request": "0.035" - }, - { - "type": "search-threshold", - "threshold": "low", - "request": "0.03" - } - ], + "variablePricings": [], "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -73523,86 +76979,120 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 34, - "newest": 258, - "throughputHighToLow": 44, - "latencyLowToHigh": 107, - "pricingLowToHigh": 265, - "pricingHighToLow": 23 + "topWeekly": 279, + "newest": 305, + "throughputHighToLow": 266, + "latencyLowToHigh": 158, + "pricingLowToHigh": 193, + "pricingHighToLow": 125 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { - "name": "OpenAI", - "slug": "openAi", + "name": "Mancer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", + "slug": "mancer", "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 16384, - "providerModelId": "gpt-4o", + "context": 8192, + "maxCompletionTokens": 1024, + "providerModelId": "mythalion", "pricing": { - "prompt": "0.0000025", - "completion": "0.00001", - "image": "0.003613", + "prompt": "0.0000005625", + "completion": "0.000001125", + "image": "0", "request": "0", - "inputCacheRead": "0.00000125", "webSearch": "0", "internalReasoning": "0", - "discount": 0 + "discount": 0.25 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "web_search_options", - "seed", + "repetition_penalty", "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" + "top_k", + "min_p", + "seed", + "top_a" ], - "inputCost": 2.5, - "outputCost": 10 + "inputCost": 0.56, + "outputCost": 1.13, + "throughput": 23.275, + "latency": 948 }, { - "name": "Azure", - "slug": "azure", + "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", + "slug": "featherless", + "quantization": "fp8", + "context": 4096, + "maxCompletionTokens": 4096, + "providerModelId": "PygmalionAI/mythalion-13b", + "pricing": { + "prompt": "0.0000008", + "completion": "0.0000012", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "top_k", + "min_p", + "seed" + ], + "inputCost": 0.8, + "outputCost": 1.2, + "throughput": 16.256, + "latency": 1470 + }, + { + "name": "Mancer (private)", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", + "slug": "mancer (private)", "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 16384, - "providerModelId": "gpt-4o", + "context": 8192, + "maxCompletionTokens": 1024, + "providerModelId": "mythalion", "pricing": { - "prompt": "0.0000025", - "completion": "0.00001", - "image": "0.003613", + "prompt": "0.000001", + "completion": "0.0000015", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "web_search_options", - "seed", + "repetition_penalty", "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" + "top_k", + "min_p", + "seed", + "top_a" ], - "inputCost": 2.5, - "outputCost": 10 + "inputCost": 1, + "outputCost": 1.5, + "throughput": 22.1565, + "latency": 668 } ] }, @@ -73763,229 +77253,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 278, + "topWeekly": 280, "newest": 77, - "throughputHighToLow": 279, - "latencyLowToHigh": 255, + "throughputHighToLow": 280, + "latencyLowToHigh": 269, "pricingLowToHigh": 34, "pricingHighToLow": 282 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://featherless.ai/\\u0026size=256", "providers": [] }, { - "slug": "arcee-ai/caller-large", - "hfSlug": "", - "updatedAt": "2025-05-05T23:32:30.837689+00:00", - "createdAt": "2025-05-05T23:31:09.181141+00:00", - "hfUpdatedAt": null, - "name": "Arcee AI: Caller Large", - "shortName": "Caller Large", - "author": "arcee-ai", - "description": "Caller Large is Arcee's specialist \"function‑calling\" SLM built to orchestrate external tools and APIs. Instead of maximizing next‑token accuracy, training focuses on structured JSON outputs, parameter extraction and multi‑step tool chains, making Caller a natural choice for retrieval‑augmented generation, robotic process automation or data‑pull chatbots. It incorporates a routing head that decides when (and how) to invoke a tool versus answering directly, reducing hallucinated calls. The model is already the backbone of Arcee Conductor's auto‑tool mode, where it parses user intent, emits clean function signatures and hands control back once the tool response is ready. Developers thus gain an OpenAI‑style function‑calling UX without handing requests to a frontier‑scale model. ", - "modelVersionGroupId": null, - "contextLength": 32768, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "arcee-ai/caller-large", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "587c1cf8-5f90-42f7-9657-4d83c51a1bfb", - "name": "Together | arcee-ai/caller-large", - "contextLength": 32768, - "model": { - "slug": "arcee-ai/caller-large", - "hfSlug": "", - "updatedAt": "2025-05-05T23:32:30.837689+00:00", - "createdAt": "2025-05-05T23:31:09.181141+00:00", - "hfUpdatedAt": null, - "name": "Arcee AI: Caller Large", - "shortName": "Caller Large", - "author": "arcee-ai", - "description": "Caller Large is Arcee's specialist \"function‑calling\" SLM built to orchestrate external tools and APIs. Instead of maximizing next‑token accuracy, training focuses on structured JSON outputs, parameter extraction and multi‑step tool chains, making Caller a natural choice for retrieval‑augmented generation, robotic process automation or data‑pull chatbots. It incorporates a routing head that decides when (and how) to invoke a tool versus answering directly, reducing hallucinated calls. The model is already the backbone of Arcee Conductor's auto‑tool mode, where it parses user intent, emits clean function signatures and hands control back once the tool response is ready. Developers thus gain an OpenAI‑style function‑calling UX without handing requests to a frontier‑scale model. ", - "modelVersionGroupId": null, - "contextLength": 32768, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "arcee-ai/caller-large", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "arcee-ai/caller-large", - "modelVariantPermaslug": "arcee-ai/caller-large", - "providerName": "Together", - "providerInfo": { - "name": "Together", - "displayName": "Together", - "slug": "together", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", - "paidModels": { - "training": false, - "retainsPrompts": false - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": false, - "group": "Together", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" - } - }, - "providerDisplayName": "Together", - "providerModelId": "arcee-ai/caller", - "providerGroup": "Together", - "quantization": null, - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": null, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", - "paidModels": { - "training": false, - "retainsPrompts": false - }, - "training": false, - "retainsPrompts": false - }, - "pricing": { - "prompt": "0.00000055", - "completion": "0.00000085", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 279, - "newest": 3, - "throughputHighToLow": 86, - "latencyLowToHigh": 65, - "pricingLowToHigh": 186, - "pricingHighToLow": 132 - }, - "providers": [ - { - "name": "Together", - "slug": "together", - "quantization": null, - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "arcee-ai/caller", - "pricing": { - "prompt": "0.00000055", - "completion": "0.00000085", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" - ], - "inputCost": 0.55, - "outputCost": 0.85 - } - ] - }, - { - "slug": "pygmalionai/mythalion-13b", - "hfSlug": "PygmalionAI/mythalion-13b", + "slug": "neversleep/llama-3-lumimaid-70b", + "hfSlug": "NeverSleep/Llama-3-Lumimaid-70B-v0.1", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-09-02T00:00:00+00:00", + "createdAt": "2024-05-16T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Pygmalion: Mythalion 13B", - "shortName": "Mythalion 13B", - "author": "pygmalionai", - "description": "A blend of the new Pygmalion-13b and MythoMax. #merge", + "name": "NeverSleep: Llama 3 Lumimaid 70B", + "shortName": "Llama 3 Lumimaid 70B", + "author": "neversleep", + "description": "The NeverSleep team is back, with a Llama 3 70B finetune trained on their curated roleplay data. Striking a balance between eRP and RP, Lumimaid was designed to be serious, yet uncensored when necessary.\n\nTo enhance it's overall intelligence and chat capability, roughly 40% of the training data was not roleplay. This provides a breadth of knowledge to access, while still keeping roleplay as the primary strength.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", "modelVersionGroupId": null, "contextLength": 8192, "inputModalities": [ @@ -73995,33 +77282,33 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama2", - "instructType": "alpaca", + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, "defaultStops": [ - "###", - "" + "<|eot_id|>", + "<|end_of_text|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "pygmalionai/mythalion-13b", + "permaslug": "neversleep/llama-3-lumimaid-70b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "81260a05-530c-42d2-b654-2e149377cbfd", - "name": "Mancer | pygmalionai/mythalion-13b", + "id": "9dec8b8b-2d23-49bf-ba8f-1116b6195c03", + "name": "Featherless | neversleep/llama-3-lumimaid-70b", "contextLength": 8192, "model": { - "slug": "pygmalionai/mythalion-13b", - "hfSlug": "PygmalionAI/mythalion-13b", + "slug": "neversleep/llama-3-lumimaid-70b", + "hfSlug": "NeverSleep/Llama-3-Lumimaid-70B-v0.1", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-09-02T00:00:00+00:00", + "createdAt": "2024-05-16T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Pygmalion: Mythalion 13B", - "shortName": "Mythalion 13B", - "author": "pygmalionai", - "description": "A blend of the new Pygmalion-13b and MythoMax. #merge", + "name": "NeverSleep: Llama 3 Lumimaid 70B", + "shortName": "Llama 3 Lumimaid 70B", + "author": "neversleep", + "description": "The NeverSleep team is back, with a Llama 3 70B finetune trained on their curated roleplay data. Striking a balance between eRP and RP, Lumimaid was designed to be serious, yet uncensored when necessary.\n\nTo enhance it's overall intelligence and chat capability, roughly 40% of the training data was not roleplay. This provides a breadth of knowledge to access, while still keeping roleplay as the primary strength.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", "modelVersionGroupId": null, "contextLength": 8192, "inputModalities": [ @@ -74031,60 +77318,61 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama2", - "instructType": "alpaca", + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, "defaultStops": [ - "###", - "" + "<|eot_id|>", + "<|end_of_text|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "pygmalionai/mythalion-13b", + "permaslug": "neversleep/llama-3-lumimaid-70b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "pygmalionai/mythalion-13b", - "modelVariantPermaslug": "pygmalionai/mythalion-13b", - "providerName": "Mancer", + "modelVariantSlug": "neversleep/llama-3-lumimaid-70b", + "modelVariantPermaslug": "neversleep/llama-3-lumimaid-70b", + "providerName": "Featherless", "providerInfo": { - "name": "Mancer", - "displayName": "Mancer", - "slug": "mancer", + "name": "Featherless", + "displayName": "Featherless", + "slug": "featherless", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://mancer.tech/terms", - "privacyPolicyUrl": "https://mancer.tech/privacy", + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false } }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": true, + "isAbortable": false, "moderationRequired": false, - "group": "Mancer", + "group": "Featherless", "editors": [], "owners": [], - "isMultipartSupported": true, + "isMultipartSupported": false, "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" } }, - "providerDisplayName": "Mancer", - "providerModelId": "mythalion", - "providerGroup": "Mancer", - "quantization": "unknown", + "providerDisplayName": "Featherless", + "providerModelId": "NeverSleep/Llama-3-Lumimaid-70B-v0.1", + "providerGroup": "Featherless", + "quantization": "fp8", "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 1024, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -74095,32 +77383,30 @@ export default { "frequency_penalty", "presence_penalty", "repetition_penalty", - "logit_bias", "top_k", "min_p", - "seed", - "top_a" + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://mancer.tech/terms", - "privacyPolicyUrl": "https://mancer.tech/privacy", + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": false }, "pricing": { - "prompt": "0.0000005625", - "completion": "0.000001125", + "prompt": "0.000004", + "completion": "0.000006", "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", - "discount": 0.25 + "discount": 0 }, "variablePricings": [], "isHidden": false, @@ -74128,7 +77414,7 @@ export default { "isDisabled": false, "supportsToolParameters": false, "supportsReasoning": false, - "supportsMultipart": true, + "supportsMultipart": false, "limitRpm": null, "limitRpd": null, "hasCompletions": true, @@ -74139,90 +77425,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 280, - "newest": 305, - "throughputHighToLow": 266, - "latencyLowToHigh": 148, - "pricingLowToHigh": 192, - "pricingHighToLow": 125 + "topWeekly": 281, + "newest": 255, + "throughputHighToLow": 299, + "latencyLowToHigh": 206, + "pricingLowToHigh": 283, + "pricingHighToLow": 38 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ - { - "name": "Mancer", - "slug": "mancer", - "quantization": "unknown", - "context": 8192, - "maxCompletionTokens": 1024, - "providerModelId": "mythalion", - "pricing": { - "prompt": "0.0000005625", - "completion": "0.000001125", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0.25 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "logit_bias", - "top_k", - "min_p", - "seed", - "top_a" - ], - "inputCost": 0.56, - "outputCost": 1.13 - }, - { - "name": "Mancer (private)", - "slug": "mancer (private)", - "quantization": "unknown", - "context": 8192, - "maxCompletionTokens": 1024, - "providerModelId": "mythalion", - "pricing": { - "prompt": "0.00000075", - "completion": "0.0000015", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "logit_bias", - "top_k", - "min_p", - "seed", - "top_a" - ], - "inputCost": 0.75, - "outputCost": 1.5 - }, { "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", "slug": "featherless", "quantization": "fp8", - "context": 4096, + "context": 8192, "maxCompletionTokens": 4096, - "providerModelId": "PygmalionAI/mythalion-13b", + "providerModelId": "NeverSleep/Llama-3-Lumimaid-70B-v0.1", "pricing": { - "prompt": "0.0000008", - "completion": "0.0000012", + "prompt": "0.000004", + "completion": "0.000006", "image": "0", "request": "0", "webSearch": "0", @@ -74241,23 +77463,25 @@ export default { "min_p", "seed" ], - "inputCost": 0.8, - "outputCost": 1.2 + "inputCost": 4, + "outputCost": 6, + "throughput": 13.779, + "latency": 1396 } ] }, { - "slug": "ai21/jamba-instruct", + "slug": "anthropic/claude-2", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-06-25T00:00:00+00:00", + "createdAt": "2023-11-22T00:00:00+00:00", "hfUpdatedAt": null, - "name": "AI21: Jamba Instruct", - "shortName": "Jamba Instruct", - "author": "ai21", - "description": "The Jamba-Instruct model, introduced by AI21 Labs, is an instruction-tuned variant of their hybrid SSM-Transformer Jamba model, specifically optimized for enterprise applications.\n\n- 256K Context Window: It can process extensive information, equivalent to a 400-page novel, which is beneficial for tasks involving large documents such as financial reports or legal documents\n- Safety and Accuracy: Jamba-Instruct is designed with enhanced safety features to ensure secure deployment in enterprise environments, reducing the risk and cost of implementation\n\nRead their [announcement](https://www.ai21.com/blog/announcing-jamba) to learn more.\n\nJamba has a knowledge cutoff of February 2024.", + "name": "Anthropic: Claude v2", + "shortName": "Claude v2", + "author": "anthropic", + "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.", "modelVersionGroupId": null, - "contextLength": 256000, + "contextLength": 200000, "inputModalities": [ "text" ], @@ -74265,32 +77489,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "Claude", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "ai21/jamba-instruct", + "permaslug": "anthropic/claude-2", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "4bdfe9f1-ba9f-4da1-ae1f-904bc6a7cd84", - "name": "AI21 | ai21/jamba-instruct", - "contextLength": 256000, + "id": "258ddc79-e520-4e68-8069-ea6771698c5a", + "name": "Anthropic | anthropic/claude-2", + "contextLength": 200000, "model": { - "slug": "ai21/jamba-instruct", + "slug": "anthropic/claude-2", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-06-25T00:00:00+00:00", + "createdAt": "2023-11-22T00:00:00+00:00", "hfUpdatedAt": null, - "name": "AI21: Jamba Instruct", - "shortName": "Jamba Instruct", - "author": "ai21", - "description": "The Jamba-Instruct model, introduced by AI21 Labs, is an instruction-tuned variant of their hybrid SSM-Transformer Jamba model, specifically optimized for enterprise applications.\n\n- 256K Context Window: It can process extensive information, equivalent to a 400-page novel, which is beneficial for tasks involving large documents such as financial reports or legal documents\n- Safety and Accuracy: Jamba-Instruct is designed with enhanced safety features to ensure secure deployment in enterprise environments, reducing the risk and cost of implementation\n\nRead their [announcement](https://www.ai21.com/blog/announcing-jamba) to learn more.\n\nJamba has a knowledge cutoff of February 2024.", + "name": "Anthropic: Claude v2", + "shortName": "Claude v2", + "author": "anthropic", + "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.", "modelVersionGroupId": null, - "contextLength": 256000, + "contextLength": 200000, "inputModalities": [ "text" ], @@ -74298,55 +77522,58 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "Claude", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "ai21/jamba-instruct", + "permaslug": "anthropic/claude-2", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "ai21/jamba-instruct", - "modelVariantPermaslug": "ai21/jamba-instruct", - "providerName": "AI21", + "modelVariantSlug": "anthropic/claude-2", + "modelVariantPermaslug": "anthropic/claude-2", + "providerName": "Anthropic", "providerInfo": { - "name": "AI21", - "displayName": "AI21", - "slug": "ai21", + "name": "Anthropic", + "displayName": "Anthropic", + "slug": "anthropic", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://www.ai21.com/privacy-policy/", - "termsOfServiceUrl": "https://www.ai21.com/terms-of-service/", + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", "paidModels": { - "training": false - } + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, - "headquarters": "IL", + "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, - "moderationRequired": false, - "group": "AI21", + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": true, + "group": "Anthropic", "editors": [], "owners": [], - "isMultipartSupported": false, - "statusPageUrl": "https://status.ai21.com/", + "isMultipartSupported": true, + "statusPageUrl": "https://status.anthropic.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://ai21.com/&size=256" + "url": "/images/icons/Anthropic.svg" } }, - "providerDisplayName": "AI21", - "providerModelId": "jamba-instruct", - "providerGroup": "AI21", + "providerDisplayName": "Anthropic", + "providerModelId": "claude-2.1", + "providerGroup": "Anthropic", "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": 4096, "maxPromptImages": null, @@ -74355,21 +77582,27 @@ export default { "max_tokens", "temperature", "top_p", + "top_k", "stop" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "privacyPolicyUrl": "https://www.ai21.com/privacy-policy/", - "termsOfServiceUrl": "https://www.ai21.com/terms-of-service/", + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", "paidModels": { - "training": false + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "training": false + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.0000005", - "completion": "0.0000007", + "prompt": "0.000008", + "completion": "0.000024", "image": "0", "request": "0", "webSearch": "0", @@ -74382,10 +77615,10 @@ export default { "isDisabled": false, "supportsToolParameters": false, "supportsReasoning": false, - "supportsMultipart": false, + "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { "supportsDocumentUrl": null @@ -74393,24 +77626,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 281, - "newest": 243, - "throughputHighToLow": 311, - "latencyLowToHigh": 318, - "pricingLowToHigh": 181, - "pricingHighToLow": 137 + "topWeekly": 282, + "newest": 296, + "throughputHighToLow": 291, + "latencyLowToHigh": 202, + "pricingLowToHigh": 296, + "pricingHighToLow": 19 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { - "name": "AI21", - "slug": "ai21", + "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", + "slug": "anthropic", "quantization": "unknown", - "context": 256000, + "context": 200000, "maxCompletionTokens": 4096, - "providerModelId": "jamba-instruct", + "providerModelId": "claude-2.1", "pricing": { - "prompt": "0.0000005", - "completion": "0.0000007", + "prompt": "0.000008", + "completion": "0.000024", "image": "0", "request": "0", "webSearch": "0", @@ -74421,25 +77656,28 @@ export default { "max_tokens", "temperature", "top_p", + "top_k", "stop" ], - "inputCost": 0.5, - "outputCost": 0.7 + "inputCost": 8, + "outputCost": 24, + "throughput": 16.037, + "latency": 1567 } ] }, { - "slug": "neversleep/llama-3-lumimaid-70b", - "hfSlug": "NeverSleep/Llama-3-Lumimaid-70B-v0.1", + "slug": "openai/gpt-3.5-turbo-16k", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-05-16T00:00:00+00:00", + "createdAt": "2023-08-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "NeverSleep: Llama 3 Lumimaid 70B", - "shortName": "Llama 3 Lumimaid 70B", - "author": "neversleep", - "description": "The NeverSleep team is back, with a Llama 3 70B finetune trained on their curated roleplay data. Striking a balance between eRP and RP, Lumimaid was designed to be serious, yet uncensored when necessary.\n\nTo enhance it's overall intelligence and chat capability, roughly 40% of the training data was not roleplay. This provides a breadth of knowledge to access, while still keeping roleplay as the primary strength.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "name": "OpenAI: GPT-3.5 Turbo 16k", + "shortName": "GPT-3.5 Turbo 16k", + "author": "openai", + "description": "This model offers four times the context length of gpt-3.5-turbo, allowing it to support approximately 20 pages of text in a single request at a higher cost. Training data: up to Sep 2021.", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 16385, "inputModalities": [ "text" ], @@ -74447,35 +77685,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "GPT", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "neversleep/llama-3-lumimaid-70b", + "permaslug": "openai/gpt-3.5-turbo-16k", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "9dec8b8b-2d23-49bf-ba8f-1116b6195c03", - "name": "Featherless | neversleep/llama-3-lumimaid-70b", - "contextLength": 8192, + "id": "eb1a93d0-a295-4afb-86d3-e2d10538c12d", + "name": "OpenAI | openai/gpt-3.5-turbo-16k", + "contextLength": 16385, "model": { - "slug": "neversleep/llama-3-lumimaid-70b", - "hfSlug": "NeverSleep/Llama-3-Lumimaid-70B-v0.1", + "slug": "openai/gpt-3.5-turbo-16k", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-05-16T00:00:00+00:00", + "createdAt": "2023-08-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "NeverSleep: Llama 3 Lumimaid 70B", - "shortName": "Llama 3 Lumimaid 70B", - "author": "neversleep", - "description": "The NeverSleep team is back, with a Llama 3 70B finetune trained on their curated roleplay data. Striking a balance between eRP and RP, Lumimaid was designed to be serious, yet uncensored when necessary.\n\nTo enhance it's overall intelligence and chat capability, roughly 40% of the training data was not roleplay. This provides a breadth of knowledge to access, while still keeping roleplay as the primary strength.\n\nUsage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "name": "OpenAI: GPT-3.5 Turbo 16k", + "shortName": "GPT-3.5 Turbo 16k", + "author": "openai", + "description": "This model offers four times the context length of gpt-3.5-turbo, allowing it to support approximately 20 pages of text in a single request at a higher cost. Training data: up to Sep 2021.", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 16385, "inputModalities": [ "text" ], @@ -74483,90 +77718,98 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "GPT", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "neversleep/llama-3-lumimaid-70b", + "permaslug": "openai/gpt-3.5-turbo-16k", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "neversleep/llama-3-lumimaid-70b", - "modelVariantPermaslug": "neversleep/llama-3-lumimaid-70b", - "providerName": "Featherless", + "modelVariantSlug": "openai/gpt-3.5-turbo-16k", + "modelVariantPermaslug": "openai/gpt-3.5-turbo-16k", + "providerName": "OpenAI", "providerInfo": { - "name": "Featherless", - "displayName": "Featherless", - "slug": "featherless", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false - } + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, - "moderationRequired": false, - "group": "Featherless", + "isAbortable": true, + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], - "isMultipartSupported": false, - "statusPageUrl": null, + "isMultipartSupported": true, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Featherless", - "providerModelId": "NeverSleep/Llama-3-Lumimaid-70B-v0.1", - "providerGroup": "Featherless", - "quantization": "fp8", + "providerDisplayName": "OpenAI", + "providerModelId": "gpt-3.5-turbo-16k", + "providerGroup": "OpenAI", + "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "top_k", - "min_p", - "seed" + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, + "requiresUserIds": true, "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.000004", - "completion": "0.000006", + "prompt": "0.000003", + "completion": "0.000004", "image": "0", "request": "0", "webSearch": "0", @@ -74577,9 +77820,9 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, - "supportsMultipart": false, + "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, @@ -74590,24 +77833,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 282, - "newest": 255, - "throughputHighToLow": 300, - "latencyLowToHigh": 214, - "pricingLowToHigh": 283, - "pricingHighToLow": 38 + "topWeekly": 283, + "newest": 306, + "throughputHighToLow": 42, + "latencyLowToHigh": 72, + "pricingLowToHigh": 258, + "pricingHighToLow": 60 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "Featherless", - "slug": "featherless", - "quantization": "fp8", - "context": 8192, + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": "unknown", + "context": 16385, "maxCompletionTokens": 4096, - "providerModelId": "NeverSleep/Llama-3-Lumimaid-70B-v0.1", + "providerModelId": "gpt-3.5-turbo-16k", "pricing": { - "prompt": "0.000004", - "completion": "0.000006", + "prompt": "0.000003", + "completion": "0.000004", "image": "0", "request": "0", "webSearch": "0", @@ -74615,156 +77860,203 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", - "top_k", - "min_p", - "seed" + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" ], - "inputCost": 4, - "outputCost": 6 + "inputCost": 3, + "outputCost": 4, + "throughput": 140.1575, + "latency": 510 + }, + { + "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", + "slug": "azure", + "quantization": "unknown", + "context": 16385, + "maxCompletionTokens": 4096, + "providerModelId": "gpt-35-turbo-16k", + "pricing": { + "prompt": "0.000003", + "completion": "0.000004", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "tools", + "tool_choice", + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "logit_bias", + "logprobs", + "top_logprobs", + "response_format" + ], + "inputCost": 3, + "outputCost": 4, + "throughput": 149.102, + "latency": 1347 } ] }, { - "slug": "anthropic/claude-2", - "hfSlug": null, + "slug": "x-ai/grok-vision-beta", + "hfSlug": "", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-11-22T00:00:00+00:00", + "createdAt": "2024-11-19T00:37:04.585936+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude v2", - "shortName": "Claude v2", - "author": "anthropic", - "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.", + "name": "xAI: Grok Vision Beta", + "shortName": "Grok Vision Beta", + "author": "x-ai", + "description": "Grok Vision Beta is xAI's experimental language model with vision capability.\n\n", "modelVersionGroupId": null, - "contextLength": 200000, + "contextLength": 8192, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", + "group": "Grok", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-2", + "permaslug": "x-ai/grok-vision-beta", "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "258ddc79-e520-4e68-8069-ea6771698c5a", - "name": "Anthropic | anthropic/claude-2", - "contextLength": 200000, + "features": {}, + "endpoint": { + "id": "641fa158-ffaa-45e0-a7cd-8d51bef6213b", + "name": "xAI | x-ai/grok-vision-beta", + "contextLength": 8192, "model": { - "slug": "anthropic/claude-2", - "hfSlug": null, + "slug": "x-ai/grok-vision-beta", + "hfSlug": "", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-11-22T00:00:00+00:00", + "createdAt": "2024-11-19T00:37:04.585936+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude v2", - "shortName": "Claude v2", - "author": "anthropic", - "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.", + "name": "xAI: Grok Vision Beta", + "shortName": "Grok Vision Beta", + "author": "x-ai", + "description": "Grok Vision Beta is xAI's experimental language model with vision capability.\n\n", "modelVersionGroupId": null, - "contextLength": 200000, + "contextLength": 8192, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Claude", + "group": "Grok", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-2", + "permaslug": "x-ai/grok-vision-beta", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "anthropic/claude-2", - "modelVariantPermaslug": "anthropic/claude-2", - "providerName": "Anthropic", + "modelVariantSlug": "x-ai/grok-vision-beta", + "modelVariantPermaslug": "x-ai/grok-vision-beta", + "providerName": "xAI", "providerInfo": { - "name": "Anthropic", - "displayName": "Anthropic", - "slug": "anthropic", + "name": "xAI", + "displayName": "xAI", + "slug": "xai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "termsOfServiceUrl": "https://x.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://x.ai/legal/privacy-policy", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 - }, - "requiresUserIds": true + } }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "Anthropic", + "moderationRequired": false, + "group": "xAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.anthropic.com/", + "statusPageUrl": "https://status.x.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Anthropic.svg" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://x.ai/&size=256" } }, - "providerDisplayName": "Anthropic", - "providerModelId": "claude-2.1", - "providerGroup": "Anthropic", - "quantization": "unknown", + "providerDisplayName": "xAI", + "providerModelId": "grok-vision-beta", + "providerGroup": "xAI", + "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "top_k", - "stop" + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "logprobs", + "top_logprobs", + "response_format" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "termsOfServiceUrl": "https://x.ai/legal/terms-of-service", + "privacyPolicyUrl": "https://x.ai/legal/privacy-policy", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 }, - "requiresUserIds": true, "training": false, "retainsPrompts": true, "retentionDays": 30 }, "pricing": { - "prompt": "0.000008", - "completion": "0.000024", - "image": "0", + "prompt": "0.000005", + "completion": "0.000015", + "image": "0.009", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -74787,25 +78079,27 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 283, - "newest": 294, - "throughputHighToLow": 293, - "latencyLowToHigh": 235, - "pricingLowToHigh": 296, - "pricingHighToLow": 19 + "topWeekly": 284, + "newest": 170, + "throughputHighToLow": 125, + "latencyLowToHigh": 64, + "pricingLowToHigh": 290, + "pricingHighToLow": 26 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://x.ai/\\u0026size=256", "providers": [ { - "name": "Anthropic", - "slug": "anthropic", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 4096, - "providerModelId": "claude-2.1", + "name": "xAI", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://x.ai/&size=256", + "slug": "xAi", + "quantization": null, + "context": 8192, + "maxCompletionTokens": null, + "providerModelId": "grok-vision-beta", "pricing": { - "prompt": "0.000008", - "completion": "0.000024", - "image": "0", + "prompt": "0.000005", + "completion": "0.000015", + "image": "0.009", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -74815,11 +78109,18 @@ export default { "max_tokens", "temperature", "top_p", - "top_k", - "stop" + "stop", + "frequency_penalty", + "presence_penalty", + "seed", + "logprobs", + "top_logprobs", + "response_format" ], - "inputCost": 8, - "outputCost": 24 + "inputCost": 5, + "outputCost": 15, + "throughput": 62.5, + "latency": 487 } ] }, @@ -74984,16 +78285,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 42, - "newest": 202, - "throughputHighToLow": 53, - "latencyLowToHigh": 50, + "topWeekly": 39, + "newest": 199, + "throughputHighToLow": 4, + "latencyLowToHigh": 91, "pricingLowToHigh": 63, "pricingHighToLow": 247 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", "slug": "nebiusAiStudio", "quantization": "fp8", "context": 131072, @@ -75022,10 +78325,13 @@ export default { "top_logprobs" ], "inputCost": 0.01, - "outputCost": 0.01 + "outputCost": 0.01, + "throughput": 31.5455, + "latency": 584.5 }, { "name": "DeepInfra", + "icon": "https://openrouter.ai/images/icons/DeepInfra.webp", "slug": "deepInfra", "quantization": "bf16", "context": 131072, @@ -75054,10 +78360,13 @@ export default { "min_p" ], "inputCost": 0.01, - "outputCost": 0.01 + "outputCost": 0.01, + "throughput": 144.546, + "latency": 649 }, { "name": "inference.net", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inference.net/&size=256", "slug": "inferenceNet", "quantization": "fp16", "context": 16384, @@ -75086,10 +78395,13 @@ export default { "top_logprobs" ], "inputCost": 0.01, - "outputCost": 0.01 + "outputCost": 0.01, + "throughput": 143.1995, + "latency": 1023 }, { "name": "Cloudflare", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.cloudflare.com/&size=256", "slug": "cloudflare", "quantization": "unknown", "context": 60000, @@ -75115,10 +78427,13 @@ export default { "presence_penalty" ], "inputCost": 0.03, - "outputCost": 0.2 + "outputCost": 0.2, + "throughput": 277.904, + "latency": 789 }, { "name": "SambaNova", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://sambanova.ai/&size=256", "slug": "sambaNova", "quantization": "bf16", "context": 16384, @@ -75141,397 +78456,24 @@ export default { "stop" ], "inputCost": 0.04, - "outputCost": 0.08 - } - ] - }, - { - "slug": "openai/o1-mini-2024-09-12", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-12T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "OpenAI: o1-mini (2024-09-12)", - "shortName": "o1-mini (2024-09-12)", - "author": "openai", - "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", - "modelVersionGroupId": null, - "contextLength": 128000, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "GPT", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "openai/o1-mini-2024-09-12", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "66ac30ec-dd07-4b8d-9dc0-5c7369bb2efe", - "name": "OpenAI | openai/o1-mini-2024-09-12", - "contextLength": 128000, - "model": { - "slug": "openai/o1-mini-2024-09-12", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-12T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "OpenAI: o1-mini (2024-09-12)", - "shortName": "o1-mini (2024-09-12)", - "author": "openai", - "description": "The latest and strongest model family from OpenAI, o1 is designed to spend more time thinking before responding.\n\nThe o1 models are optimized for math, science, programming, and other STEM-related tasks. They consistently exhibit PhD-level accuracy on benchmarks in physics, chemistry, and biology. Learn more in the [launch announcement](https://openai.com/o1).\n\nNote: This model is currently experimental and not suitable for production use-cases, and may be heavily rate-limited.", - "modelVersionGroupId": null, - "contextLength": 128000, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "GPT", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "openai/o1-mini-2024-09-12", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "openai/o1-mini-2024-09-12", - "modelVariantPermaslug": "openai/o1-mini-2024-09-12", - "providerName": "OpenAI", - "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true - } - }, - "providerDisplayName": "OpenAI", - "providerModelId": "o1-mini-2024-09-12", - "providerGroup": "OpenAI", - "quantization": "unknown", - "variant": "standard", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 65536, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "seed", - "max_tokens" - ], - "isByok": false, - "moderationRequired": true, - "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "pricing": { - "prompt": "0.0000011", - "completion": "0.0000044", - "image": "0", - "request": "0", - "inputCacheRead": "0.00000055", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 285, - "newest": 208, - "throughputHighToLow": 316, - "latencyLowToHigh": 292, - "pricingLowToHigh": 233, - "pricingHighToLow": 89 - }, - "providers": [ - { - "name": "OpenAI", - "slug": "openAi", - "quantization": "unknown", - "context": 128000, - "maxCompletionTokens": 65536, - "providerModelId": "o1-mini-2024-09-12", - "pricing": { - "prompt": "0.0000011", - "completion": "0.0000044", - "image": "0", - "request": "0", - "inputCacheRead": "0.00000055", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "seed", - "max_tokens" - ], - "inputCost": 1.1, - "outputCost": 4.4 - } - ] - }, - { - "slug": "aion-labs/aion-1.0-mini", - "hfSlug": "FuseAI/FuseO1-DeepSeekR1-QwQ-SkyT1-32B-Preview", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-04T19:25:07.903671+00:00", - "hfUpdatedAt": null, - "name": "AionLabs: Aion-1.0-Mini", - "shortName": "Aion-1.0-Mini", - "author": "aion-labs", - "description": "Aion-1.0-Mini 32B parameter model is a distilled version of the DeepSeek-R1 model, designed for strong performance in reasoning domains such as mathematics, coding, and logic. It is a modified variant of a FuseAI model that outperforms R1-Distill-Qwen-32B and R1-Distill-Llama-70B, with benchmark results available on its [Hugging Face page](https://huggingface.co/FuseAI/FuseO1-DeepSeekR1-QwQ-SkyT1-32B-Preview), independently replicated for verification.", - "modelVersionGroupId": null, - "contextLength": 131072, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "aion-labs/aion-1.0-mini", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "83eeb9eb-edea-49f5-abad-3f245f46420f", - "name": "AionLabs | aion-labs/aion-1.0-mini", - "contextLength": 131072, - "model": { - "slug": "aion-labs/aion-1.0-mini", - "hfSlug": "FuseAI/FuseO1-DeepSeekR1-QwQ-SkyT1-32B-Preview", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-04T19:25:07.903671+00:00", - "hfUpdatedAt": null, - "name": "AionLabs: Aion-1.0-Mini", - "shortName": "Aion-1.0-Mini", - "author": "aion-labs", - "description": "Aion-1.0-Mini 32B parameter model is a distilled version of the DeepSeek-R1 model, designed for strong performance in reasoning domains such as mathematics, coding, and logic. It is a modified variant of a FuseAI model that outperforms R1-Distill-Qwen-32B and R1-Distill-Llama-70B, with benchmark results available on its [Hugging Face page](https://huggingface.co/FuseAI/FuseO1-DeepSeekR1-QwQ-SkyT1-32B-Preview), independently replicated for verification.", - "modelVersionGroupId": null, - "contextLength": 16384, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "aion-labs/aion-1.0-mini", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "aion-labs/aion-1.0-mini", - "modelVariantPermaslug": "aion-labs/aion-1.0-mini", - "providerName": "AionLabs", - "providerInfo": { - "name": "AionLabs", - "displayName": "AionLabs", - "slug": "aion-labs", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://www.aionlabs.ai/terms/", - "privacyPolicyUrl": "https://www.aionlabs.ai/privacy-policy/", - "paidModels": { - "training": false - } - }, - "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, - "moderationRequired": false, - "group": "AionLabs", - "editors": [], - "owners": [], - "isMultipartSupported": false, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.aionlabs.ai/&size=256" - } - }, - "providerDisplayName": "AionLabs", - "providerModelId": "aion-labs/aion-1.0-mini", - "providerGroup": "AionLabs", - "quantization": "bf16", - "variant": "standard", - "isFree": false, - "canAbort": false, - "maxPromptTokens": null, - "maxCompletionTokens": 32768, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://www.aionlabs.ai/terms/", - "privacyPolicyUrl": "https://www.aionlabs.ai/privacy-policy/", - "paidModels": { - "training": false - }, - "training": false - }, - "pricing": { - "prompt": "0.0000007", - "completion": "0.0000014", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": true, - "supportsMultipart": false, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": false, - "hasChatCompletions": true, - "features": { - "supportedParameters": {}, - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 286, - "newest": 121, - "throughputHighToLow": 4, - "latencyLowToHigh": 267, - "pricingLowToHigh": 198, - "pricingHighToLow": 120 - }, - "providers": [ - { - "name": "AionLabs", - "slug": "aionLabs", - "quantization": "bf16", - "context": 131072, - "maxCompletionTokens": 32768, - "providerModelId": "aion-labs/aion-1.0-mini", - "pricing": { - "prompt": "0.0000007", - "completion": "0.0000014", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "reasoning", - "include_reasoning" - ], - "inputCost": 0.7, - "outputCost": 1.4 + "outputCost": 0.08, + "throughput": 4054.054, + "latency": 826 } ] }, { - "slug": "openai/gpt-3.5-turbo-16k", - "hfSlug": null, + "slug": "moonshotai/moonlight-16b-a3b-instruct", + "hfSlug": "moonshotai/Moonlight-16B-A3B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-08-28T00:00:00+00:00", + "createdAt": "2025-02-28T05:16:41.979606+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-3.5 Turbo 16k", - "shortName": "GPT-3.5 Turbo 16k", - "author": "openai", - "description": "This model offers four times the context length of gpt-3.5-turbo, allowing it to support approximately 20 pages of text in a single request at a higher cost. Training data: up to Sep 2021.", + "name": "Moonshot AI: Moonlight 16B A3B Instruct (free)", + "shortName": "Moonlight 16B A3B Instruct (free)", + "author": "moonshotai", + "description": "Moonlight-16B-A3B-Instruct is a 16B-parameter Mixture-of-Experts (MoE) language model developed by Moonshot AI. It is optimized for instruction-following tasks with 3B activated parameters per inference. The model advances the Pareto frontier in performance per FLOP across English, coding, math, and Chinese benchmarks. It outperforms comparable models like Llama3-3B and Deepseek-v2-Lite while maintaining efficient deployment capabilities through Hugging Face integration and compatibility with popular inference engines like vLLM12.", "modelVersionGroupId": null, - "contextLength": 16385, + "contextLength": 8192, "inputModalities": [ "text" ], @@ -75539,32 +78481,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-3.5-turbo-16k", + "permaslug": "moonshotai/moonlight-16b-a3b-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "eb1a93d0-a295-4afb-86d3-e2d10538c12d", - "name": "OpenAI | openai/gpt-3.5-turbo-16k", - "contextLength": 16385, + "id": "1d289d23-b3d0-43c3-b7f4-f169b7a1af3d", + "name": "Chutes | moonshotai/moonlight-16b-a3b-instruct:free", + "contextLength": 8192, "model": { - "slug": "openai/gpt-3.5-turbo-16k", - "hfSlug": null, + "slug": "moonshotai/moonlight-16b-a3b-instruct", + "hfSlug": "moonshotai/Moonlight-16B-A3B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-08-28T00:00:00+00:00", + "createdAt": "2025-02-28T05:16:41.979606+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-3.5 Turbo 16k", - "shortName": "GPT-3.5 Turbo 16k", - "author": "openai", - "description": "This model offers four times the context length of gpt-3.5-turbo, allowing it to support approximately 20 pages of text in a single request at a higher cost. Training data: up to Sep 2021.", + "name": "Moonshot AI: Moonlight 16B A3B Instruct", + "shortName": "Moonlight 16B A3B Instruct", + "author": "moonshotai", + "description": "Moonlight-16B-A3B-Instruct is a 16B-parameter Mixture-of-Experts (MoE) language model developed by Moonshot AI. It is optimized for instruction-following tasks with 3B activated parameters per inference. The model advances the Pareto frontier in performance per FLOP across English, coding, math, and Chinese benchmarks. It outperforms comparable models like Llama3-3B and Deepseek-v2-Lite while maintaining efficient deployment capabilities through Hugging Face integration and compatibility with popular inference engines like vLLM12.", "modelVersionGroupId": null, - "contextLength": 16385, + "contextLength": 8192, "inputModalities": [ "text" ], @@ -75572,67 +78514,59 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-3.5-turbo-16k", + "permaslug": "moonshotai/moonlight-16b-a3b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/gpt-3.5-turbo-16k", - "modelVariantPermaslug": "openai/gpt-3.5-turbo-16k", - "providerName": "OpenAI", + "modelVariantSlug": "moonshotai/moonlight-16b-a3b-instruct:free", + "modelVariantPermaslug": "moonshotai/moonlight-16b-a3b-instruct:free", + "providerName": "Chutes", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Chutes", + "displayName": "Chutes", + "slug": "chutes", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "training": true, + "retainsPrompts": true + } }, - "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", + "moderationRequired": false, + "group": "Chutes", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" } }, - "providerDisplayName": "OpenAI", - "providerModelId": "gpt-3.5-turbo-16k", - "providerGroup": "OpenAI", - "quantization": "unknown", - "variant": "standard", - "isFree": false, + "providerDisplayName": "Chutes", + "providerModelId": "moonshotai/Moonlight-16B-A3B-Instruct", + "providerGroup": "Chutes", + "quantization": "bf16", + "variant": "free", + "isFree": true, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", @@ -75640,30 +78574,27 @@ export default { "frequency_penalty", "presence_penalty", "seed", - "logit_bias", + "top_k", + "min_p", + "repetition_penalty", "logprobs", - "top_logprobs", - "response_format" + "logit_bias", + "top_logprobs" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://chutes.ai/tos", "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "training": true, + "retainsPrompts": true }, "pricing": { - "prompt": "0.000003", - "completion": "0.000004", + "prompt": "0", + "completion": "0", "image": "0", "request": "0", "webSearch": "0", @@ -75674,7 +78605,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -75687,96 +78618,28 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 287, - "newest": 306, - "throughputHighToLow": 38, - "latencyLowToHigh": 66, - "pricingLowToHigh": 258, - "pricingHighToLow": 60 + "topWeekly": 286, + "newest": 104, + "throughputHighToLow": 210, + "latencyLowToHigh": 198, + "pricingLowToHigh": 44, + "pricingHighToLow": 292 }, - "providers": [ - { - "name": "OpenAI", - "slug": "openAi", - "quantization": "unknown", - "context": 16385, - "maxCompletionTokens": 4096, - "providerModelId": "gpt-3.5-turbo-16k", - "pricing": { - "prompt": "0.000003", - "completion": "0.000004", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format" - ], - "inputCost": 3, - "outputCost": 4 - }, - { - "name": "Azure", - "slug": "azure", - "quantization": "unknown", - "context": 16385, - "maxCompletionTokens": 4096, - "providerModelId": "gpt-35-turbo-16k", - "pricing": { - "prompt": "0.000003", - "completion": "0.000004", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logit_bias", - "logprobs", - "top_logprobs", - "response_format" - ], - "inputCost": 3, - "outputCost": 4 - } - ] + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [] }, { - "slug": "eva-unit-01/eva-llama-3.33-70b", - "hfSlug": "EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.1", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-16T19:28:23.771236+00:00", + "slug": "arcee-ai/caller-large", + "hfSlug": "", + "updatedAt": "2025-05-05T23:32:30.837689+00:00", + "createdAt": "2025-05-05T23:31:09.181141+00:00", "hfUpdatedAt": null, - "name": "EVA Llama 3.33 70B", - "shortName": "EVA Llama 3.33 70B", - "author": "eva-unit-01", - "description": "EVA Llama 3.33 70b is a roleplay and storywriting specialist model. It is a full-parameter finetune of [Llama-3.3-70B-Instruct](https://openrouter.ai/meta-llama/llama-3.3-70b-instruct) on mixture of synthetic and natural data.\n\nIt uses Celeste 70B 0.1 data mixture, greatly expanding it to improve versatility, creativity and \"flavor\" of the resulting model\n\nThis model was built with Llama by Meta.\n", + "name": "Arcee AI: Caller Large", + "shortName": "Caller Large", + "author": "arcee-ai", + "description": "Caller Large is Arcee's specialist \"function‑calling\" SLM built to orchestrate external tools and APIs. Instead of maximizing next‑token accuracy, training focuses on structured JSON outputs, parameter extraction and multi‑step tool chains, making Caller a natural choice for retrieval‑augmented generation, robotic process automation or data‑pull chatbots. It incorporates a routing head that decides when (and how) to invoke a tool versus answering directly, reducing hallucinated calls. The model is already the backbone of Arcee Conductor's auto‑tool mode, where it parses user intent, emits clean function signatures and hands control back once the tool response is ready. Developers thus gain an OpenAI‑style function‑calling UX without handing requests to a frontier‑scale model. ", "modelVersionGroupId": null, - "contextLength": 16384, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -75784,35 +78647,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "eva-unit-01/eva-llama-3.33-70b", + "permaslug": "arcee-ai/caller-large", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "cc5c89d5-5596-48b0-af30-45b1fef9d8c5", - "name": "Featherless | eva-unit-01/eva-llama-3.33-70b", - "contextLength": 16384, + "id": "587c1cf8-5f90-42f7-9657-4d83c51a1bfb", + "name": "Together | arcee-ai/caller-large", + "contextLength": 32768, "model": { - "slug": "eva-unit-01/eva-llama-3.33-70b", - "hfSlug": "EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.1", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-12-16T19:28:23.771236+00:00", + "slug": "arcee-ai/caller-large", + "hfSlug": "", + "updatedAt": "2025-05-05T23:32:30.837689+00:00", + "createdAt": "2025-05-05T23:31:09.181141+00:00", "hfUpdatedAt": null, - "name": "EVA Llama 3.33 70B", - "shortName": "EVA Llama 3.33 70B", - "author": "eva-unit-01", - "description": "EVA Llama 3.33 70b is a roleplay and storywriting specialist model. It is a full-parameter finetune of [Llama-3.3-70B-Instruct](https://openrouter.ai/meta-llama/llama-3.3-70b-instruct) on mixture of synthetic and natural data.\n\nIt uses Celeste 70B 0.1 data mixture, greatly expanding it to improve versatility, creativity and \"flavor\" of the resulting model\n\nThis model was built with Llama by Meta.\n", + "name": "Arcee AI: Caller Large", + "shortName": "Caller Large", + "author": "arcee-ai", + "description": "Caller Large is Arcee's specialist \"function‑calling\" SLM built to orchestrate external tools and APIs. Instead of maximizing next‑token accuracy, training focuses on structured JSON outputs, parameter extraction and multi‑step tool chains, making Caller a natural choice for retrieval‑augmented generation, robotic process automation or data‑pull chatbots. It incorporates a routing head that decides when (and how) to invoke a tool versus answering directly, reducing hallucinated calls. The model is already the backbone of Arcee Conductor's auto‑tool mode, where it parses user intent, emits clean function signatures and hands control back once the tool response is ready. Developers thus gain an OpenAI‑style function‑calling UX without handing requests to a frontier‑scale model. ", "modelVersionGroupId": null, - "contextLength": 16384, + "contextLength": 32768, "inputModalities": [ "text" ], @@ -75820,31 +78680,28 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "eva-unit-01/eva-llama-3.33-70b", + "permaslug": "arcee-ai/caller-large", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "eva-unit-01/eva-llama-3.33-70b", - "modelVariantPermaslug": "eva-unit-01/eva-llama-3.33-70b", - "providerName": "Featherless", + "modelVariantSlug": "arcee-ai/caller-large", + "modelVariantPermaslug": "arcee-ai/caller-large", + "providerName": "Together", "providerInfo": { - "name": "Featherless", - "displayName": "Featherless", - "slug": "featherless", + "name": "Together", + "displayName": "Together", + "slug": "together", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -75853,47 +78710,50 @@ export default { "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, + "isAbortable": true, "moderationRequired": false, - "group": "Featherless", + "group": "Together", "editors": [], "owners": [], - "isMultipartSupported": false, + "isMultipartSupported": true, "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" } }, - "providerDisplayName": "Featherless", - "providerModelId": "EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.0", - "providerGroup": "Featherless", - "quantization": "fp8", + "providerDisplayName": "Together", + "providerModelId": "arcee-ai/caller", + "providerGroup": "Together", + "quantization": null, "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", "top_k", + "repetition_penalty", + "logit_bias", "min_p", - "seed" + "response_format" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -75902,8 +78762,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.000004", - "completion": "0.000006", + "prompt": "0.00000055", + "completion": "0.00000085", "image": "0", "request": "0", "webSearch": "0", @@ -75914,37 +78774,40 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, - "supportsMultipart": false, + "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 288, - "newest": 152, - "throughputHighToLow": 298, - "latencyLowToHigh": 313, - "pricingLowToHigh": 280, - "pricingHighToLow": 35 + "topWeekly": 287, + "newest": 3, + "throughputHighToLow": 88, + "latencyLowToHigh": 52, + "pricingLowToHigh": 186, + "pricingHighToLow": 132 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://arcee.ai/\\u0026size=256", "providers": [ { - "name": "Featherless", - "slug": "featherless", - "quantization": "fp8", - "context": 16384, - "maxCompletionTokens": 4096, - "providerModelId": "EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.0", + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": null, + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "arcee-ai/caller", "pricing": { - "prompt": "0.000004", - "completion": "0.000006", + "prompt": "0.00000055", + "completion": "0.00000085", "image": "0", "request": "0", "webSearch": "0", @@ -75952,36 +78815,40 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", "top_k", + "repetition_penalty", + "logit_bias", "min_p", - "seed" + "response_format" ], - "inputCost": 4, - "outputCost": 6 + "inputCost": 0.55, + "outputCost": 0.85, + "throughput": 86.4875, + "latency": 471 } ] }, { - "slug": "arcee-ai/spotlight", - "hfSlug": "", - "updatedAt": "2025-05-05T21:47:17.011692+00:00", - "createdAt": "2025-05-05T21:45:52.249082+00:00", + "slug": "aion-labs/aion-1.0-mini", + "hfSlug": "FuseAI/FuseO1-DeepSeekR1-QwQ-SkyT1-32B-Preview", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-04T19:25:07.903671+00:00", "hfUpdatedAt": null, - "name": "Arcee AI: Spotlight", - "shortName": "Spotlight", - "author": "arcee-ai", - "description": "Spotlight is a 7‑billion‑parameter vision‑language model derived from Qwen 2.5‑VL and fine‑tuned by Arcee AI for tight image‑text grounding tasks. It offers a 32 k‑token context window, enabling rich multimodal conversations that combine lengthy documents with one or more images. Training emphasized fast inference on consumer GPUs while retaining strong captioning, visual‐question‑answering, and diagram‑analysis accuracy. As a result, Spotlight slots neatly into agent workflows where screenshots, charts or UI mock‑ups need to be interpreted on the fly. Early benchmarks show it matching or out‑scoring larger VLMs such as LLaVA‑1.6 13 B on popular VQA and POPE alignment tests. ", + "name": "AionLabs: Aion-1.0-Mini", + "shortName": "Aion-1.0-Mini", + "author": "aion-labs", + "description": "Aion-1.0-Mini 32B parameter model is a distilled version of the DeepSeek-R1 model, designed for strong performance in reasoning domains such as mathematics, coding, and logic. It is a modified variant of a FuseAI model that outperforms R1-Distill-Qwen-32B and R1-Distill-Llama-70B, with benchmark results available on its [Hugging Face page](https://huggingface.co/FuseAI/FuseO1-DeepSeekR1-QwQ-SkyT1-32B-Preview), independently replicated for verification.", "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ - "image", "text" ], "outputModalities": [ @@ -75995,27 +78862,26 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "arcee-ai/spotlight", + "permaslug": "aion-labs/aion-1.0-mini", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "a9b3fe6f-e21f-4f3c-9ea7-f70d856939d6", - "name": "Together | arcee-ai/spotlight", + "id": "83eeb9eb-edea-49f5-abad-3f245f46420f", + "name": "AionLabs | aion-labs/aion-1.0-mini", "contextLength": 131072, "model": { - "slug": "arcee-ai/spotlight", - "hfSlug": "", - "updatedAt": "2025-05-05T21:47:17.011692+00:00", - "createdAt": "2025-05-05T21:45:52.249082+00:00", + "slug": "aion-labs/aion-1.0-mini", + "hfSlug": "FuseAI/FuseO1-DeepSeekR1-QwQ-SkyT1-32B-Preview", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-02-04T19:25:07.903671+00:00", "hfUpdatedAt": null, - "name": "Arcee AI: Spotlight", - "shortName": "Spotlight", - "author": "arcee-ai", - "description": "Spotlight is a 7‑billion‑parameter vision‑language model derived from Qwen 2.5‑VL and fine‑tuned by Arcee AI for tight image‑text grounding tasks. It offers a 32 k‑token context window, enabling rich multimodal conversations that combine lengthy documents with one or more images. Training emphasized fast inference on consumer GPUs while retaining strong captioning, visual‐question‑answering, and diagram‑analysis accuracy. As a result, Spotlight slots neatly into agent workflows where screenshots, charts or UI mock‑ups need to be interpreted on the fly. Early benchmarks show it matching or out‑scoring larger VLMs such as LLaVA‑1.6 13 B on popular VQA and POPE alignment tests. ", + "name": "AionLabs: Aion-1.0-Mini", + "shortName": "Aion-1.0-Mini", + "author": "aion-labs", + "description": "Aion-1.0-Mini 32B parameter model is a distilled version of the DeepSeek-R1 model, designed for strong performance in reasoning domains such as mathematics, coding, and logic. It is a modified variant of a FuseAI model that outperforms R1-Distill-Qwen-32B and R1-Distill-Llama-70B, with benchmark results available on its [Hugging Face page](https://huggingface.co/FuseAI/FuseO1-DeepSeekR1-QwQ-SkyT1-32B-Preview), independently replicated for verification.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 16384, "inputModalities": [ - "image", "text" ], "outputModalities": [ @@ -76029,81 +78895,71 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "arcee-ai/spotlight", + "permaslug": "aion-labs/aion-1.0-mini", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "arcee-ai/spotlight", - "modelVariantPermaslug": "arcee-ai/spotlight", - "providerName": "Together", + "modelVariantSlug": "aion-labs/aion-1.0-mini", + "modelVariantPermaslug": "aion-labs/aion-1.0-mini", + "providerName": "AionLabs", "providerInfo": { - "name": "Together", - "displayName": "Together", - "slug": "together", + "name": "AionLabs", + "displayName": "AionLabs", + "slug": "aion-labs", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://www.aionlabs.ai/terms/", + "privacyPolicyUrl": "https://www.aionlabs.ai/privacy-policy/", "paidModels": { - "training": false, - "retainsPrompts": false + "training": false } }, - "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, + "hasCompletions": false, + "isAbortable": false, "moderationRequired": false, - "group": "Together", + "group": "AionLabs", "editors": [], "owners": [], - "isMultipartSupported": true, + "isMultipartSupported": false, "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.aionlabs.ai/&size=256" } }, - "providerDisplayName": "Together", - "providerModelId": "arcee_ai/arcee-spotlight", - "providerGroup": "Together", - "quantization": null, + "providerDisplayName": "AionLabs", + "providerModelId": "aion-labs/aion-1.0-mini", + "providerGroup": "AionLabs", + "quantization": "bf16", "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 65537, + "maxCompletionTokens": 32768, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "reasoning", + "include_reasoning" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://www.aionlabs.ai/terms/", + "privacyPolicyUrl": "https://www.aionlabs.ai/privacy-policy/", "paidModels": { - "training": false, - "retainsPrompts": false + "training": false }, - "training": false, - "retainsPrompts": false + "training": false }, "pricing": { - "prompt": "0.00000018", - "completion": "0.00000018", + "prompt": "0.0000007", + "completion": "0.0000014", "image": "0", "request": "0", "webSearch": "0", @@ -76115,11 +78971,11 @@ export default { "isDeranked": false, "isDisabled": false, "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, + "supportsReasoning": true, + "supportsMultipart": false, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { "supportedParameters": {}, @@ -76128,24 +78984,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 289, - "newest": 4, - "throughputHighToLow": 16, - "latencyLowToHigh": 0, - "pricingLowToHigh": 140, - "pricingHighToLow": 176 + "topWeekly": 288, + "newest": 121, + "throughputHighToLow": 93, + "latencyLowToHigh": 309, + "pricingLowToHigh": 198, + "pricingHighToLow": 120 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://www.aionlabs.ai/\\u0026size=256", "providers": [ { - "name": "Together", - "slug": "together", - "quantization": null, + "name": "AionLabs", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.aionlabs.ai/&size=256", + "slug": "aionLabs", + "quantization": "bf16", "context": 131072, - "maxCompletionTokens": 65537, - "providerModelId": "arcee_ai/arcee-spotlight", + "maxCompletionTokens": 32768, + "providerModelId": "aion-labs/aion-1.0-mini", "pricing": { - "prompt": "0.00000018", - "completion": "0.00000018", + "prompt": "0.0000007", + "completion": "0.0000014", "image": "0", "request": "0", "webSearch": "0", @@ -76156,17 +79014,13 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "reasoning", + "include_reasoning" ], - "inputCost": 0.18, - "outputCost": 0.18 + "inputCost": 0.7, + "outputCost": 1.4, + "throughput": 83.769, + "latency": 5546 } ] }, @@ -76337,16 +79191,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 290, + "topWeekly": 289, "newest": 307, - "throughputHighToLow": 237, - "latencyLowToHigh": 218, + "throughputHighToLow": 234, + "latencyLowToHigh": 231, "pricingLowToHigh": 315, "pricingHighToLow": 2 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", "slug": "openAi", "quantization": "unknown", "context": 32767, @@ -76377,10 +79233,13 @@ export default { "response_format" ], "inputCost": 60, - "outputCost": 120 + "outputCost": 120, + "throughput": 32.5215, + "latency": 1550 }, { "name": "Azure", + "icon": "https://openrouter.ai/images/icons/Azure.svg", "slug": "azure", "quantization": "unknown", "context": 32767, @@ -76411,22 +79270,24 @@ export default { "response_format" ], "inputCost": 60, - "outputCost": 120 + "outputCost": 120, + "throughput": 267.708, + "latency": 9512 } ] }, { - "slug": "moonshotai/moonlight-16b-a3b-instruct", - "hfSlug": "moonshotai/Moonlight-16B-A3B-Instruct", + "slug": "anthropic/claude-2.1", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-28T05:16:41.979606+00:00", + "createdAt": "2023-11-22T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Moonshot AI: Moonlight 16B A3B Instruct (free)", - "shortName": "Moonlight 16B A3B Instruct (free)", - "author": "moonshotai", - "description": "Moonlight-16B-A3B-Instruct is a 16B-parameter Mixture-of-Experts (MoE) language model developed by Moonshot AI. It is optimized for instruction-following tasks with 3B activated parameters per inference. The model advances the Pareto frontier in performance per FLOP across English, coding, math, and Chinese benchmarks. It outperforms comparable models like Llama3-3B and Deepseek-v2-Lite while maintaining efficient deployment capabilities through Hugging Face integration and compatibility with popular inference engines like vLLM12.", + "name": "Anthropic: Claude v2.1 (self-moderated)", + "shortName": "Claude v2.1 (self-moderated)", + "author": "anthropic", + "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 200000, "inputModalities": [ "text" ], @@ -76434,32 +79295,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "Claude", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "moonshotai/moonlight-16b-a3b-instruct", + "permaslug": "anthropic/claude-2.1", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "1d289d23-b3d0-43c3-b7f4-f169b7a1af3d", - "name": "Chutes | moonshotai/moonlight-16b-a3b-instruct:free", - "contextLength": 8192, + "id": "7159a751-fb9d-4750-918b-a774fd71cfa3", + "name": "Anthropic | anthropic/claude-2.1:beta", + "contextLength": 200000, "model": { - "slug": "moonshotai/moonlight-16b-a3b-instruct", - "hfSlug": "moonshotai/Moonlight-16B-A3B-Instruct", + "slug": "anthropic/claude-2.1", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-02-28T05:16:41.979606+00:00", + "createdAt": "2023-11-22T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Moonshot AI: Moonlight 16B A3B Instruct", - "shortName": "Moonlight 16B A3B Instruct", - "author": "moonshotai", - "description": "Moonlight-16B-A3B-Instruct is a 16B-parameter Mixture-of-Experts (MoE) language model developed by Moonshot AI. It is optimized for instruction-following tasks with 3B activated parameters per inference. The model advances the Pareto frontier in performance per FLOP across English, coding, math, and Chinese benchmarks. It outperforms comparable models like Llama3-3B and Deepseek-v2-Lite while maintaining efficient deployment capabilities through Hugging Face integration and compatibility with popular inference engines like vLLM12.", + "name": "Anthropic: Claude v2.1", + "shortName": "Claude v2.1", + "author": "anthropic", + "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 200000, "inputModalities": [ "text" ], @@ -76467,87 +79328,87 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "Claude", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "moonshotai/moonlight-16b-a3b-instruct", + "permaslug": "anthropic/claude-2.1", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "moonshotai/moonlight-16b-a3b-instruct:free", - "modelVariantPermaslug": "moonshotai/moonlight-16b-a3b-instruct:free", - "providerName": "Chutes", + "modelVariantSlug": "anthropic/claude-2.1:beta", + "modelVariantPermaslug": "anthropic/claude-2.1:beta", + "providerName": "Anthropic", "providerInfo": { - "name": "Chutes", - "displayName": "Chutes", - "slug": "chutes", + "name": "Anthropic", + "displayName": "Anthropic", + "slug": "anthropic", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", "paidModels": { - "training": true, - "retainsPrompts": true - } + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": false, - "group": "Chutes", + "moderationRequired": true, + "group": "Anthropic", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.anthropic.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://chutes.ai/&size=256" + "url": "/images/icons/Anthropic.svg" } }, - "providerDisplayName": "Chutes", - "providerModelId": "moonshotai/Moonlight-16B-A3B-Instruct", - "providerGroup": "Chutes", - "quantization": "bf16", - "variant": "free", - "isFree": true, + "providerDisplayName": "Anthropic", + "providerModelId": "claude-2.1", + "providerGroup": "Anthropic", + "quantization": "unknown", + "variant": "beta", + "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", "top_k", - "min_p", - "repetition_penalty", - "logprobs", - "logit_bias", - "top_logprobs" + "stop" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://chutes.ai/tos", + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", "paidModels": { - "training": true, - "retainsPrompts": true + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, - "training": true, - "retainsPrompts": true + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0", - "completion": "0", + "prompt": "0.000008", + "completion": "0.000024", "image": "0", "request": "0", "webSearch": "0", @@ -76571,14 +79432,45 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 291, - "newest": 104, - "throughputHighToLow": 209, - "latencyLowToHigh": 203, - "pricingLowToHigh": 44, - "pricingHighToLow": 292 + "topWeekly": 267, + "newest": 294, + "throughputHighToLow": 293, + "latencyLowToHigh": 224, + "pricingLowToHigh": 298, + "pricingHighToLow": 17 }, - "providers": [] + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", + "providers": [ + { + "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", + "slug": "anthropic", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 4096, + "providerModelId": "claude-2.1", + "pricing": { + "prompt": "0.000008", + "completion": "0.000024", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 8, + "outputCost": 24, + "throughput": 16.038, + "latency": 1532 + } + ] }, { "slug": "cohere/command-r-03-2024", @@ -76741,16 +79633,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 292, + "topWeekly": 291, "newest": 283, - "throughputHighToLow": 40, - "latencyLowToHigh": 35, - "pricingLowToHigh": 188, - "pricingHighToLow": 129 + "throughputHighToLow": 45, + "latencyLowToHigh": 30, + "pricingLowToHigh": 189, + "pricingHighToLow": 128 }, + "authorIcon": "https://openrouter.ai/images/icons/Cohere.png", "providers": [ { "name": "Cohere", + "icon": "https://openrouter.ai/images/icons/Cohere.png", "slug": "cohere", "quantization": "unknown", "context": 128000, @@ -76779,7 +79673,9 @@ export default { "structured_outputs" ], "inputCost": 0.5, - "outputCost": 1.5 + "outputCost": 1.5, + "throughput": 139.4435, + "latency": 324 } ] }, @@ -76944,22 +79840,232 @@ export default { }, "providerRegion": null }, + "sorting": { + "topWeekly": 292, + "newest": 265, + "throughputHighToLow": 265, + "latencyLowToHigh": 144, + "pricingLowToHigh": 209, + "pricingHighToLow": 113 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [ + { + "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", + "slug": "featherless", + "quantization": "fp8", + "context": 4096, + "maxCompletionTokens": 4096, + "providerModelId": "Sao10K/Fimbulvetr-11B-v2", + "pricing": { + "prompt": "0.0000008", + "completion": "0.0000012", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "top_k", + "min_p", + "seed" + ], + "inputCost": 0.8, + "outputCost": 1.2, + "throughput": 23.299, + "latency": 883 + } + ] + }, + { + "slug": "undi95/toppy-m-7b", + "hfSlug": "Undi95/Toppy-M-7B", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2023-11-10T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Toppy M 7B", + "shortName": "Toppy M 7B", + "author": "undi95", + "description": "A wild 7B parameter model that merges several models using the new task_arithmetic merge method from mergekit.\nList of merged models:\n- NousResearch/Nous-Capybara-7B-V1.9\n- [HuggingFaceH4/zephyr-7b-beta](/models/huggingfaceh4/zephyr-7b-beta)\n- lemonilia/AshhLimaRP-Mistral-7B\n- Vulkane/120-Days-of-Sodom-LoRA-Mistral-7b\n- Undi95/Mistral-pippa-sharegpt-7b-qlora\n\n#merge #uncensored", + "modelVersionGroupId": null, + "contextLength": 4096, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Mistral", + "instructType": "alpaca", + "defaultSystem": null, + "defaultStops": [ + "###", + "" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "undi95/toppy-m-7b", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "d2b3973b-76cb-43ef-affa-1ca7e1b25ea2", + "name": "Featherless | undi95/toppy-m-7b", + "contextLength": 4096, + "model": { + "slug": "undi95/toppy-m-7b", + "hfSlug": "Undi95/Toppy-M-7B", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2023-11-10T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Toppy M 7B", + "shortName": "Toppy M 7B", + "author": "undi95", + "description": "A wild 7B parameter model that merges several models using the new task_arithmetic merge method from mergekit.\nList of merged models:\n- NousResearch/Nous-Capybara-7B-V1.9\n- [HuggingFaceH4/zephyr-7b-beta](/models/huggingfaceh4/zephyr-7b-beta)\n- lemonilia/AshhLimaRP-Mistral-7B\n- Vulkane/120-Days-of-Sodom-LoRA-Mistral-7b\n- Undi95/Mistral-pippa-sharegpt-7b-qlora\n\n#merge #uncensored", + "modelVersionGroupId": null, + "contextLength": 4096, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Mistral", + "instructType": "alpaca", + "defaultSystem": null, + "defaultStops": [ + "###", + "" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "undi95/toppy-m-7b", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "undi95/toppy-m-7b", + "modelVariantPermaslug": "undi95/toppy-m-7b", + "providerName": "Featherless", + "providerInfo": { + "name": "Featherless", + "displayName": "Featherless", + "slug": "featherless", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", + "paidModels": { + "training": false, + "retainsPrompts": false + } + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": false, + "moderationRequired": false, + "group": "Featherless", + "editors": [], + "owners": [], + "isMultipartSupported": false, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" + } + }, + "providerDisplayName": "Featherless", + "providerModelId": "Undi95/Toppy-M-7B", + "providerGroup": "Featherless", + "quantization": null, + "variant": "standard", + "isFree": false, + "canAbort": false, + "maxPromptTokens": null, + "maxCompletionTokens": 4096, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "top_k", + "min_p", + "seed" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", + "paidModels": { + "training": false, + "retainsPrompts": false + }, + "training": false, + "retainsPrompts": false + }, + "pricing": { + "prompt": "0.0000008", + "completion": "0.0000012", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": false, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": false, + "features": { + "supportedParameters": {}, + "supportsDocumentUrl": null + }, + "providerRegion": null + }, "sorting": { "topWeekly": 293, - "newest": 265, - "throughputHighToLow": 265, - "latencyLowToHigh": 146, + "newest": 298, + "throughputHighToLow": 267, + "latencyLowToHigh": 186, "pricingLowToHigh": 210, - "pricingHighToLow": 112 + "pricingHighToLow": 114 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", "slug": "featherless", - "quantization": "fp8", + "quantization": null, "context": 4096, "maxCompletionTokens": 4096, - "providerModelId": "Sao10K/Fimbulvetr-11B-v2", + "providerModelId": "Undi95/Toppy-M-7B", "pricing": { "prompt": "0.0000008", "completion": "0.0000012", @@ -76982,22 +80088,24 @@ export default { "seed" ], "inputCost": 0.8, - "outputCost": 1.2 + "outputCost": 1.2, + "throughput": 23.1435, + "latency": 1232 } ] }, { - "slug": "arcee-ai/virtuoso-medium-v2", - "hfSlug": "arcee-ai/Virtuoso-Medium-v2", - "updatedAt": "2025-05-05T21:48:27.917466+00:00", - "createdAt": "2025-05-05T20:53:54.621827+00:00", + "slug": "ai21/jamba-1.6-mini", + "hfSlug": "ai21labs/AI21-Jamba-Mini-1.6", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-03-13T22:32:51+00:00", "hfUpdatedAt": null, - "name": "Arcee AI: Virtuoso Medium V2", - "shortName": "Virtuoso Medium V2", - "author": "arcee-ai", - "description": "Virtuoso‑Medium‑v2 is a 32 B model distilled from DeepSeek‑v3 logits and merged back onto a Qwen 2.5 backbone, yielding a sharper, more factual successor to the original Virtuoso Medium. The team harvested ~1.1 B logit tokens and applied \"fusion‑merging\" plus DPO alignment, which pushed scores past Arcee‑Nova 2024 and many 40 B‑plus peers on MMLU‑Pro, MATH and HumanEval. With a 128 k context and aggressive quantization options (from BF16 down to 4‑bit GGUF), it balances capability with deployability on single‑GPU nodes. Typical use cases include enterprise chat assistants, technical writing aids and medium‑complexity code drafting where Virtuoso‑Large would be overkill. ", - "modelVersionGroupId": null, - "contextLength": 131072, + "name": "AI21: Jamba Mini 1.6", + "shortName": "Jamba Mini 1.6", + "author": "ai21", + "description": "AI21 Jamba Mini 1.6 is a hybrid foundation model combining State Space Models (Mamba) with Transformer attention mechanisms. With 12 billion active parameters (52 billion total), this model excels in extremely long-context tasks (up to 256K tokens) and achieves superior inference efficiency, outperforming comparable open models on tasks such as retrieval-augmented generation (RAG) and grounded question answering. Jamba Mini 1.6 supports multilingual tasks across English, Spanish, French, Portuguese, Italian, Dutch, German, Arabic, and Hebrew, along with structured JSON output and tool-use capabilities.\n\nUsage of this model is subject to the [Jamba Open Model License](https://www.ai21.com/licenses/jamba-open-model-license).", + "modelVersionGroupId": "cd1eb031-30bc-4e2e-aa06-3c20f986e5c7", + "contextLength": 256000, "inputModalities": [ "text" ], @@ -77012,25 +80120,25 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "arcee-ai/virtuoso-medium-v2", + "permaslug": "ai21/jamba-1.6-mini", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "b3494a2a-b7d6-4055-a379-d77711c85c7b", - "name": "Together | arcee-ai/virtuoso-medium-v2", - "contextLength": 131072, + "id": "8c1760f3-6c79-4ba7-9d89-6d5168b075d8", + "name": "AI21 | ai21/jamba-1.6-mini", + "contextLength": 256000, "model": { - "slug": "arcee-ai/virtuoso-medium-v2", - "hfSlug": "arcee-ai/Virtuoso-Medium-v2", - "updatedAt": "2025-05-05T21:48:27.917466+00:00", - "createdAt": "2025-05-05T20:53:54.621827+00:00", + "slug": "ai21/jamba-1.6-mini", + "hfSlug": "ai21labs/AI21-Jamba-Mini-1.6", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2025-03-13T22:32:51+00:00", "hfUpdatedAt": null, - "name": "Arcee AI: Virtuoso Medium V2", - "shortName": "Virtuoso Medium V2", - "author": "arcee-ai", - "description": "Virtuoso‑Medium‑v2 is a 32 B model distilled from DeepSeek‑v3 logits and merged back onto a Qwen 2.5 backbone, yielding a sharper, more factual successor to the original Virtuoso Medium. The team harvested ~1.1 B logit tokens and applied \"fusion‑merging\" plus DPO alignment, which pushed scores past Arcee‑Nova 2024 and many 40 B‑plus peers on MMLU‑Pro, MATH and HumanEval. With a 128 k context and aggressive quantization options (from BF16 down to 4‑bit GGUF), it balances capability with deployability on single‑GPU nodes. Typical use cases include enterprise chat assistants, technical writing aids and medium‑complexity code drafting where Virtuoso‑Large would be overkill. ", - "modelVersionGroupId": null, - "contextLength": 131072, + "name": "AI21: Jamba Mini 1.6", + "shortName": "Jamba Mini 1.6", + "author": "ai21", + "description": "AI21 Jamba Mini 1.6 is a hybrid foundation model combining State Space Models (Mamba) with Transformer attention mechanisms. With 12 billion active parameters (52 billion total), this model excels in extremely long-context tasks (up to 256K tokens) and achieves superior inference efficiency, outperforming comparable open models on tasks such as retrieval-augmented generation (RAG) and grounded question answering. Jamba Mini 1.6 supports multilingual tasks across English, Spanish, French, Portuguese, Italian, Dutch, German, Arabic, and Hebrew, along with structured JSON output and tool-use capabilities.\n\nUsage of this model is subject to the [Jamba Open Model License](https://www.ai21.com/licenses/jamba-open-model-license).", + "modelVersionGroupId": "cd1eb031-30bc-4e2e-aa06-3c20f986e5c7", + "contextLength": 256000, "inputModalities": [ "text" ], @@ -77045,81 +80153,73 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "arcee-ai/virtuoso-medium-v2", + "permaslug": "ai21/jamba-1.6-mini", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "arcee-ai/virtuoso-medium-v2", - "modelVariantPermaslug": "arcee-ai/virtuoso-medium-v2", - "providerName": "Together", + "modelVariantSlug": "ai21/jamba-1.6-mini", + "modelVariantPermaslug": "ai21/jamba-1.6-mini", + "providerName": "AI21", "providerInfo": { - "name": "Together", - "displayName": "Together", - "slug": "together", + "name": "AI21", + "displayName": "AI21", + "slug": "ai21", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "privacyPolicyUrl": "https://www.ai21.com/privacy-policy/", + "termsOfServiceUrl": "https://www.ai21.com/terms-of-service/", "paidModels": { - "training": false, - "retainsPrompts": false + "training": false } }, - "headquarters": "US", + "headquarters": "IL", "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, + "hasCompletions": false, + "isAbortable": false, "moderationRequired": false, - "group": "Together", + "group": "AI21", "editors": [], "owners": [], - "isMultipartSupported": true, - "statusPageUrl": null, + "isMultipartSupported": false, + "statusPageUrl": "https://status.ai21.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://ai21.com/&size=256" } }, - "providerDisplayName": "Together", - "providerModelId": "arcee-ai/virtuoso-medium-v2", - "providerGroup": "Together", - "quantization": null, + "providerDisplayName": "AI21", + "providerModelId": "jamba-mini-1.6", + "providerGroup": "AI21", + "quantization": "bf16", "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 32768, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "stop" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "privacyPolicyUrl": "https://www.ai21.com/privacy-policy/", + "termsOfServiceUrl": "https://www.ai21.com/terms-of-service/", "paidModels": { - "training": false, - "retainsPrompts": false + "training": false }, - "training": false, - "retainsPrompts": false + "training": false }, "pricing": { - "prompt": "0.0000005", - "completion": "0.0000008", + "prompt": "0.0000002", + "completion": "0.0000004", "image": "0", "request": "0", "webSearch": "0", @@ -77130,38 +80230,39 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, - "supportsMultipart": true, + "supportsMultipart": false, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 294, - "newest": 8, - "throughputHighToLow": 144, - "latencyLowToHigh": 93, - "pricingLowToHigh": 183, - "pricingHighToLow": 135 + "newest": 86, + "throughputHighToLow": 12, + "latencyLowToHigh": 73, + "pricingLowToHigh": 150, + "pricingHighToLow": 168 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai21.com/\\u0026size=256", "providers": [ { - "name": "Together", - "slug": "together", - "quantization": null, - "context": 131072, - "maxCompletionTokens": 32768, - "providerModelId": "arcee-ai/virtuoso-medium-v2", + "name": "AI21", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://ai21.com/&size=256", + "slug": "ai21", + "quantization": "bf16", + "context": 256000, + "maxCompletionTokens": 4096, + "providerModelId": "jamba-mini-1.6", "pricing": { - "prompt": "0.0000005", - "completion": "0.0000008", + "prompt": "0.0000002", + "completion": "0.0000004", "image": "0", "request": "0", "webSearch": "0", @@ -77169,20 +80270,17 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "top_k", - "repetition_penalty", - "logit_bias", - "min_p", - "response_format" + "stop" ], - "inputCost": 0.5, - "outputCost": 0.8 + "inputCost": 0.2, + "outputCost": 0.4, + "throughput": 208.3575, + "latency": 503 } ] }, @@ -77357,14 +80455,16 @@ export default { "sorting": { "topWeekly": 295, "newest": 78, - "throughputHighToLow": 296, - "latencyLowToHigh": 315, + "throughputHighToLow": 305, + "latencyLowToHigh": 314, "pricingLowToHigh": 318, "pricingHighToLow": 0 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", "slug": "openAi", "quantization": null, "context": 200000, @@ -77395,20 +80495,266 @@ export default { "response_format" ], "inputCost": 150, - "outputCost": 600 + "outputCost": 600, + "throughput": 12.56, + "latency": 156692 } ] }, { - "slug": "undi95/toppy-m-7b", - "hfSlug": "Undi95/Toppy-M-7B", + "slug": "anthracite-org/magnum-v2-72b", + "hfSlug": "anthracite-org/magnum-v2-72b", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-11-10T00:00:00+00:00", + "createdAt": "2024-09-30T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Toppy M 7B", - "shortName": "Toppy M 7B", - "author": "undi95", - "description": "A wild 7B parameter model that merges several models using the new task_arithmetic merge method from mergekit.\nList of merged models:\n- NousResearch/Nous-Capybara-7B-V1.9\n- [HuggingFaceH4/zephyr-7b-beta](/models/huggingfaceh4/zephyr-7b-beta)\n- lemonilia/AshhLimaRP-Mistral-7B\n- Vulkane/120-Days-of-Sodom-LoRA-Mistral-7b\n- Undi95/Mistral-pippa-sharegpt-7b-qlora\n\n#merge #uncensored", + "name": "Magnum v2 72B", + "shortName": "Magnum v2 72B", + "author": "anthracite-org", + "description": "From the maker of [Goliath](https://openrouter.ai/models/alpindale/goliath-120b), Magnum 72B is the seventh in a family of models designed to achieve the prose quality of the Claude 3 models, notably Opus & Sonnet.\n\nThe model is based on [Qwen2 72B](https://openrouter.ai/models/qwen/qwen-2-72b-instruct) and trained with 55 million tokens of highly curated roleplay (RP) data.", + "modelVersionGroupId": null, + "contextLength": 32768, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Qwen", + "instructType": "chatml", + "defaultSystem": null, + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "anthracite-org/magnum-v2-72b", + "reasoningConfig": null, + "features": {}, + "endpoint": { + "id": "740509db-9804-483c-9004-65a3884d7be7", + "name": "Infermatic | anthracite-org/magnum-v2-72b", + "contextLength": 32768, + "model": { + "slug": "anthracite-org/magnum-v2-72b", + "hfSlug": "anthracite-org/magnum-v2-72b", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-09-30T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Magnum v2 72B", + "shortName": "Magnum v2 72B", + "author": "anthracite-org", + "description": "From the maker of [Goliath](https://openrouter.ai/models/alpindale/goliath-120b), Magnum 72B is the seventh in a family of models designed to achieve the prose quality of the Claude 3 models, notably Opus & Sonnet.\n\nThe model is based on [Qwen2 72B](https://openrouter.ai/models/qwen/qwen-2-72b-instruct) and trained with 55 million tokens of highly curated roleplay (RP) data.", + "modelVersionGroupId": null, + "contextLength": 32768, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Qwen", + "instructType": "chatml", + "defaultSystem": null, + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "anthracite-org/magnum-v2-72b", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "anthracite-org/magnum-v2-72b", + "modelVariantPermaslug": "anthracite-org/magnum-v2-72b", + "providerName": "Infermatic", + "providerInfo": { + "name": "Infermatic", + "displayName": "Infermatic", + "slug": "infermatic", + "baseUrl": "url", + "dataPolicy": { + "privacyPolicyUrl": "https://infermatic.ai/privacy-policy/", + "termsOfServiceUrl": "https://infermatic.ai/terms-and-conditions/", + "paidModels": { + "training": false, + "retainsPrompts": false + } + }, + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": false, + "group": "Infermatic", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": null, + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://infermatic.ai/&size=256" + } + }, + "providerDisplayName": "Infermatic", + "providerModelId": "anthracite-org-magnum-v2-72b-FP8-Dynamic", + "providerGroup": "Infermatic", + "quantization": "fp8", + "variant": "standard", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": null, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "logit_bias", + "top_k", + "min_p", + "seed" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "privacyPolicyUrl": "https://infermatic.ai/privacy-policy/", + "termsOfServiceUrl": "https://infermatic.ai/terms-and-conditions/", + "paidModels": { + "training": false, + "retainsPrompts": false + }, + "training": false, + "retainsPrompts": false + }, + "pricing": { + "prompt": "0.000003", + "completion": "0.000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null + }, + "providerRegion": null + }, + "sorting": { + "topWeekly": 296, + "newest": 195, + "throughputHighToLow": 198, + "latencyLowToHigh": 171, + "pricingLowToHigh": 256, + "pricingHighToLow": 61 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [ + { + "name": "Infermatic", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://infermatic.ai/&size=256", + "slug": "infermatic", + "quantization": "fp8", + "context": 32768, + "maxCompletionTokens": null, + "providerModelId": "anthracite-org-magnum-v2-72b-FP8-Dynamic", + "pricing": { + "prompt": "0.000003", + "completion": "0.000003", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "logit_bias", + "top_k", + "min_p", + "seed" + ], + "inputCost": 3, + "outputCost": 3, + "throughput": 42.939, + "latency": 1068 + }, + { + "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", + "slug": "featherless", + "quantization": "fp8", + "context": 16384, + "maxCompletionTokens": 4096, + "providerModelId": "anthracite-org/magnum-v2-72b", + "pricing": { + "prompt": "0.000004", + "completion": "0.000006", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "repetition_penalty", + "top_k", + "min_p", + "seed" + ], + "inputCost": 4, + "outputCost": 6, + "throughput": 13.8215, + "latency": 6797 + } + ] + }, + { + "slug": "meta-llama/llama-2-70b-chat", + "hfSlug": "meta-llama/Llama-2-70b-chat-hf", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2023-06-20T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Meta: Llama 2 70B Chat", + "shortName": "Llama 2 70B Chat", + "author": "meta-llama", + "description": "The flagship, 70 billion parameter language model from Meta, fine tuned for chat completions. Llama 2 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align to human preferences for helpfulness and safety.", "modelVersionGroupId": null, "contextLength": 4096, "inputModalities": [ @@ -77418,33 +80764,33 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "alpaca", + "group": "Llama2", + "instructType": "llama2", "defaultSystem": null, "defaultStops": [ - "###", - "" + "", + "[INST]" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "undi95/toppy-m-7b", + "permaslug": "meta-llama/llama-2-70b-chat", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "d2b3973b-76cb-43ef-affa-1ca7e1b25ea2", - "name": "Featherless | undi95/toppy-m-7b", + "id": "08b4e0b7-69ca-4c23-8500-ff49a5175285", + "name": "Together | meta-llama/llama-2-70b-chat", "contextLength": 4096, "model": { - "slug": "undi95/toppy-m-7b", - "hfSlug": "Undi95/Toppy-M-7B", + "slug": "meta-llama/llama-2-70b-chat", + "hfSlug": "meta-llama/Llama-2-70b-chat-hf", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-11-10T00:00:00+00:00", + "createdAt": "2023-06-20T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Toppy M 7B", - "shortName": "Toppy M 7B", - "author": "undi95", - "description": "A wild 7B parameter model that merges several models using the new task_arithmetic merge method from mergekit.\nList of merged models:\n- NousResearch/Nous-Capybara-7B-V1.9\n- [HuggingFaceH4/zephyr-7b-beta](/models/huggingfaceh4/zephyr-7b-beta)\n- lemonilia/AshhLimaRP-Mistral-7B\n- Vulkane/120-Days-of-Sodom-LoRA-Mistral-7b\n- Undi95/Mistral-pippa-sharegpt-7b-qlora\n\n#merge #uncensored", + "name": "Meta: Llama 2 70B Chat", + "shortName": "Llama 2 70B Chat", + "author": "meta-llama", + "description": "The flagship, 70 billion parameter language model from Meta, fine tuned for chat completions. Llama 2 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align to human preferences for helpfulness and safety.", "modelVersionGroupId": null, "contextLength": 4096, "inputModalities": [ @@ -77454,31 +80800,31 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "alpaca", + "group": "Llama2", + "instructType": "llama2", "defaultSystem": null, "defaultStops": [ - "###", - "" + "", + "[INST]" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "undi95/toppy-m-7b", + "permaslug": "meta-llama/llama-2-70b-chat", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "undi95/toppy-m-7b", - "modelVariantPermaslug": "undi95/toppy-m-7b", - "providerName": "Featherless", + "modelVariantSlug": "meta-llama/llama-2-70b-chat", + "modelVariantPermaslug": "meta-llama/llama-2-70b-chat", + "providerName": "Together", "providerInfo": { - "name": "Featherless", - "displayName": "Featherless", - "slug": "featherless", + "name": "Together", + "displayName": "Together", + "slug": "together", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -77487,28 +80833,28 @@ export default { "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, + "isAbortable": true, "moderationRequired": false, - "group": "Featherless", + "group": "Together", "editors": [], "owners": [], - "isMultipartSupported": false, + "isMultipartSupported": true, "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" } }, - "providerDisplayName": "Featherless", - "providerModelId": "Undi95/Toppy-M-7B", - "providerGroup": "Featherless", + "providerDisplayName": "Together", + "providerModelId": "meta-llama-llama-2-70b-hf", + "providerGroup": "Together", "quantization": null, "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -77518,16 +80864,17 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", "top_k", + "repetition_penalty", + "logit_bias", "min_p", - "seed" + "response_format" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -77536,8 +80883,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.0000008", - "completion": "0.0000012", + "prompt": "0.0000009", + "completion": "0.0000009", "image": "0", "request": "0", "webSearch": "0", @@ -77550,36 +80897,37 @@ export default { "isDisabled": false, "supportsToolParameters": false, "supportsReasoning": false, - "supportsMultipart": false, + "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, - "hasChatCompletions": false, + "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 296, - "newest": 299, - "throughputHighToLow": 264, - "latencyLowToHigh": 180, - "pricingLowToHigh": 211, - "pricingHighToLow": 113 + "topWeekly": 297, + "newest": 314, + "throughputHighToLow": 183, + "latencyLowToHigh": 82, + "pricingLowToHigh": 215, + "pricingHighToLow": 106 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { - "name": "Featherless", - "slug": "featherless", + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", "quantization": null, "context": 4096, - "maxCompletionTokens": 4096, - "providerModelId": "Undi95/Toppy-M-7B", + "maxCompletionTokens": null, + "providerModelId": "meta-llama-llama-2-70b-hf", "pricing": { - "prompt": "0.0000008", - "completion": "0.0000012", + "prompt": "0.0000009", + "completion": "0.0000009", "image": "0", "request": "0", "webSearch": "0", @@ -77593,28 +80941,31 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", "top_k", + "repetition_penalty", + "logit_bias", "min_p", - "seed" + "response_format" ], - "inputCost": 0.8, - "outputCost": 1.2 + "inputCost": 0.9, + "outputCost": 0.9, + "throughput": 46.022, + "latency": 560 } ] }, { - "slug": "arcee-ai/maestro-reasoning", - "hfSlug": "", - "updatedAt": "2025-05-05T21:48:13.389346+00:00", - "createdAt": "2025-05-05T21:41:09.235957+00:00", + "slug": "all-hands/openhands-lm-32b-v0.1", + "hfSlug": "all-hands/openhands-lm-32b-v0.1", + "updatedAt": "2025-04-02T17:07:18.295736+00:00", + "createdAt": "2025-04-02T16:56:53.893822+00:00", "hfUpdatedAt": null, - "name": "Arcee AI: Maestro Reasoning", - "shortName": "Maestro Reasoning", - "author": "arcee-ai", - "description": "Maestro Reasoning is Arcee's flagship analysis model: a 32 B‑parameter derivative of Qwen 2.5‑32 B tuned with DPO and chain‑of‑thought RL for step‑by‑step logic. Compared to the earlier 7 B preview, the production 32 B release widens the context window to 128 k tokens and doubles pass‑rate on MATH and GSM‑8K, while also lifting code completion accuracy. Its instruction style encourages structured \"thought → answer\" traces that can be parsed or hidden according to user preference. That transparency pairs well with audit‑focused industries like finance or healthcare where seeing the reasoning path matters. In Arcee Conductor, Maestro is automatically selected for complex, multi‑constraint queries that smaller SLMs bounce. ", + "name": "OpenHands LM 32B V0.1", + "shortName": "OpenHands LM 32B V0.1", + "author": "all-hands", + "description": "OpenHands LM v0.1 is a 32B open-source coding model fine-tuned from Qwen2.5-Coder-32B-Instruct using reinforcement learning techniques outlined in SWE-Gym. It is optimized for autonomous software development agents and achieves strong performance on SWE-Bench Verified, with a 37.2% resolve rate. The model supports a 128K token context window, making it well-suited for long-horizon code reasoning and large codebase tasks.\n\nOpenHands LM is designed for local deployment and runs on consumer-grade GPUs such as a single 3090. It enables fully offline agent workflows without dependency on proprietary APIs. This release is intended as a research preview, and future updates aim to improve generalizability, reduce repetition, and offer smaller variants.", "modelVersionGroupId": null, - "contextLength": 131072, + "contextLength": 16384, "inputModalities": [ "text" ], @@ -77629,23 +80980,23 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "arcee-ai/maestro-reasoning", + "permaslug": "all-hands/openhands-lm-32b-v0.1", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "0edd5142-34f6-431d-828b-14c16e184eb0", - "name": "Together | arcee-ai/maestro-reasoning", - "contextLength": 131072, + "id": "3b9c711a-07e8-43d3-a0d7-f387dff982c7", + "name": "Featherless | all-hands/openhands-lm-32b-v0.1", + "contextLength": 16384, "model": { - "slug": "arcee-ai/maestro-reasoning", - "hfSlug": "", - "updatedAt": "2025-05-05T21:48:13.389346+00:00", - "createdAt": "2025-05-05T21:41:09.235957+00:00", + "slug": "all-hands/openhands-lm-32b-v0.1", + "hfSlug": "all-hands/openhands-lm-32b-v0.1", + "updatedAt": "2025-04-02T17:07:18.295736+00:00", + "createdAt": "2025-04-02T16:56:53.893822+00:00", "hfUpdatedAt": null, - "name": "Arcee AI: Maestro Reasoning", - "shortName": "Maestro Reasoning", - "author": "arcee-ai", - "description": "Maestro Reasoning is Arcee's flagship analysis model: a 32 B‑parameter derivative of Qwen 2.5‑32 B tuned with DPO and chain‑of‑thought RL for step‑by‑step logic. Compared to the earlier 7 B preview, the production 32 B release widens the context window to 128 k tokens and doubles pass‑rate on MATH and GSM‑8K, while also lifting code completion accuracy. Its instruction style encourages structured \"thought → answer\" traces that can be parsed or hidden according to user preference. That transparency pairs well with audit‑focused industries like finance or healthcare where seeing the reasoning path matters. In Arcee Conductor, Maestro is automatically selected for complex, multi‑constraint queries that smaller SLMs bounce. ", + "name": "OpenHands LM 32B V0.1", + "shortName": "OpenHands LM 32B V0.1", + "author": "all-hands", + "description": "OpenHands LM v0.1 is a 32B open-source coding model fine-tuned from Qwen2.5-Coder-32B-Instruct using reinforcement learning techniques outlined in SWE-Gym. It is optimized for autonomous software development agents and achieves strong performance on SWE-Bench Verified, with a 37.2% resolve rate. The model supports a 128K token context window, making it well-suited for long-horizon code reasoning and large codebase tasks.\n\nOpenHands LM is designed for local deployment and runs on consumer-grade GPUs such as a single 3090. It enables fully offline agent workflows without dependency on proprietary APIs. This release is intended as a research preview, and future updates aim to improve generalizability, reduce repetition, and offer smaller variants.", "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ @@ -77662,21 +81013,21 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "arcee-ai/maestro-reasoning", + "permaslug": "all-hands/openhands-lm-32b-v0.1", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "arcee-ai/maestro-reasoning", - "modelVariantPermaslug": "arcee-ai/maestro-reasoning", - "providerName": "Together", + "modelVariantSlug": "all-hands/openhands-lm-32b-v0.1", + "modelVariantPermaslug": "all-hands/openhands-lm-32b-v0.1", + "providerName": "Featherless", "providerInfo": { - "name": "Together", - "displayName": "Together", - "slug": "together", + "name": "Featherless", + "displayName": "Featherless", + "slug": "featherless", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -77685,48 +81036,49 @@ export default { "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": true, + "isAbortable": false, "moderationRequired": false, - "group": "Together", + "group": "Featherless", "editors": [], "owners": [], - "isMultipartSupported": true, + "isMultipartSupported": false, "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" } }, - "providerDisplayName": "Together", - "providerModelId": "arcee-ai/maestro-reasoning", - "providerGroup": "Together", + "providerDisplayName": "Featherless", + "providerModelId": "all-hands/openhands-lm-32b-v0.1", + "providerGroup": "Featherless", "quantization": null, "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 32000, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "top_k", "repetition_penalty", - "logit_bias", + "top_k", "min_p", - "response_format" + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -77735,8 +81087,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.0000009", - "completion": "0.0000033", + "prompt": "0.0000026", + "completion": "0.0000034", "image": "0", "request": "0", "webSearch": "0", @@ -77747,9 +81099,9 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, - "supportsMultipart": true, + "supportsMultipart": false, "limitRpm": null, "limitRpd": null, "hasCompletions": true, @@ -77761,24 +81113,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 297, - "newest": 5, - "throughputHighToLow": 59, - "latencyLowToHigh": 55, - "pricingLowToHigh": 226, - "pricingHighToLow": 92 + "topWeekly": 298, + "newest": 65, + "throughputHighToLow": 303, + "latencyLowToHigh": 248, + "pricingLowToHigh": 252, + "pricingHighToLow": 65 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { - "name": "Together", - "slug": "together", + "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", + "slug": "featherless", "quantization": null, - "context": 131072, - "maxCompletionTokens": 32000, - "providerModelId": "arcee-ai/maestro-reasoning", + "context": 16384, + "maxCompletionTokens": 4096, + "providerModelId": "all-hands/openhands-lm-32b-v0.1", "pricing": { - "prompt": "0.0000009", - "completion": "0.0000033", + "prompt": "0.0000026", + "completion": "0.0000034", "image": "0", "request": "0", "webSearch": "0", @@ -77786,162 +81140,160 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "top_k", "repetition_penalty", - "logit_bias", + "top_k", "min_p", - "response_format" + "seed" ], - "inputCost": 0.9, - "outputCost": 3.3 + "inputCost": 2.6, + "outputCost": 3.4, + "throughput": 13.057, + "latency": 1851 } ] }, { - "slug": "x-ai/grok-vision-beta", - "hfSlug": "", + "slug": "anthropic/claude-2.0", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-19T00:37:04.585936+00:00", + "createdAt": "2023-07-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "xAI: Grok Vision Beta", - "shortName": "Grok Vision Beta", - "author": "x-ai", - "description": "Grok Vision Beta is xAI's experimental language model with vision capability.\n\n", + "name": "Anthropic: Claude v2.0", + "shortName": "Claude v2.0", + "author": "anthropic", + "description": "Anthropic's flagship model. Superior performance on tasks that require complex reasoning. Supports hundreds of pages of text.", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 100000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Grok", + "group": "Claude", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "x-ai/grok-vision-beta", + "permaslug": "anthropic/claude-2.0", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "641fa158-ffaa-45e0-a7cd-8d51bef6213b", - "name": "xAI | x-ai/grok-vision-beta", - "contextLength": 8192, + "id": "5a724bb2-64a0-4e69-a04c-bbe4e5d2e391", + "name": "Anthropic | anthropic/claude-2.0", + "contextLength": 100000, "model": { - "slug": "x-ai/grok-vision-beta", - "hfSlug": "", + "slug": "anthropic/claude-2.0", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-19T00:37:04.585936+00:00", + "createdAt": "2023-07-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "xAI: Grok Vision Beta", - "shortName": "Grok Vision Beta", - "author": "x-ai", - "description": "Grok Vision Beta is xAI's experimental language model with vision capability.\n\n", + "name": "Anthropic: Claude v2.0", + "shortName": "Claude v2.0", + "author": "anthropic", + "description": "Anthropic's flagship model. Superior performance on tasks that require complex reasoning. Supports hundreds of pages of text.", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 100000, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Grok", + "group": "Claude", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "x-ai/grok-vision-beta", + "permaslug": "anthropic/claude-2.0", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "x-ai/grok-vision-beta", - "modelVariantPermaslug": "x-ai/grok-vision-beta", - "providerName": "xAI", + "modelVariantSlug": "anthropic/claude-2.0", + "modelVariantPermaslug": "anthropic/claude-2.0", + "providerName": "Anthropic", "providerInfo": { - "name": "xAI", - "displayName": "xAI", - "slug": "xai", + "name": "Anthropic", + "displayName": "Anthropic", + "slug": "anthropic", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://x.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://x.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 - } + }, + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": false, - "group": "xAI", + "moderationRequired": true, + "group": "Anthropic", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.x.ai/", + "statusPageUrl": "https://status.anthropic.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://x.ai/&size=256" + "url": "/images/icons/Anthropic.svg" } }, - "providerDisplayName": "xAI", - "providerModelId": "grok-vision-beta", - "providerGroup": "xAI", - "quantization": null, + "providerDisplayName": "Anthropic", + "providerModelId": "claude-2.0", + "providerGroup": "Anthropic", + "quantization": "unknown", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logprobs", - "top_logprobs", - "response_format" + "top_k", + "stop" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://x.ai/legal/terms-of-service", - "privacyPolicyUrl": "https://x.ai/legal/privacy-policy", + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", "paidModels": { "training": false, "retainsPrompts": true, "retentionDays": 30 }, + "requiresUserIds": true, "training": false, "retainsPrompts": true, "retentionDays": 30 }, "pricing": { - "prompt": "0.000005", - "completion": "0.000015", - "image": "0.009", + "prompt": "0.000008", + "completion": "0.000024", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -77964,25 +81316,27 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 298, - "newest": 170, - "throughputHighToLow": 132, - "latencyLowToHigh": 61, - "pricingLowToHigh": 290, - "pricingHighToLow": 26 + "topWeekly": 299, + "newest": 310, + "throughputHighToLow": 250, + "latencyLowToHigh": 194, + "pricingLowToHigh": 300, + "pricingHighToLow": 21 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { - "name": "xAI", - "slug": "xAi", - "quantization": null, - "context": 8192, - "maxCompletionTokens": null, - "providerModelId": "grok-vision-beta", + "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", + "slug": "anthropic", + "quantization": "unknown", + "context": 100000, + "maxCompletionTokens": 4096, + "providerModelId": "claude-2.0", "pricing": { - "prompt": "0.000005", - "completion": "0.000015", - "image": "0.009", + "prompt": "0.000008", + "completion": "0.000024", + "image": "0", "request": "0", "webSearch": "0", "internalReasoning": "0", @@ -77992,16 +81346,13 @@ export default { "max_tokens", "temperature", "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "seed", - "logprobs", - "top_logprobs", - "response_format" + "top_k", + "stop" ], - "inputCost": 5, - "outputCost": 15 + "inputCost": 8, + "outputCost": 24, + "throughput": 30.145, + "latency": 1285 } ] }, @@ -78168,16 +81519,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 299, + "topWeekly": 300, "newest": 309, - "throughputHighToLow": 162, - "latencyLowToHigh": 131, + "throughputHighToLow": 195, + "latencyLowToHigh": 115, "pricingLowToHigh": 227, "pricingHighToLow": 91 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://mancer.tech/\\u0026size=256", "providers": [ { "name": "Mancer", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", "slug": "mancer", "quantization": "unknown", "context": 8000, @@ -78207,10 +81560,13 @@ export default { "top_a" ], "inputCost": 1.13, - "outputCost": 1.13 + "outputCost": 1.13, + "throughput": 43.797, + "latency": 748 }, { "name": "Mancer (private)", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://mancer.tech/&size=256", "slug": "mancer (private)", "quantization": "unknown", "context": 8000, @@ -78240,208 +81596,24 @@ export default { "top_a" ], "inputCost": 1.5, - "outputCost": 1.5 - } - ] - }, - { - "slug": "ai21/jamba-1.6-mini", - "hfSlug": "ai21labs/AI21-Jamba-Mini-1.6", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-13T22:32:51+00:00", - "hfUpdatedAt": null, - "name": "AI21: Jamba Mini 1.6", - "shortName": "Jamba Mini 1.6", - "author": "ai21", - "description": "AI21 Jamba Mini 1.6 is a hybrid foundation model combining State Space Models (Mamba) with Transformer attention mechanisms. With 12 billion active parameters (52 billion total), this model excels in extremely long-context tasks (up to 256K tokens) and achieves superior inference efficiency, outperforming comparable open models on tasks such as retrieval-augmented generation (RAG) and grounded question answering. Jamba Mini 1.6 supports multilingual tasks across English, Spanish, French, Portuguese, Italian, Dutch, German, Arabic, and Hebrew, along with structured JSON output and tool-use capabilities.\n\nUsage of this model is subject to the [Jamba Open Model License](https://www.ai21.com/licenses/jamba-open-model-license).", - "modelVersionGroupId": "cd1eb031-30bc-4e2e-aa06-3c20f986e5c7", - "contextLength": 256000, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "ai21/jamba-1.6-mini", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "8c1760f3-6c79-4ba7-9d89-6d5168b075d8", - "name": "AI21 | ai21/jamba-1.6-mini", - "contextLength": 256000, - "model": { - "slug": "ai21/jamba-1.6-mini", - "hfSlug": "ai21labs/AI21-Jamba-Mini-1.6", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2025-03-13T22:32:51+00:00", - "hfUpdatedAt": null, - "name": "AI21: Jamba Mini 1.6", - "shortName": "Jamba Mini 1.6", - "author": "ai21", - "description": "AI21 Jamba Mini 1.6 is a hybrid foundation model combining State Space Models (Mamba) with Transformer attention mechanisms. With 12 billion active parameters (52 billion total), this model excels in extremely long-context tasks (up to 256K tokens) and achieves superior inference efficiency, outperforming comparable open models on tasks such as retrieval-augmented generation (RAG) and grounded question answering. Jamba Mini 1.6 supports multilingual tasks across English, Spanish, French, Portuguese, Italian, Dutch, German, Arabic, and Hebrew, along with structured JSON output and tool-use capabilities.\n\nUsage of this model is subject to the [Jamba Open Model License](https://www.ai21.com/licenses/jamba-open-model-license).", - "modelVersionGroupId": "cd1eb031-30bc-4e2e-aa06-3c20f986e5c7", - "contextLength": 256000, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Other", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "ai21/jamba-1.6-mini", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "ai21/jamba-1.6-mini", - "modelVariantPermaslug": "ai21/jamba-1.6-mini", - "providerName": "AI21", - "providerInfo": { - "name": "AI21", - "displayName": "AI21", - "slug": "ai21", - "baseUrl": "url", - "dataPolicy": { - "privacyPolicyUrl": "https://www.ai21.com/privacy-policy/", - "termsOfServiceUrl": "https://www.ai21.com/terms-of-service/", - "paidModels": { - "training": false - } - }, - "headquarters": "IL", - "hasChatCompletions": true, - "hasCompletions": false, - "isAbortable": false, - "moderationRequired": false, - "group": "AI21", - "editors": [], - "owners": [], - "isMultipartSupported": false, - "statusPageUrl": "https://status.ai21.com/", - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://ai21.com/&size=256" - } - }, - "providerDisplayName": "AI21", - "providerModelId": "jamba-mini-1.6", - "providerGroup": "AI21", - "quantization": "bf16", - "variant": "standard", - "isFree": false, - "canAbort": false, - "maxPromptTokens": null, - "maxCompletionTokens": 4096, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "privacyPolicyUrl": "https://www.ai21.com/privacy-policy/", - "termsOfServiceUrl": "https://www.ai21.com/terms-of-service/", - "paidModels": { - "training": false - }, - "training": false - }, - "pricing": { - "prompt": "0.0000002", - "completion": "0.0000004", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": true, - "supportsReasoning": false, - "supportsMultipart": false, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": false, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 300, - "newest": 86, - "throughputHighToLow": 14, - "latencyLowToHigh": 68, - "pricingLowToHigh": 153, - "pricingHighToLow": 165 - }, - "providers": [ - { - "name": "AI21", - "slug": "ai21", - "quantization": "bf16", - "context": 256000, - "maxCompletionTokens": 4096, - "providerModelId": "jamba-mini-1.6", - "pricing": { - "prompt": "0.0000002", - "completion": "0.0000004", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "tools", - "tool_choice", - "max_tokens", - "temperature", - "top_p", - "stop" - ], - "inputCost": 0.2, - "outputCost": 0.4 + "outputCost": 1.5, + "throughput": 44.091, + "latency": 540 } ] }, { - "slug": "allenai/olmo-7b-instruct", - "hfSlug": "allenai/OLMo-7B-Instruct", + "slug": "eva-unit-01/eva-llama-3.33-70b", + "hfSlug": "EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.1", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-05-10T00:00:00+00:00", + "createdAt": "2024-12-16T19:28:23.771236+00:00", "hfUpdatedAt": null, - "name": "OLMo 7B Instruct", - "shortName": "OLMo 7B Instruct", - "author": "allenai", - "description": "OLMo 7B Instruct by the Allen Institute for AI is a model finetuned for question answering. It demonstrates **notable performance** across multiple benchmarks including TruthfulQA and ToxiGen.\n\n**Open Source**: The model, its code, checkpoints, logs are released under the [Apache 2.0 license](https://choosealicense.com/licenses/apache-2.0).\n\n- [Core repo (training, inference, fine-tuning etc.)](https://github.com/allenai/OLMo)\n- [Evaluation code](https://github.com/allenai/OLMo-Eval)\n- [Further fine-tuning code](https://github.com/allenai/open-instruct)\n- [Paper](https://arxiv.org/abs/2402.00838)\n- [Technical blog post](https://blog.allenai.org/olmo-open-language-model-87ccfc95f580)\n- [W&B Logs](https://wandb.ai/ai2-llm/OLMo-7B/reports/OLMo-7B--Vmlldzo2NzQyMzk5)", + "name": "EVA Llama 3.33 70B", + "shortName": "EVA Llama 3.33 70B", + "author": "eva-unit-01", + "description": "EVA Llama 3.33 70b is a roleplay and storywriting specialist model. It is a full-parameter finetune of [Llama-3.3-70B-Instruct](https://openrouter.ai/meta-llama/llama-3.3-70b-instruct) on mixture of synthetic and natural data.\n\nIt uses Celeste 70B 0.1 data mixture, greatly expanding it to improve versatility, creativity and \"flavor\" of the resulting model\n\nThis model was built with Llama by Meta.\n", "modelVersionGroupId": null, - "contextLength": 2048, + "contextLength": 16384, "inputModalities": [ "text" ], @@ -78449,35 +81621,35 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "zephyr", + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, "defaultStops": [ - "<|user|>\n", - "" + "<|eot_id|>", + "<|end_of_text|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "allenai/olmo-7b-instruct", + "permaslug": "eva-unit-01/eva-llama-3.33-70b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "e4f30272-1ce4-40ad-92a9-065c7f7dad69", - "name": "Nebius | allenai/olmo-7b-instruct", - "contextLength": 2048, + "id": "cc5c89d5-5596-48b0-af30-45b1fef9d8c5", + "name": "Featherless | eva-unit-01/eva-llama-3.33-70b", + "contextLength": 16384, "model": { - "slug": "allenai/olmo-7b-instruct", - "hfSlug": "allenai/OLMo-7B-Instruct", + "slug": "eva-unit-01/eva-llama-3.33-70b", + "hfSlug": "EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.1", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-05-10T00:00:00+00:00", + "createdAt": "2024-12-16T19:28:23.771236+00:00", "hfUpdatedAt": null, - "name": "OLMo 7B Instruct", - "shortName": "OLMo 7B Instruct", - "author": "allenai", - "description": "OLMo 7B Instruct by the Allen Institute for AI is a model finetuned for question answering. It demonstrates **notable performance** across multiple benchmarks including TruthfulQA and ToxiGen.\n\n**Open Source**: The model, its code, checkpoints, logs are released under the [Apache 2.0 license](https://choosealicense.com/licenses/apache-2.0).\n\n- [Core repo (training, inference, fine-tuning etc.)](https://github.com/allenai/OLMo)\n- [Evaluation code](https://github.com/allenai/OLMo-Eval)\n- [Further fine-tuning code](https://github.com/allenai/open-instruct)\n- [Paper](https://arxiv.org/abs/2402.00838)\n- [Technical blog post](https://blog.allenai.org/olmo-open-language-model-87ccfc95f580)\n- [W&B Logs](https://wandb.ai/ai2-llm/OLMo-7B/reports/OLMo-7B--Vmlldzo2NzQyMzk5)", + "name": "EVA Llama 3.33 70B", + "shortName": "EVA Llama 3.33 70B", + "author": "eva-unit-01", + "description": "EVA Llama 3.33 70b is a roleplay and storywriting specialist model. It is a full-parameter finetune of [Llama-3.3-70B-Instruct](https://openrouter.ai/meta-llama/llama-3.3-70b-instruct) on mixture of synthetic and natural data.\n\nIt uses Celeste 70B 0.1 data mixture, greatly expanding it to improve versatility, creativity and \"flavor\" of the resulting model\n\nThis model was built with Llama by Meta.\n", "modelVersionGroupId": null, - "contextLength": 2048, + "contextLength": 16384, "inputModalities": [ "text" ], @@ -78485,62 +81657,61 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", - "instructType": "zephyr", + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, "defaultStops": [ - "<|user|>\n", - "" + "<|eot_id|>", + "<|end_of_text|>" ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "allenai/olmo-7b-instruct", + "permaslug": "eva-unit-01/eva-llama-3.33-70b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "allenai/olmo-7b-instruct", - "modelVariantPermaslug": "allenai/olmo-7b-instruct", - "providerName": "Nebius", + "modelVariantSlug": "eva-unit-01/eva-llama-3.33-70b", + "modelVariantPermaslug": "eva-unit-01/eva-llama-3.33-70b", + "providerName": "Featherless", "providerInfo": { - "name": "Nebius", - "displayName": "Nebius AI Studio", - "slug": "nebius", + "name": "Featherless", + "displayName": "Featherless", + "slug": "featherless", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false } }, - "headquarters": "DE", + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": false, "moderationRequired": false, - "group": "Nebius", + "group": "Featherless", "editors": [], "owners": [], - "isMultipartSupported": true, + "isMultipartSupported": false, "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" } }, - "providerDisplayName": "Nebius AI Studio", - "providerModelId": "allenai/OLMo-7B-Instruct-hf", - "providerGroup": "Nebius", - "quantization": null, + "providerDisplayName": "Featherless", + "providerModelId": "EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.0", + "providerGroup": "Featherless", + "quantization": "fp8", "variant": "standard", "isFree": false, "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -78550,17 +81721,16 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", + "repetition_penalty", "top_k", - "logit_bias", - "logprobs", - "top_logprobs" + "min_p", + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", - "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -78569,8 +81739,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.00000008", - "completion": "0.00000024", + "prompt": "0.000004", + "completion": "0.000006", "image": "0", "request": "0", "webSearch": "0", @@ -78583,36 +81753,37 @@ export default { "isDisabled": false, "supportsToolParameters": false, "supportsReasoning": false, - "supportsMultipart": true, + "supportsMultipart": false, "limitRpm": null, "limitRpd": null, "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 301, - "newest": 262, - "throughputHighToLow": 195, - "latencyLowToHigh": 28, - "pricingLowToHigh": 107, - "pricingHighToLow": 210 + "newest": 152, + "throughputHighToLow": 296, + "latencyLowToHigh": 310, + "pricingLowToHigh": 280, + "pricingHighToLow": 35 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { - "name": "Nebius AI Studio", - "slug": "nebiusAiStudio", - "quantization": null, - "context": 2048, - "maxCompletionTokens": null, - "providerModelId": "allenai/OLMo-7B-Instruct-hf", + "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", + "slug": "featherless", + "quantization": "fp8", + "context": 16384, + "maxCompletionTokens": 4096, + "providerModelId": "EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.0", "pricing": { - "prompt": "0.00000008", - "completion": "0.00000024", + "prompt": "0.000004", + "completion": "0.000006", "image": "0", "request": "0", "webSearch": "0", @@ -78626,29 +81797,30 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "seed", + "repetition_penalty", "top_k", - "logit_bias", - "logprobs", - "top_logprobs" + "min_p", + "seed" ], - "inputCost": 0.08, - "outputCost": 0.24 + "inputCost": 4, + "outputCost": 6, + "throughput": 13.858, + "latency": 14196 } ] }, { - "slug": "all-hands/openhands-lm-32b-v0.1", - "hfSlug": "all-hands/openhands-lm-32b-v0.1", - "updatedAt": "2025-04-02T17:07:18.295736+00:00", - "createdAt": "2025-04-02T16:56:53.893822+00:00", + "slug": "arcee-ai/maestro-reasoning", + "hfSlug": "", + "updatedAt": "2025-05-05T21:48:13.389346+00:00", + "createdAt": "2025-05-05T21:41:09.235957+00:00", "hfUpdatedAt": null, - "name": "OpenHands LM 32B V0.1", - "shortName": "OpenHands LM 32B V0.1", - "author": "all-hands", - "description": "OpenHands LM v0.1 is a 32B open-source coding model fine-tuned from Qwen2.5-Coder-32B-Instruct using reinforcement learning techniques outlined in SWE-Gym. It is optimized for autonomous software development agents and achieves strong performance on SWE-Bench Verified, with a 37.2% resolve rate. The model supports a 128K token context window, making it well-suited for long-horizon code reasoning and large codebase tasks.\n\nOpenHands LM is designed for local deployment and runs on consumer-grade GPUs such as a single 3090. It enables fully offline agent workflows without dependency on proprietary APIs. This release is intended as a research preview, and future updates aim to improve generalizability, reduce repetition, and offer smaller variants.", + "name": "Arcee AI: Maestro Reasoning", + "shortName": "Maestro Reasoning", + "author": "arcee-ai", + "description": "Maestro Reasoning is Arcee's flagship analysis model: a 32 B‑parameter derivative of Qwen 2.5‑32 B tuned with DPO and chain‑of‑thought RL for step‑by‑step logic. Compared to the earlier 7 B preview, the production 32 B release widens the context window to 128 k tokens and doubles pass‑rate on MATH and GSM‑8K, while also lifting code completion accuracy. Its instruction style encourages structured \"thought → answer\" traces that can be parsed or hidden according to user preference. That transparency pairs well with audit‑focused industries like finance or healthcare where seeing the reasoning path matters. In Arcee Conductor, Maestro is automatically selected for complex, multi‑constraint queries that smaller SLMs bounce. ", "modelVersionGroupId": null, - "contextLength": 16384, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -78663,23 +81835,23 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "all-hands/openhands-lm-32b-v0.1", + "permaslug": "arcee-ai/maestro-reasoning", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "3b9c711a-07e8-43d3-a0d7-f387dff982c7", - "name": "Featherless | all-hands/openhands-lm-32b-v0.1", - "contextLength": 16384, + "id": "0edd5142-34f6-431d-828b-14c16e184eb0", + "name": "Together | arcee-ai/maestro-reasoning", + "contextLength": 131072, "model": { - "slug": "all-hands/openhands-lm-32b-v0.1", - "hfSlug": "all-hands/openhands-lm-32b-v0.1", - "updatedAt": "2025-04-02T17:07:18.295736+00:00", - "createdAt": "2025-04-02T16:56:53.893822+00:00", + "slug": "arcee-ai/maestro-reasoning", + "hfSlug": "", + "updatedAt": "2025-05-05T21:48:13.389346+00:00", + "createdAt": "2025-05-05T21:41:09.235957+00:00", "hfUpdatedAt": null, - "name": "OpenHands LM 32B V0.1", - "shortName": "OpenHands LM 32B V0.1", - "author": "all-hands", - "description": "OpenHands LM v0.1 is a 32B open-source coding model fine-tuned from Qwen2.5-Coder-32B-Instruct using reinforcement learning techniques outlined in SWE-Gym. It is optimized for autonomous software development agents and achieves strong performance on SWE-Bench Verified, with a 37.2% resolve rate. The model supports a 128K token context window, making it well-suited for long-horizon code reasoning and large codebase tasks.\n\nOpenHands LM is designed for local deployment and runs on consumer-grade GPUs such as a single 3090. It enables fully offline agent workflows without dependency on proprietary APIs. This release is intended as a research preview, and future updates aim to improve generalizability, reduce repetition, and offer smaller variants.", + "name": "Arcee AI: Maestro Reasoning", + "shortName": "Maestro Reasoning", + "author": "arcee-ai", + "description": "Maestro Reasoning is Arcee's flagship analysis model: a 32 B‑parameter derivative of Qwen 2.5‑32 B tuned with DPO and chain‑of‑thought RL for step‑by‑step logic. Compared to the earlier 7 B preview, the production 32 B release widens the context window to 128 k tokens and doubles pass‑rate on MATH and GSM‑8K, while also lifting code completion accuracy. Its instruction style encourages structured \"thought → answer\" traces that can be parsed or hidden according to user preference. That transparency pairs well with audit‑focused industries like finance or healthcare where seeing the reasoning path matters. In Arcee Conductor, Maestro is automatically selected for complex, multi‑constraint queries that smaller SLMs bounce. ", "modelVersionGroupId": null, "contextLength": 131072, "inputModalities": [ @@ -78696,21 +81868,21 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "all-hands/openhands-lm-32b-v0.1", + "permaslug": "arcee-ai/maestro-reasoning", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "all-hands/openhands-lm-32b-v0.1", - "modelVariantPermaslug": "all-hands/openhands-lm-32b-v0.1", - "providerName": "Featherless", + "modelVariantSlug": "arcee-ai/maestro-reasoning", + "modelVariantPermaslug": "arcee-ai/maestro-reasoning", + "providerName": "Together", "providerInfo": { - "name": "Featherless", - "displayName": "Featherless", - "slug": "featherless", + "name": "Together", + "displayName": "Together", + "slug": "together", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -78719,49 +81891,48 @@ export default { "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": false, + "isAbortable": true, "moderationRequired": false, - "group": "Featherless", + "group": "Together", "editors": [], "owners": [], - "isMultipartSupported": false, + "isMultipartSupported": true, "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" } }, - "providerDisplayName": "Featherless", - "providerModelId": "all-hands/openhands-lm-32b-v0.1", - "providerGroup": "Featherless", + "providerDisplayName": "Together", + "providerModelId": "arcee-ai/maestro-reasoning", + "providerGroup": "Together", "quantization": null, "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": 32000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", "top_k", + "repetition_penalty", + "logit_bias", "min_p", - "seed" + "response_format" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -78770,8 +81941,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.0000026", - "completion": "0.0000034", + "prompt": "0.0000009", + "completion": "0.0000033", "image": "0", "request": "0", "webSearch": "0", @@ -78782,9 +81953,9 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, - "supportsMultipart": false, + "supportsMultipart": true, "limitRpm": null, "limitRpd": null, "hasCompletions": true, @@ -78797,23 +81968,25 @@ export default { }, "sorting": { "topWeekly": 302, - "newest": 65, - "throughputHighToLow": 304, - "latencyLowToHigh": 244, - "pricingLowToHigh": 252, - "pricingHighToLow": 65 + "newest": 5, + "throughputHighToLow": 65, + "latencyLowToHigh": 70, + "pricingLowToHigh": 226, + "pricingHighToLow": 92 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://arcee.ai/\\u0026size=256", "providers": [ { - "name": "Featherless", - "slug": "featherless", + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", "quantization": null, - "context": 16384, - "maxCompletionTokens": 4096, - "providerModelId": "all-hands/openhands-lm-32b-v0.1", + "context": 131072, + "maxCompletionTokens": 32000, + "providerModelId": "arcee-ai/maestro-reasoning", "pricing": { - "prompt": "0.0000026", - "completion": "0.0000034", + "prompt": "0.0000009", + "completion": "0.0000033", "image": "0", "request": "0", "webSearch": "0", @@ -78821,36 +81994,37 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", "top_k", + "repetition_penalty", + "logit_bias", "min_p", - "seed" + "response_format" ], - "inputCost": 2.6, - "outputCost": 3.4 + "inputCost": 0.9, + "outputCost": 3.3, + "throughput": 113.949, + "latency": 499.5 } ] }, { - "slug": "anthropic/claude-2.1", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-11-22T00:00:00+00:00", + "slug": "arcee-ai/virtuoso-medium-v2", + "hfSlug": "arcee-ai/Virtuoso-Medium-v2", + "updatedAt": "2025-05-05T21:48:27.917466+00:00", + "createdAt": "2025-05-05T20:53:54.621827+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude v2.1 (self-moderated)", - "shortName": "Claude v2.1 (self-moderated)", - "author": "anthropic", - "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.", + "name": "Arcee AI: Virtuoso Medium V2", + "shortName": "Virtuoso Medium V2", + "author": "arcee-ai", + "description": "Virtuoso‑Medium‑v2 is a 32 B model distilled from DeepSeek‑v3 logits and merged back onto a Qwen 2.5 backbone, yielding a sharper, more factual successor to the original Virtuoso Medium. The team harvested ~1.1 B logit tokens and applied \"fusion‑merging\" plus DPO alignment, which pushed scores past Arcee‑Nova 2024 and many 40 B‑plus peers on MMLU‑Pro, MATH and HumanEval. With a 128 k context and aggressive quantization options (from BF16 down to 4‑bit GGUF), it balances capability with deployability on single‑GPU nodes. Typical use cases include enterprise chat assistants, technical writing aids and medium‑complexity code drafting where Virtuoso‑Large would be overkill. ", "modelVersionGroupId": null, - "contextLength": 200000, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -78858,32 +82032,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Claude", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-2.1", + "permaslug": "arcee-ai/virtuoso-medium-v2", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "7159a751-fb9d-4750-918b-a774fd71cfa3", - "name": "Anthropic | anthropic/claude-2.1:beta", - "contextLength": 200000, + "id": "b3494a2a-b7d6-4055-a379-d77711c85c7b", + "name": "Together | arcee-ai/virtuoso-medium-v2", + "contextLength": 131072, "model": { - "slug": "anthropic/claude-2.1", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-11-22T00:00:00+00:00", + "slug": "arcee-ai/virtuoso-medium-v2", + "hfSlug": "arcee-ai/Virtuoso-Medium-v2", + "updatedAt": "2025-05-05T21:48:27.917466+00:00", + "createdAt": "2025-05-05T20:53:54.621827+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude v2.1", - "shortName": "Claude v2.1", - "author": "anthropic", - "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.", + "name": "Arcee AI: Virtuoso Medium V2", + "shortName": "Virtuoso Medium V2", + "author": "arcee-ai", + "description": "Virtuoso‑Medium‑v2 is a 32 B model distilled from DeepSeek‑v3 logits and merged back onto a Qwen 2.5 backbone, yielding a sharper, more factual successor to the original Virtuoso Medium. The team harvested ~1.1 B logit tokens and applied \"fusion‑merging\" plus DPO alignment, which pushed scores past Arcee‑Nova 2024 and many 40 B‑plus peers on MMLU‑Pro, MATH and HumanEval. With a 128 k context and aggressive quantization options (from BF16 down to 4‑bit GGUF), it balances capability with deployability on single‑GPU nodes. Typical use cases include enterprise chat assistants, technical writing aids and medium‑complexity code drafting where Virtuoso‑Large would be overkill. ", "modelVersionGroupId": null, - "contextLength": 200000, + "contextLength": 131072, "inputModalities": [ "text" ], @@ -78891,87 +82065,88 @@ export default { "text" ], "hasTextOutput": true, - "group": "Claude", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-2.1", + "permaslug": "arcee-ai/virtuoso-medium-v2", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "anthropic/claude-2.1:beta", - "modelVariantPermaslug": "anthropic/claude-2.1:beta", - "providerName": "Anthropic", + "modelVariantSlug": "arcee-ai/virtuoso-medium-v2", + "modelVariantPermaslug": "arcee-ai/virtuoso-medium-v2", + "providerName": "Together", "providerInfo": { - "name": "Anthropic", - "displayName": "Anthropic", - "slug": "anthropic", + "name": "Together", + "displayName": "Together", + "slug": "together", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "retainsPrompts": false + } }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "Anthropic", + "moderationRequired": false, + "group": "Together", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.anthropic.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Anthropic.svg" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" } }, - "providerDisplayName": "Anthropic", - "providerModelId": "claude-2.1", - "providerGroup": "Anthropic", - "quantization": "unknown", - "variant": "beta", + "providerDisplayName": "Together", + "providerModelId": "arcee-ai/virtuoso-medium-v2", + "providerGroup": "Together", + "quantization": null, + "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": 32768, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "top_k", - "stop" + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, - "requiresUserIds": true, "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, "pricing": { - "prompt": "0.000008", - "completion": "0.000024", + "prompt": "0.0000005", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -78990,29 +82165,32 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 267, - "newest": 296, - "throughputHighToLow": 291, - "latencyLowToHigh": 199, - "pricingLowToHigh": 298, - "pricingHighToLow": 17 + "topWeekly": 303, + "newest": 8, + "throughputHighToLow": 142, + "latencyLowToHigh": 99, + "pricingLowToHigh": 183, + "pricingHighToLow": 135 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://arcee.ai/\\u0026size=256", "providers": [ { - "name": "Anthropic", - "slug": "anthropic", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 4096, - "providerModelId": "claude-2.1", + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": null, + "context": 131072, + "maxCompletionTokens": 32768, + "providerModelId": "arcee-ai/virtuoso-medium-v2", "pricing": { - "prompt": "0.000008", - "completion": "0.000024", + "prompt": "0.0000005", + "completion": "0.0000008", "image": "0", "request": "0", "webSearch": "0", @@ -79023,26 +82201,34 @@ export default { "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", "top_k", - "stop" + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" ], - "inputCost": 8, - "outputCost": 24 + "inputCost": 0.5, + "outputCost": 0.8, + "throughput": 58.564, + "latency": 641 } ] }, { - "slug": "anthropic/claude-2.0", - "hfSlug": null, + "slug": "allenai/olmo-7b-instruct", + "hfSlug": "allenai/OLMo-7B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-07-28T00:00:00+00:00", + "createdAt": "2024-05-10T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude v2.0", - "shortName": "Claude v2.0", - "author": "anthropic", - "description": "Anthropic's flagship model. Superior performance on tasks that require complex reasoning. Supports hundreds of pages of text.", + "name": "OLMo 7B Instruct", + "shortName": "OLMo 7B Instruct", + "author": "allenai", + "description": "OLMo 7B Instruct by the Allen Institute for AI is a model finetuned for question answering. It demonstrates **notable performance** across multiple benchmarks including TruthfulQA and ToxiGen.\n\n**Open Source**: The model, its code, checkpoints, logs are released under the [Apache 2.0 license](https://choosealicense.com/licenses/apache-2.0).\n\n- [Core repo (training, inference, fine-tuning etc.)](https://github.com/allenai/OLMo)\n- [Evaluation code](https://github.com/allenai/OLMo-Eval)\n- [Further fine-tuning code](https://github.com/allenai/open-instruct)\n- [Paper](https://arxiv.org/abs/2402.00838)\n- [Technical blog post](https://blog.allenai.org/olmo-open-language-model-87ccfc95f580)\n- [W&B Logs](https://wandb.ai/ai2-llm/OLMo-7B/reports/OLMo-7B--Vmlldzo2NzQyMzk5)", "modelVersionGroupId": null, - "contextLength": 100000, + "contextLength": 2048, "inputModalities": [ "text" ], @@ -79050,32 +82236,35 @@ export default { "text" ], "hasTextOutput": true, - "group": "Claude", - "instructType": null, + "group": "Other", + "instructType": "zephyr", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|user|>\n", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-2.0", + "permaslug": "allenai/olmo-7b-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "5a724bb2-64a0-4e69-a04c-bbe4e5d2e391", - "name": "Anthropic | anthropic/claude-2.0", - "contextLength": 100000, + "id": "e4f30272-1ce4-40ad-92a9-065c7f7dad69", + "name": "Nebius | allenai/olmo-7b-instruct", + "contextLength": 2048, "model": { - "slug": "anthropic/claude-2.0", - "hfSlug": null, + "slug": "allenai/olmo-7b-instruct", + "hfSlug": "allenai/OLMo-7B-Instruct", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-07-28T00:00:00+00:00", + "createdAt": "2024-05-10T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude v2.0", - "shortName": "Claude v2.0", - "author": "anthropic", - "description": "Anthropic's flagship model. Superior performance on tasks that require complex reasoning. Supports hundreds of pages of text.", + "name": "OLMo 7B Instruct", + "shortName": "OLMo 7B Instruct", + "author": "allenai", + "description": "OLMo 7B Instruct by the Allen Institute for AI is a model finetuned for question answering. It demonstrates **notable performance** across multiple benchmarks including TruthfulQA and ToxiGen.\n\n**Open Source**: The model, its code, checkpoints, logs are released under the [Apache 2.0 license](https://choosealicense.com/licenses/apache-2.0).\n\n- [Core repo (training, inference, fine-tuning etc.)](https://github.com/allenai/OLMo)\n- [Evaluation code](https://github.com/allenai/OLMo-Eval)\n- [Further fine-tuning code](https://github.com/allenai/open-instruct)\n- [Paper](https://arxiv.org/abs/2402.00838)\n- [Technical blog post](https://blog.allenai.org/olmo-open-language-model-87ccfc95f580)\n- [W&B Logs](https://wandb.ai/ai2-llm/OLMo-7B/reports/OLMo-7B--Vmlldzo2NzQyMzk5)", "modelVersionGroupId": null, - "contextLength": 100000, + "contextLength": 2048, "inputModalities": [ "text" ], @@ -79083,87 +82272,92 @@ export default { "text" ], "hasTextOutput": true, - "group": "Claude", - "instructType": null, + "group": "Other", + "instructType": "zephyr", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|user|>\n", + "" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-2.0", + "permaslug": "allenai/olmo-7b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "anthropic/claude-2.0", - "modelVariantPermaslug": "anthropic/claude-2.0", - "providerName": "Anthropic", + "modelVariantSlug": "allenai/olmo-7b-instruct", + "modelVariantPermaslug": "allenai/olmo-7b-instruct", + "providerName": "Nebius", "providerInfo": { - "name": "Anthropic", - "displayName": "Anthropic", - "slug": "anthropic", + "name": "Nebius", + "displayName": "Nebius AI Studio", + "slug": "nebius", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "retainsPrompts": false + } }, - "headquarters": "US", + "headquarters": "DE", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": true, - "moderationRequired": true, - "group": "Anthropic", + "isAbortable": false, + "moderationRequired": false, + "group": "Nebius", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.anthropic.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Anthropic.svg" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "invertRequired": true } }, - "providerDisplayName": "Anthropic", - "providerModelId": "claude-2.0", - "providerGroup": "Anthropic", - "quantization": "unknown", + "providerDisplayName": "Nebius AI Studio", + "providerModelId": "allenai/OLMo-7B-Instruct-hf", + "providerGroup": "Nebius", + "quantization": null, "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", "top_k", - "stop" + "logit_bias", + "logprobs", + "top_logprobs" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "termsOfServiceUrl": "https://docs.nebius.com/legal/studio/terms-of-use/", + "privacyPolicyUrl": "https://docs.nebius.com/legal/studio/privacy/", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, - "requiresUserIds": true, "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, "pricing": { - "prompt": "0.000008", - "completion": "0.000024", + "prompt": "0.00000008", + "completion": "0.00000024", "image": "0", "request": "0", "webSearch": "0", @@ -79182,29 +82376,32 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { "topWeekly": 304, - "newest": 310, - "throughputHighToLow": 240, - "latencyLowToHigh": 182, - "pricingLowToHigh": 300, - "pricingHighToLow": 21 + "newest": 262, + "throughputHighToLow": 166, + "latencyLowToHigh": 13, + "pricingLowToHigh": 108, + "pricingHighToLow": 209 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://allenai.org/\\u0026size=256", "providers": [ { - "name": "Anthropic", - "slug": "anthropic", - "quantization": "unknown", - "context": 100000, - "maxCompletionTokens": 4096, - "providerModelId": "claude-2.0", + "name": "Nebius AI Studio", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://docs.nebius.com/&size=256", + "slug": "nebiusAiStudio", + "quantization": null, + "context": 2048, + "maxCompletionTokens": null, + "providerModelId": "allenai/OLMo-7B-Instruct-hf", "pricing": { - "prompt": "0.000008", - "completion": "0.000024", + "prompt": "0.00000008", + "completion": "0.00000024", "image": "0", "request": "0", "webSearch": "0", @@ -79215,26 +82412,34 @@ export default { "max_tokens", "temperature", "top_p", + "stop", + "frequency_penalty", + "presence_penalty", + "seed", "top_k", - "stop" + "logit_bias", + "logprobs", + "top_logprobs" ], - "inputCost": 8, - "outputCost": 24 + "inputCost": 0.08, + "outputCost": 0.24, + "throughput": 50.721, + "latency": 239 } ] }, { - "slug": "anthracite-org/magnum-v2-72b", - "hfSlug": "anthracite-org/magnum-v2-72b", + "slug": "alpindale/magnum-72b", + "hfSlug": "alpindale/magnum-72b-v1", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-30T00:00:00+00:00", + "createdAt": "2024-07-11T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Magnum v2 72B", - "shortName": "Magnum v2 72B", - "author": "anthracite-org", - "description": "From the maker of [Goliath](https://openrouter.ai/models/alpindale/goliath-120b), Magnum 72B is the seventh in a family of models designed to achieve the prose quality of the Claude 3 models, notably Opus & Sonnet.\n\nThe model is based on [Qwen2 72B](https://openrouter.ai/models/qwen/qwen-2-72b-instruct) and trained with 55 million tokens of highly curated roleplay (RP) data.", + "name": "Magnum 72B", + "shortName": "Magnum 72B", + "author": "alpindale", + "description": "From the maker of [Goliath](https://openrouter.ai/models/alpindale/goliath-120b), Magnum 72B is the first in a new family of models designed to achieve the prose quality of the Claude 3 models, notably Opus & Sonnet.\n\nThe model is based on [Qwen2 72B](https://openrouter.ai/models/qwen/qwen-2-72b-instruct) and trained with 55 million tokens of highly curated roleplay (RP) data.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 16384, "inputModalities": [ "text" ], @@ -79253,25 +82458,25 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthracite-org/magnum-v2-72b", + "permaslug": "alpindale/magnum-72b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "740509db-9804-483c-9004-65a3884d7be7", - "name": "Infermatic | anthracite-org/magnum-v2-72b", - "contextLength": 32768, + "id": "3a1f8d4e-f9ef-44ae-a9a2-9a58bb03e7dc", + "name": "Featherless | alpindale/magnum-72b", + "contextLength": 16384, "model": { - "slug": "anthracite-org/magnum-v2-72b", - "hfSlug": "anthracite-org/magnum-v2-72b", + "slug": "alpindale/magnum-72b", + "hfSlug": "alpindale/magnum-72b-v1", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-09-30T00:00:00+00:00", + "createdAt": "2024-07-11T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Magnum v2 72B", - "shortName": "Magnum v2 72B", - "author": "anthracite-org", - "description": "From the maker of [Goliath](https://openrouter.ai/models/alpindale/goliath-120b), Magnum 72B is the seventh in a family of models designed to achieve the prose quality of the Claude 3 models, notably Opus & Sonnet.\n\nThe model is based on [Qwen2 72B](https://openrouter.ai/models/qwen/qwen-2-72b-instruct) and trained with 55 million tokens of highly curated roleplay (RP) data.", + "name": "Magnum 72B", + "shortName": "Magnum 72B", + "author": "alpindale", + "description": "From the maker of [Goliath](https://openrouter.ai/models/alpindale/goliath-120b), Magnum 72B is the first in a new family of models designed to achieve the prose quality of the Claude 3 models, notably Opus & Sonnet.\n\nThe model is based on [Qwen2 72B](https://openrouter.ai/models/qwen/qwen-2-72b-instruct) and trained with 55 million tokens of highly curated roleplay (RP) data.", "modelVersionGroupId": null, - "contextLength": 32768, + "contextLength": 16384, "inputModalities": [ "text" ], @@ -79290,50 +82495,51 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthracite-org/magnum-v2-72b", + "permaslug": "alpindale/magnum-72b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "anthracite-org/magnum-v2-72b", - "modelVariantPermaslug": "anthracite-org/magnum-v2-72b", - "providerName": "Infermatic", + "modelVariantSlug": "alpindale/magnum-72b", + "modelVariantPermaslug": "alpindale/magnum-72b", + "providerName": "Featherless", "providerInfo": { - "name": "Infermatic", - "displayName": "Infermatic", - "slug": "infermatic", + "name": "Featherless", + "displayName": "Featherless", + "slug": "featherless", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://infermatic.ai/privacy-policy/", - "termsOfServiceUrl": "https://infermatic.ai/terms-and-conditions/", + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false } }, + "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": true, + "isAbortable": false, "moderationRequired": false, - "group": "Infermatic", + "group": "Featherless", "editors": [], "owners": [], - "isMultipartSupported": true, + "isMultipartSupported": false, "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://infermatic.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" } }, - "providerDisplayName": "Infermatic", - "providerModelId": "anthracite-org-magnum-v2-72b-FP8-Dynamic", - "providerGroup": "Infermatic", + "providerDisplayName": "Featherless", + "providerModelId": "anthracite-org/magnum-v1-72b", + "providerGroup": "Featherless", "quantization": "fp8", "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -79344,7 +82550,6 @@ export default { "frequency_penalty", "presence_penalty", "repetition_penalty", - "logit_bias", "top_k", "min_p", "seed" @@ -79352,8 +82557,8 @@ export default { "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://infermatic.ai/privacy-policy/", - "termsOfServiceUrl": "https://infermatic.ai/terms-and-conditions/", + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -79362,8 +82567,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.000003", - "completion": "0.000003", + "prompt": "0.000004", + "completion": "0.000006", "image": "0", "request": "0", "webSearch": "0", @@ -79376,7 +82581,7 @@ export default { "isDisabled": false, "supportsToolParameters": false, "supportsReasoning": false, - "supportsMultipart": true, + "supportsMultipart": false, "limitRpm": null, "limitRpd": null, "hasCompletions": true, @@ -79388,52 +82593,22 @@ export default { }, "sorting": { "topWeekly": 305, - "newest": 196, - "throughputHighToLow": 194, - "latencyLowToHigh": 151, - "pricingLowToHigh": 256, - "pricingHighToLow": 61 + "newest": 239, + "throughputHighToLow": 300, + "latencyLowToHigh": 307, + "pricingLowToHigh": 282, + "pricingHighToLow": 37 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ - { - "name": "Infermatic", - "slug": "infermatic", - "quantization": "fp8", - "context": 32768, - "maxCompletionTokens": null, - "providerModelId": "anthracite-org-magnum-v2-72b-FP8-Dynamic", - "pricing": { - "prompt": "0.000003", - "completion": "0.000003", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "logit_bias", - "top_k", - "min_p", - "seed" - ], - "inputCost": 3, - "outputCost": 3 - }, { "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", "slug": "featherless", "quantization": "fp8", "context": 16384, "maxCompletionTokens": 4096, - "providerModelId": "anthracite-org/magnum-v2-72b", + "providerModelId": "anthracite-org/magnum-v1-72b", "pricing": { "prompt": "0.000004", "completion": "0.000006", @@ -79456,7 +82631,9 @@ export default { "seed" ], "inputCost": 4, - "outputCost": 6 + "outputCost": 6, + "throughput": 13.7415, + "latency": 9707.5 } ] }, @@ -79616,16 +82793,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 304, + "topWeekly": 299, "newest": 310, - "throughputHighToLow": 240, - "latencyLowToHigh": 182, + "throughputHighToLow": 250, + "latencyLowToHigh": 194, "pricingLowToHigh": 300, "pricingHighToLow": 21 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [ { "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", "slug": "anthropic", "quantization": "unknown", "context": 100000, @@ -79648,286 +82827,79 @@ export default { "stop" ], "inputCost": 8, - "outputCost": 24 - } - ] - }, - { - "slug": "alpindale/magnum-72b", - "hfSlug": "alpindale/magnum-72b-v1", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-07-11T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Magnum 72B", - "shortName": "Magnum 72B", - "author": "alpindale", - "description": "From the maker of [Goliath](https://openrouter.ai/models/alpindale/goliath-120b), Magnum 72B is the first in a new family of models designed to achieve the prose quality of the Claude 3 models, notably Opus & Sonnet.\n\nThe model is based on [Qwen2 72B](https://openrouter.ai/models/qwen/qwen-2-72b-instruct) and trained with 55 million tokens of highly curated roleplay (RP) data.", - "modelVersionGroupId": null, - "contextLength": 16384, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", - "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "alpindale/magnum-72b", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "3a1f8d4e-f9ef-44ae-a9a2-9a58bb03e7dc", - "name": "Featherless | alpindale/magnum-72b", - "contextLength": 16384, - "model": { - "slug": "alpindale/magnum-72b", - "hfSlug": "alpindale/magnum-72b-v1", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-07-11T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Magnum 72B", - "shortName": "Magnum 72B", - "author": "alpindale", - "description": "From the maker of [Goliath](https://openrouter.ai/models/alpindale/goliath-120b), Magnum 72B is the first in a new family of models designed to achieve the prose quality of the Claude 3 models, notably Opus & Sonnet.\n\nThe model is based on [Qwen2 72B](https://openrouter.ai/models/qwen/qwen-2-72b-instruct) and trained with 55 million tokens of highly curated roleplay (RP) data.", - "modelVersionGroupId": null, - "contextLength": 16384, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", - "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "alpindale/magnum-72b", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "alpindale/magnum-72b", - "modelVariantPermaslug": "alpindale/magnum-72b", - "providerName": "Featherless", - "providerInfo": { - "name": "Featherless", - "displayName": "Featherless", - "slug": "featherless", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", - "paidModels": { - "training": false, - "retainsPrompts": false - } - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": false, - "moderationRequired": false, - "group": "Featherless", - "editors": [], - "owners": [], - "isMultipartSupported": false, - "statusPageUrl": null, - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" - } - }, - "providerDisplayName": "Featherless", - "providerModelId": "anthracite-org/magnum-v1-72b", - "providerGroup": "Featherless", - "quantization": "fp8", - "variant": "standard", - "isFree": false, - "canAbort": false, - "maxPromptTokens": null, - "maxCompletionTokens": 4096, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "top_k", - "min_p", - "seed" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", - "paidModels": { - "training": false, - "retainsPrompts": false - }, - "training": false, - "retainsPrompts": false - }, - "pricing": { - "prompt": "0.000004", - "completion": "0.000006", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": false, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 307, - "newest": 239, - "throughputHighToLow": 299, - "latencyLowToHigh": 311, - "pricingLowToHigh": 282, - "pricingHighToLow": 37 - }, - "providers": [ - { - "name": "Featherless", - "slug": "featherless", - "quantization": "fp8", - "context": 16384, - "maxCompletionTokens": 4096, - "providerModelId": "anthracite-org/magnum-v1-72b", - "pricing": { - "prompt": "0.000004", - "completion": "0.000006", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "stop", - "frequency_penalty", - "presence_penalty", - "repetition_penalty", - "top_k", - "min_p", - "seed" - ], - "inputCost": 4, - "outputCost": 6 + "outputCost": 24, + "throughput": 30.145, + "latency": 1285 } ] }, { - "slug": "scb10x/llama3.1-typhoon2-8b-instruct", - "hfSlug": "scb10x/llama3.1-typhoon2-8b-instruct", - "updatedAt": "2025-03-28T21:16:32.49476+00:00", - "createdAt": "2025-03-28T21:15:11.257967+00:00", + "slug": "arcee-ai/spotlight", + "hfSlug": "", + "updatedAt": "2025-05-05T21:47:17.011692+00:00", + "createdAt": "2025-05-05T21:45:52.249082+00:00", "hfUpdatedAt": null, - "name": "Typhoon2 8B Instruct", - "shortName": "Typhoon2 8B Instruct", - "author": "scb10x", - "description": "Llama3.1-Typhoon2-8B-Instruct is a Thai-English instruction-tuned model with 8 billion parameters, built on Llama 3.1. It significantly improves over its base model in Thai reasoning, instruction-following, and function-calling tasks, while maintaining competitive English performance. The model is optimized for bilingual interaction and performs well on Thai-English code-switching, MT-Bench, IFEval, and tool-use benchmarks.\n\nDespite its smaller size, it demonstrates strong generalization across math, coding, and multilingual benchmarks, outperforming comparable 8B models across most Thai-specific tasks. Full benchmark results and methodology are available in the [technical report.](https://arxiv.org/abs/2412.13702)", + "name": "Arcee AI: Spotlight", + "shortName": "Spotlight", + "author": "arcee-ai", + "description": "Spotlight is a 7‑billion‑parameter vision‑language model derived from Qwen 2.5‑VL and fine‑tuned by Arcee AI for tight image‑text grounding tasks. It offers a 32 k‑token context window, enabling rich multimodal conversations that combine lengthy documents with one or more images. Training emphasized fast inference on consumer GPUs while retaining strong captioning, visual‐question‑answering, and diagram‑analysis accuracy. As a result, Spotlight slots neatly into agent workflows where screenshots, charts or UI mock‑ups need to be interpreted on the fly. Early benchmarks show it matching or out‑scoring larger VLMs such as LLaVA‑1.6 13 B on popular VQA and POPE alignment tests. ", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 131072, "inputModalities": [ + "image", "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "scb10x/llama3.1-typhoon2-8b-instruct", + "permaslug": "arcee-ai/spotlight", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "cd687004-e160-49f4-b6a1-0f6ab950f046", - "name": "Together | scb10x/llama3.1-typhoon2-8b-instruct", - "contextLength": 8192, + "id": "a9b3fe6f-e21f-4f3c-9ea7-f70d856939d6", + "name": "Together | arcee-ai/spotlight", + "contextLength": 131072, "model": { - "slug": "scb10x/llama3.1-typhoon2-8b-instruct", - "hfSlug": "scb10x/llama3.1-typhoon2-8b-instruct", - "updatedAt": "2025-03-28T21:16:32.49476+00:00", - "createdAt": "2025-03-28T21:15:11.257967+00:00", + "slug": "arcee-ai/spotlight", + "hfSlug": "", + "updatedAt": "2025-05-05T21:47:17.011692+00:00", + "createdAt": "2025-05-05T21:45:52.249082+00:00", "hfUpdatedAt": null, - "name": "Typhoon2 8B Instruct", - "shortName": "Typhoon2 8B Instruct", - "author": "scb10x", - "description": "Llama3.1-Typhoon2-8B-Instruct is a Thai-English instruction-tuned model with 8 billion parameters, built on Llama 3.1. It significantly improves over its base model in Thai reasoning, instruction-following, and function-calling tasks, while maintaining competitive English performance. The model is optimized for bilingual interaction and performs well on Thai-English code-switching, MT-Bench, IFEval, and tool-use benchmarks.\n\nDespite its smaller size, it demonstrates strong generalization across math, coding, and multilingual benchmarks, outperforming comparable 8B models across most Thai-specific tasks. Full benchmark results and methodology are available in the [technical report.](https://arxiv.org/abs/2412.13702)", + "name": "Arcee AI: Spotlight", + "shortName": "Spotlight", + "author": "arcee-ai", + "description": "Spotlight is a 7‑billion‑parameter vision‑language model derived from Qwen 2.5‑VL and fine‑tuned by Arcee AI for tight image‑text grounding tasks. It offers a 32 k‑token context window, enabling rich multimodal conversations that combine lengthy documents with one or more images. Training emphasized fast inference on consumer GPUs while retaining strong captioning, visual‐question‑answering, and diagram‑analysis accuracy. As a result, Spotlight slots neatly into agent workflows where screenshots, charts or UI mock‑ups need to be interpreted on the fly. Early benchmarks show it matching or out‑scoring larger VLMs such as LLaVA‑1.6 13 B on popular VQA and POPE alignment tests. ", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 131072, "inputModalities": [ + "image", "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "llama3", + "group": "Other", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|eot_id|>", - "<|end_of_text|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "scb10x/llama3.1-typhoon2-8b-instruct", + "permaslug": "arcee-ai/spotlight", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "scb10x/llama3.1-typhoon2-8b-instruct", - "modelVariantPermaslug": "scb10x/llama3.1-typhoon2-8b-instruct", + "modelVariantSlug": "arcee-ai/spotlight", + "modelVariantPermaslug": "arcee-ai/spotlight", "providerName": "Together", "providerInfo": { "name": "Together", @@ -79959,14 +82931,14 @@ export default { } }, "providerDisplayName": "Together", - "providerModelId": "scb10x/scb10x-llama3-1-typhoon2-8b-instruct", + "providerModelId": "arcee_ai/arcee-spotlight", "providerGroup": "Together", "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 65537, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -80021,21 +82993,23 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 308, - "newest": 68, - "throughputHighToLow": 110, - "latencyLowToHigh": 73, - "pricingLowToHigh": 141, - "pricingHighToLow": 177 + "topWeekly": 307, + "newest": 4, + "throughputHighToLow": 23, + "latencyLowToHigh": 56, + "pricingLowToHigh": 137, + "pricingHighToLow": 179 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://arcee.ai/\\u0026size=256", "providers": [ { "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", "slug": "together", "quantization": null, - "context": 8192, - "maxCompletionTokens": null, - "providerModelId": "scb10x/scb10x-llama3-1-typhoon2-8b-instruct", + "context": 131072, + "maxCompletionTokens": 65537, + "providerModelId": "arcee_ai/arcee-spotlight", "pricing": { "prompt": "0.00000018", "completion": "0.00000018", @@ -80059,22 +83033,24 @@ export default { "response_format" ], "inputCost": 0.18, - "outputCost": 0.18 + "outputCost": 0.18, + "throughput": 190.058, + "latency": 268 } ] }, { - "slug": "meta-llama/llama-2-70b-chat", - "hfSlug": "meta-llama/Llama-2-70b-chat-hf", + "slug": "openai/gpt-4-0314", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-06-20T00:00:00+00:00", + "createdAt": "2023-05-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 2 70B Chat", - "shortName": "Llama 2 70B Chat", - "author": "meta-llama", - "description": "The flagship, 70 billion parameter language model from Meta, fine tuned for chat completions. Llama 2 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align to human preferences for helpfulness and safety.", + "name": "OpenAI: GPT-4 (older v0314)", + "shortName": "GPT-4 (older v0314)", + "author": "openai", + "description": "GPT-4-0314 is the first version of GPT-4 released, with a context length of 8,192 tokens, and was supported until June 14. Training data: up to Sep 2021.", "modelVersionGroupId": null, - "contextLength": 4096, + "contextLength": 8191, "inputModalities": [ "text" ], @@ -80082,35 +83058,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama2", - "instructType": "llama2", + "group": "GPT", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "", - "[INST]" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-2-70b-chat", + "permaslug": "openai/gpt-4-0314", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "08b4e0b7-69ca-4c23-8500-ff49a5175285", - "name": "Together | meta-llama/llama-2-70b-chat", - "contextLength": 4096, + "id": "69206fec-9f6f-4338-9919-5ed90134c376", + "name": "OpenAI | openai/gpt-4-0314", + "contextLength": 8191, "model": { - "slug": "meta-llama/llama-2-70b-chat", - "hfSlug": "meta-llama/Llama-2-70b-chat-hf", + "slug": "openai/gpt-4-0314", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-06-20T00:00:00+00:00", + "createdAt": "2023-05-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 2 70B Chat", - "shortName": "Llama 2 70B Chat", - "author": "meta-llama", - "description": "The flagship, 70 billion parameter language model from Meta, fine tuned for chat completions. Llama 2 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align to human preferences for helpfulness and safety.", + "name": "OpenAI: GPT-4 (older v0314)", + "shortName": "GPT-4 (older v0314)", + "author": "openai", + "description": "GPT-4-0314 is the first version of GPT-4 released, with a context length of 8,192 tokens, and was supported until June 14. Training data: up to Sep 2021.", "modelVersionGroupId": null, - "contextLength": 4096, + "contextLength": 8191, "inputModalities": [ "text" ], @@ -80118,91 +83091,99 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama2", - "instructType": "llama2", + "group": "GPT", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "", - "[INST]" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-2-70b-chat", + "permaslug": "openai/gpt-4-0314", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "meta-llama/llama-2-70b-chat", - "modelVariantPermaslug": "meta-llama/llama-2-70b-chat", - "providerName": "Together", + "modelVariantSlug": "openai/gpt-4-0314", + "modelVariantPermaslug": "openai/gpt-4-0314", + "providerName": "OpenAI", "providerInfo": { - "name": "Together", - "displayName": "Together", - "slug": "together", + "name": "OpenAI", + "displayName": "OpenAI", + "slug": "openai", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false - } + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": false, - "group": "Together", + "moderationRequired": true, + "group": "OpenAI", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": null, + "statusPageUrl": "https://status.openai.com/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" + "url": "/images/icons/OpenAI.svg", + "invertRequired": true } }, - "providerDisplayName": "Together", - "providerModelId": "meta-llama-llama-2-70b-hf", - "providerGroup": "Together", - "quantization": null, + "providerDisplayName": "OpenAI", + "providerModelId": "gpt-4-0314", + "providerGroup": "OpenAI", + "quantization": "unknown", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "top_k", - "repetition_penalty", + "seed", "logit_bias", - "min_p", - "response_format" + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" ], "isByok": false, - "moderationRequired": false, + "moderationRequired": true, "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", + "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", + "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, + "requiresUserIds": true, "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.0000009", - "completion": "0.0000009", + "prompt": "0.00003", + "completion": "0.00006", "image": "0", "request": "0", "webSearch": "0", @@ -80213,7 +83194,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": false, + "supportsToolParameters": true, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -80226,24 +83207,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 309, - "newest": 314, - "throughputHighToLow": 184, - "latencyLowToHigh": 112, - "pricingLowToHigh": 216, - "pricingHighToLow": 105 + "topWeekly": 308, + "newest": 318, + "throughputHighToLow": 187, + "latencyLowToHigh": 145, + "pricingLowToHigh": 313, + "pricingHighToLow": 6 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { - "name": "Together", - "slug": "together", - "quantization": null, - "context": 4096, - "maxCompletionTokens": null, - "providerModelId": "meta-llama-llama-2-70b-hf", + "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", + "slug": "openAi", + "quantization": "unknown", + "context": 8191, + "maxCompletionTokens": 4096, + "providerModelId": "gpt-4-0314", "pricing": { - "prompt": "0.0000009", - "completion": "0.0000009", + "prompt": "0.00003", + "completion": "0.00006", "image": "0", "request": "0", "webSearch": "0", @@ -80251,35 +83234,40 @@ export default { "discount": 0 }, "supportedParameters": [ + "tools", + "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "top_k", - "repetition_penalty", + "seed", "logit_bias", - "min_p", - "response_format" + "logprobs", + "top_logprobs", + "response_format", + "structured_outputs" ], - "inputCost": 0.9, - "outputCost": 0.9 + "inputCost": 30, + "outputCost": 60, + "throughput": 42.422, + "latency": 946 } ] }, { - "slug": "eva-unit-01/eva-qwen-2.5-32b", - "hfSlug": "EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2", + "slug": "cohere/command", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-08T22:27:27.726229+00:00", + "createdAt": "2024-03-14T00:00:00+00:00", "hfUpdatedAt": null, - "name": "EVA Qwen2.5 32B", - "shortName": "EVA Qwen2.5 32B", - "author": "eva-unit-01", - "description": "EVA Qwen2.5 32B is a roleplaying/storywriting specialist model. It's a full-parameter finetune of Qwen2.5-32B on mixture of synthetic and natural data.\n\nIt uses Celeste 70B 0.1 data mixture, greatly expanding it to improve versatility, creativity and \"flavor\" of the resulting model.", + "name": "Cohere: Command", + "shortName": "Command", + "author": "cohere", + "description": "Command is an instruction-following conversational model that performs language tasks with high quality, more reliably and with a longer context than our base generative models.\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", "modelVersionGroupId": null, - "contextLength": 16384, + "contextLength": 4096, "inputModalities": [ "text" ], @@ -80287,36 +83275,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", + "group": "Cohere", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "eva-unit-01/eva-qwen-2.5-32b", + "permaslug": "cohere/command", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "3d087ccd-9e09-47cd-9f1f-8822ef9156e0", - "name": "Featherless | eva-unit-01/eva-qwen-2.5-32b", - "contextLength": 16384, + "id": "73e8a047-c823-454e-ab82-ceea5d165747", + "name": "Cohere | cohere/command", + "contextLength": 4096, "model": { - "slug": "eva-unit-01/eva-qwen-2.5-32b", - "hfSlug": "EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2", + "slug": "cohere/command", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-11-08T22:27:27.726229+00:00", + "createdAt": "2024-03-14T00:00:00+00:00", "hfUpdatedAt": null, - "name": "EVA Qwen2.5 32B", - "shortName": "EVA Qwen2.5 32B", - "author": "eva-unit-01", - "description": "EVA Qwen2.5 32B is a roleplaying/storywriting specialist model. It's a full-parameter finetune of Qwen2.5-32B on mixture of synthetic and natural data.\n\nIt uses Celeste 70B 0.1 data mixture, greatly expanding it to improve versatility, creativity and \"flavor\" of the resulting model.", + "name": "Cohere: Command", + "shortName": "Command", + "author": "cohere", + "description": "Command is an instruction-following conversational model that performs language tasks with high quality, more reliably and with a longer context than our base generative models.\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", "modelVersionGroupId": null, - "contextLength": 32000, + "contextLength": 4096, "inputModalities": [ "text" ], @@ -80324,62 +83308,59 @@ export default { "text" ], "hasTextOutput": true, - "group": "Qwen", - "instructType": "chatml", + "group": "Cohere", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "eva-unit-01/eva-qwen-2.5-32b", + "permaslug": "cohere/command", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "eva-unit-01/eva-qwen-2.5-32b", - "modelVariantPermaslug": "eva-unit-01/eva-qwen-2.5-32b", - "providerName": "Featherless", + "modelVariantSlug": "cohere/command", + "modelVariantPermaslug": "cohere/command", + "providerName": "Cohere", "providerInfo": { - "name": "Featherless", - "displayName": "Featherless", - "slug": "featherless", + "name": "Cohere", + "displayName": "Cohere", + "slug": "cohere", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", + "privacyPolicyUrl": "https://cohere.com/privacy", + "termsOfServiceUrl": "https://cohere.com/terms-of-use", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": false, + "hasCompletions": false, + "isAbortable": true, "moderationRequired": false, - "group": "Featherless", + "group": "Cohere", "editors": [], "owners": [], - "isMultipartSupported": false, - "statusPageUrl": null, + "isMultipartSupported": true, + "statusPageUrl": "https://status.cohere.ai/", "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" + "url": "/images/icons/Cohere.png" } }, - "providerDisplayName": "Featherless", - "providerModelId": "EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2", - "providerGroup": "Featherless", - "quantization": "fp8", + "providerDisplayName": "Cohere", + "providerModelId": "command", + "providerGroup": "Cohere", + "quantization": "unknown", "variant": "standard", "isFree": false, - "canAbort": false, + "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": 4000, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -80389,26 +83370,28 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", "top_k", - "min_p", - "seed" + "seed", + "response_format", + "structured_outputs" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://featherless.ai/terms", - "privacyPolicyUrl": "https://featherless.ai/privacy", + "privacyPolicyUrl": "https://cohere.com/privacy", + "termsOfServiceUrl": "https://cohere.com/terms-of-use", "paidModels": { "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "training": false, - "retainsPrompts": false + "retainsPrompts": true, + "retentionDays": 30 }, "pricing": { - "prompt": "0.0000026", - "completion": "0.0000034", + "prompt": "0.000001", + "completion": "0.000002", "image": "0", "request": "0", "webSearch": "0", @@ -80421,35 +83404,38 @@ export default { "isDisabled": false, "supportsToolParameters": false, "supportsReasoning": false, - "supportsMultipart": false, + "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": true, + "hasCompletions": false, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 310, - "newest": 175, - "throughputHighToLow": 302, - "latencyLowToHigh": 237, - "pricingLowToHigh": 253, - "pricingHighToLow": 66 + "topWeekly": 309, + "newest": 275, + "throughputHighToLow": 253, + "latencyLowToHigh": 147, + "pricingLowToHigh": 223, + "pricingHighToLow": 97 }, + "authorIcon": "https://openrouter.ai/images/icons/Cohere.png", "providers": [ { - "name": "Featherless", - "slug": "featherless", - "quantization": "fp8", - "context": 16384, - "maxCompletionTokens": 4096, - "providerModelId": "EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2", + "name": "Cohere", + "icon": "https://openrouter.ai/images/icons/Cohere.png", + "slug": "cohere", + "quantization": "unknown", + "context": 4096, + "maxCompletionTokens": 4000, + "providerModelId": "command", "pricing": { - "prompt": "0.0000026", - "completion": "0.0000034", + "prompt": "0.000001", + "completion": "0.000002", "image": "0", "request": "0", "webSearch": "0", @@ -80463,13 +83449,15 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "repetition_penalty", "top_k", - "min_p", - "seed" + "seed", + "response_format", + "structured_outputs" ], - "inputCost": 2.6, - "outputCost": 3.4 + "inputCost": 1, + "outputCost": 2, + "throughput": 25.501, + "latency": 870.5 } ] }, @@ -80636,16 +83624,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 311, + "topWeekly": 310, "newest": 51, - "throughputHighToLow": 274, - "latencyLowToHigh": 213, - "pricingLowToHigh": 205, - "pricingHighToLow": 107 + "throughputHighToLow": 273, + "latencyLowToHigh": 230, + "pricingLowToHigh": 204, + "pricingHighToLow": 108 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", "slug": "featherless", "quantization": null, "context": 4096, @@ -80673,22 +83663,24 @@ export default { "seed" ], "inputCost": 0.8, - "outputCost": 1.2 + "outputCost": 1.2, + "throughput": 21.085, + "latency": 1593.5 } ] }, { - "slug": "cohere/command", - "hfSlug": null, + "slug": "meta-llama/llama-guard-2-8b", + "hfSlug": "meta-llama/Meta-Llama-Guard-2-8B", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-03-14T00:00:00+00:00", + "createdAt": "2024-05-13T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Cohere: Command", - "shortName": "Command", - "author": "cohere", - "description": "Command is an instruction-following conversational model that performs language tasks with high quality, more reliably and with a longer context than our base generative models.\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", + "name": "Meta: LlamaGuard 2 8B", + "shortName": "LlamaGuard 2 8B", + "author": "meta-llama", + "description": "This safeguard model has 8B parameters and is based on the Llama 3 family. Just like is predecessor, [LlamaGuard 1](https://huggingface.co/meta-llama/LlamaGuard-7b), it can do both prompt and response classification.\n\nLlamaGuard 2 acts as a normal LLM would, generating text that indicates whether the given input/output is safe/unsafe. If deemed unsafe, it will also share the content categories violated.\n\nFor best results, please use raw prompt input or the `/completions` endpoint, instead of the chat API.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", "modelVersionGroupId": null, - "contextLength": 4096, + "contextLength": 8192, "inputModalities": [ "text" ], @@ -80696,32 +83688,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Cohere", - "instructType": null, + "group": "Llama3", + "instructType": "none", "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cohere/command", + "permaslug": "meta-llama/llama-guard-2-8b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "73e8a047-c823-454e-ab82-ceea5d165747", - "name": "Cohere | cohere/command", - "contextLength": 4096, + "id": "4f39d8b1-b1e1-4158-ab45-8a3de7674c7b", + "name": "Together | meta-llama/llama-guard-2-8b", + "contextLength": 8192, "model": { - "slug": "cohere/command", - "hfSlug": null, + "slug": "meta-llama/llama-guard-2-8b", + "hfSlug": "meta-llama/Meta-Llama-Guard-2-8B", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-03-14T00:00:00+00:00", + "createdAt": "2024-05-13T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Cohere: Command", - "shortName": "Command", - "author": "cohere", - "description": "Command is an instruction-following conversational model that performs language tasks with high quality, more reliably and with a longer context than our base generative models.\n\nUse of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).", + "name": "Meta: LlamaGuard 2 8B", + "shortName": "LlamaGuard 2 8B", + "author": "meta-llama", + "description": "This safeguard model has 8B parameters and is based on the Llama 3 family. Just like is predecessor, [LlamaGuard 1](https://huggingface.co/meta-llama/LlamaGuard-7b), it can do both prompt and response classification.\n\nLlamaGuard 2 acts as a normal LLM would, generating text that indicates whether the given input/output is safe/unsafe. If deemed unsafe, it will also share the content categories violated.\n\nFor best results, please use raw prompt input or the `/completions` endpoint, instead of the chat API.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", "modelVersionGroupId": null, - "contextLength": 4096, + "contextLength": 8192, "inputModalities": [ "text" ], @@ -80729,59 +83721,58 @@ export default { "text" ], "hasTextOutput": true, - "group": "Cohere", - "instructType": null, + "group": "Llama3", + "instructType": "none", "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "cohere/command", + "permaslug": "meta-llama/llama-guard-2-8b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "cohere/command", - "modelVariantPermaslug": "cohere/command", - "providerName": "Cohere", + "modelVariantSlug": "meta-llama/llama-guard-2-8b", + "modelVariantPermaslug": "meta-llama/llama-guard-2-8b", + "providerName": "Together", "providerInfo": { - "name": "Cohere", - "displayName": "Cohere", - "slug": "cohere", + "name": "Together", + "displayName": "Together", + "slug": "together", "baseUrl": "url", "dataPolicy": { - "privacyPolicyUrl": "https://cohere.com/privacy", - "termsOfServiceUrl": "https://cohere.com/terms-of-use", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false } }, "headquarters": "US", "hasChatCompletions": true, - "hasCompletions": false, + "hasCompletions": true, "isAbortable": true, "moderationRequired": false, - "group": "Cohere", + "group": "Together", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.cohere.ai/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/Cohere.png" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" } }, - "providerDisplayName": "Cohere", - "providerModelId": "command", - "providerGroup": "Cohere", + "providerDisplayName": "Together", + "providerModelId": "meta-llama/LlamaGuard-2-8b", + "providerGroup": "Together", "quantization": "unknown", "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4000, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -80792,27 +83783,26 @@ export default { "frequency_penalty", "presence_penalty", "top_k", - "seed", - "response_format", - "structured_outputs" + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "privacyPolicyUrl": "https://cohere.com/privacy", - "termsOfServiceUrl": "https://cohere.com/terms-of-use", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, "pricing": { - "prompt": "0.000001", - "completion": "0.000002", + "prompt": "0.0000002", + "completion": "0.0000002", "image": "0", "request": "0", "webSearch": "0", @@ -80828,33 +83818,34 @@ export default { "supportsMultipart": true, "limitRpm": null, "limitRpd": null, - "hasCompletions": false, + "hasCompletions": true, "hasChatCompletions": true, "features": { - "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 312, - "newest": 275, - "throughputHighToLow": 267, - "latencyLowToHigh": 105, - "pricingLowToHigh": 223, - "pricingHighToLow": 97 + "topWeekly": 311, + "newest": 260, + "throughputHighToLow": 109, + "latencyLowToHigh": 178, + "pricingLowToHigh": 148, + "pricingHighToLow": 172 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [ { - "name": "Cohere", - "slug": "cohere", + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", "quantization": "unknown", - "context": 4096, - "maxCompletionTokens": 4000, - "providerModelId": "command", + "context": 8192, + "maxCompletionTokens": null, + "providerModelId": "meta-llama/LlamaGuard-2-8b", "pricing": { - "prompt": "0.000001", - "completion": "0.000002", + "prompt": "0.0000002", + "completion": "0.0000002", "image": "0", "request": "0", "webSearch": "0", @@ -80869,60 +83860,67 @@ export default { "frequency_penalty", "presence_penalty", "top_k", - "seed", - "response_format", - "structured_outputs" + "repetition_penalty", + "logit_bias", + "min_p", + "response_format" ], - "inputCost": 1, - "outputCost": 2 + "inputCost": 0.2, + "outputCost": 0.2, + "throughput": 74.433, + "latency": 895 } ] }, { - "slug": "meta-llama/llama-guard-2-8b", - "hfSlug": "meta-llama/Meta-Llama-Guard-2-8B", + "slug": "eva-unit-01/eva-qwen-2.5-32b", + "hfSlug": "EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-05-13T00:00:00+00:00", + "createdAt": "2024-11-08T22:27:27.726229+00:00", "hfUpdatedAt": null, - "name": "Meta: LlamaGuard 2 8B", - "shortName": "LlamaGuard 2 8B", - "author": "meta-llama", - "description": "This safeguard model has 8B parameters and is based on the Llama 3 family. Just like is predecessor, [LlamaGuard 1](https://huggingface.co/meta-llama/LlamaGuard-7b), it can do both prompt and response classification.\n\nLlamaGuard 2 acts as a normal LLM would, generating text that indicates whether the given input/output is safe/unsafe. If deemed unsafe, it will also share the content categories violated.\n\nFor best results, please use raw prompt input or the `/completions` endpoint, instead of the chat API.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "name": "EVA Qwen2.5 32B", + "shortName": "EVA Qwen2.5 32B", + "author": "eva-unit-01", + "description": "EVA Qwen2.5 32B is a roleplaying/storywriting specialist model. It's a full-parameter finetune of Qwen2.5-32B on mixture of synthetic and natural data.\n\nIt uses Celeste 70B 0.1 data mixture, greatly expanding it to improve versatility, creativity and \"flavor\" of the resulting model.", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 16384, "inputModalities": [ "text" ], "outputModalities": [ "text" ], - "hasTextOutput": true, - "group": "Llama3", - "instructType": "none", - "defaultSystem": null, - "defaultStops": [], + "hasTextOutput": true, + "group": "Qwen", + "instructType": "chatml", + "defaultSystem": null, + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-guard-2-8b", + "permaslug": "eva-unit-01/eva-qwen-2.5-32b", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "4f39d8b1-b1e1-4158-ab45-8a3de7674c7b", - "name": "Together | meta-llama/llama-guard-2-8b", - "contextLength": 8192, + "id": "3d087ccd-9e09-47cd-9f1f-8822ef9156e0", + "name": "Featherless | eva-unit-01/eva-qwen-2.5-32b", + "contextLength": 16384, "model": { - "slug": "meta-llama/llama-guard-2-8b", - "hfSlug": "meta-llama/Meta-Llama-Guard-2-8B", + "slug": "eva-unit-01/eva-qwen-2.5-32b", + "hfSlug": "EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2", "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-05-13T00:00:00+00:00", + "createdAt": "2024-11-08T22:27:27.726229+00:00", "hfUpdatedAt": null, - "name": "Meta: LlamaGuard 2 8B", - "shortName": "LlamaGuard 2 8B", - "author": "meta-llama", - "description": "This safeguard model has 8B parameters and is based on the Llama 3 family. Just like is predecessor, [LlamaGuard 1](https://huggingface.co/meta-llama/LlamaGuard-7b), it can do both prompt and response classification.\n\nLlamaGuard 2 acts as a normal LLM would, generating text that indicates whether the given input/output is safe/unsafe. If deemed unsafe, it will also share the content categories violated.\n\nFor best results, please use raw prompt input or the `/completions` endpoint, instead of the chat API.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "name": "EVA Qwen2.5 32B", + "shortName": "EVA Qwen2.5 32B", + "author": "eva-unit-01", + "description": "EVA Qwen2.5 32B is a roleplaying/storywriting specialist model. It's a full-parameter finetune of Qwen2.5-32B on mixture of synthetic and natural data.\n\nIt uses Celeste 70B 0.1 data mixture, greatly expanding it to improve versatility, creativity and \"flavor\" of the resulting model.", "modelVersionGroupId": null, - "contextLength": 8192, + "contextLength": 32000, "inputModalities": [ "text" ], @@ -80930,28 +83928,32 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", - "instructType": "none", + "group": "Qwen", + "instructType": "chatml", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-guard-2-8b", + "permaslug": "eva-unit-01/eva-qwen-2.5-32b", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "meta-llama/llama-guard-2-8b", - "modelVariantPermaslug": "meta-llama/llama-guard-2-8b", - "providerName": "Together", + "modelVariantSlug": "eva-unit-01/eva-qwen-2.5-32b", + "modelVariantPermaslug": "eva-unit-01/eva-qwen-2.5-32b", + "providerName": "Featherless", "providerInfo": { - "name": "Together", - "displayName": "Together", - "slug": "together", + "name": "Featherless", + "displayName": "Featherless", + "slug": "featherless", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -80960,28 +83962,28 @@ export default { "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, - "isAbortable": true, + "isAbortable": false, "moderationRequired": false, - "group": "Together", + "group": "Featherless", "editors": [], "owners": [], - "isMultipartSupported": true, + "isMultipartSupported": false, "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256" } }, - "providerDisplayName": "Together", - "providerModelId": "meta-llama/LlamaGuard-2-8b", - "providerGroup": "Together", - "quantization": "unknown", + "providerDisplayName": "Featherless", + "providerModelId": "EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2", + "providerGroup": "Featherless", + "quantization": "fp8", "variant": "standard", "isFree": false, - "canAbort": true, + "canAbort": false, "maxPromptTokens": null, - "maxCompletionTokens": null, + "maxCompletionTokens": 4096, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ @@ -80991,17 +83993,16 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "top_k", "repetition_penalty", - "logit_bias", + "top_k", "min_p", - "response_format" + "seed" ], "isByok": false, "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://www.together.ai/terms-of-service", - "privacyPolicyUrl": "https://www.together.ai/privacy", + "termsOfServiceUrl": "https://featherless.ai/terms", + "privacyPolicyUrl": "https://featherless.ai/privacy", "paidModels": { "training": false, "retainsPrompts": false @@ -81010,8 +84011,8 @@ export default { "retainsPrompts": false }, "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", + "prompt": "0.0000026", + "completion": "0.0000034", "image": "0", "request": "0", "webSearch": "0", @@ -81024,7 +84025,7 @@ export default { "isDisabled": false, "supportsToolParameters": false, "supportsReasoning": false, - "supportsMultipart": true, + "supportsMultipart": false, "limitRpm": null, "limitRpd": null, "hasCompletions": true, @@ -81035,24 +84036,26 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 313, - "newest": 260, - "throughputHighToLow": 107, - "latencyLowToHigh": 257, - "pricingLowToHigh": 151, - "pricingHighToLow": 169 + "topWeekly": 312, + "newest": 175, + "throughputHighToLow": 298, + "latencyLowToHigh": 216, + "pricingLowToHigh": 253, + "pricingHighToLow": 66 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { - "name": "Together", - "slug": "together", - "quantization": "unknown", - "context": 8192, - "maxCompletionTokens": null, - "providerModelId": "meta-llama/LlamaGuard-2-8b", + "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", + "slug": "featherless", + "quantization": "fp8", + "context": 16384, + "maxCompletionTokens": 4096, + "providerModelId": "EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2", "pricing": { - "prompt": "0.0000002", - "completion": "0.0000002", + "prompt": "0.0000026", + "completion": "0.0000034", "image": "0", "request": "0", "webSearch": "0", @@ -81066,29 +84069,30 @@ export default { "stop", "frequency_penalty", "presence_penalty", - "top_k", "repetition_penalty", - "logit_bias", + "top_k", "min_p", - "response_format" + "seed" ], - "inputCost": 0.2, - "outputCost": 0.2 + "inputCost": 2.6, + "outputCost": 3.4, + "throughput": 13.739, + "latency": 1594 } ] }, { - "slug": "openai/gpt-4-0314", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-05-28T00:00:00+00:00", + "slug": "scb10x/llama3.1-typhoon2-8b-instruct", + "hfSlug": "scb10x/llama3.1-typhoon2-8b-instruct", + "updatedAt": "2025-03-28T21:16:32.49476+00:00", + "createdAt": "2025-03-28T21:15:11.257967+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4 (older v0314)", - "shortName": "GPT-4 (older v0314)", - "author": "openai", - "description": "GPT-4-0314 is the first version of GPT-4 released, with a context length of 8,192 tokens, and was supported until June 14. Training data: up to Sep 2021.", + "name": "Typhoon2 8B Instruct", + "shortName": "Typhoon2 8B Instruct", + "author": "scb10x", + "description": "Llama3.1-Typhoon2-8B-Instruct is a Thai-English instruction-tuned model with 8 billion parameters, built on Llama 3.1. It significantly improves over its base model in Thai reasoning, instruction-following, and function-calling tasks, while maintaining competitive English performance. The model is optimized for bilingual interaction and performs well on Thai-English code-switching, MT-Bench, IFEval, and tool-use benchmarks.\n\nDespite its smaller size, it demonstrates strong generalization across math, coding, and multilingual benchmarks, outperforming comparable 8B models across most Thai-specific tasks. Full benchmark results and methodology are available in the [technical report.](https://arxiv.org/abs/2412.13702)", "modelVersionGroupId": null, - "contextLength": 8191, + "contextLength": 8192, "inputModalities": [ "text" ], @@ -81096,32 +84100,35 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-4-0314", + "permaslug": "scb10x/llama3.1-typhoon2-8b-instruct", "reasoningConfig": null, "features": {}, "endpoint": { - "id": "69206fec-9f6f-4338-9919-5ed90134c376", - "name": "OpenAI | openai/gpt-4-0314", - "contextLength": 8191, + "id": "cd687004-e160-49f4-b6a1-0f6ab950f046", + "name": "Together | scb10x/llama3.1-typhoon2-8b-instruct", + "contextLength": 8192, "model": { - "slug": "openai/gpt-4-0314", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-05-28T00:00:00+00:00", + "slug": "scb10x/llama3.1-typhoon2-8b-instruct", + "hfSlug": "scb10x/llama3.1-typhoon2-8b-instruct", + "updatedAt": "2025-03-28T21:16:32.49476+00:00", + "createdAt": "2025-03-28T21:15:11.257967+00:00", "hfUpdatedAt": null, - "name": "OpenAI: GPT-4 (older v0314)", - "shortName": "GPT-4 (older v0314)", - "author": "openai", - "description": "GPT-4-0314 is the first version of GPT-4 released, with a context length of 8,192 tokens, and was supported until June 14. Training data: up to Sep 2021.", + "name": "Typhoon2 8B Instruct", + "shortName": "Typhoon2 8B Instruct", + "author": "scb10x", + "description": "Llama3.1-Typhoon2-8B-Instruct is a Thai-English instruction-tuned model with 8 billion parameters, built on Llama 3.1. It significantly improves over its base model in Thai reasoning, instruction-following, and function-calling tasks, while maintaining competitive English performance. The model is optimized for bilingual interaction and performs well on Thai-English code-switching, MT-Bench, IFEval, and tool-use benchmarks.\n\nDespite its smaller size, it demonstrates strong generalization across math, coding, and multilingual benchmarks, outperforming comparable 8B models across most Thai-specific tasks. Full benchmark results and methodology are available in the [technical report.](https://arxiv.org/abs/2412.13702)", "modelVersionGroupId": null, - "contextLength": 8191, + "contextLength": 8192, "inputModalities": [ "text" ], @@ -81129,99 +84136,91 @@ export default { "text" ], "hasTextOutput": true, - "group": "GPT", - "instructType": null, + "group": "Llama3", + "instructType": "llama3", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "<|eot_id|>", + "<|end_of_text|>" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "openai/gpt-4-0314", + "permaslug": "scb10x/llama3.1-typhoon2-8b-instruct", "reasoningConfig": null, "features": {} }, - "modelVariantSlug": "openai/gpt-4-0314", - "modelVariantPermaslug": "openai/gpt-4-0314", - "providerName": "OpenAI", + "modelVariantSlug": "scb10x/llama3.1-typhoon2-8b-instruct", + "modelVariantPermaslug": "scb10x/llama3.1-typhoon2-8b-instruct", + "providerName": "Together", "providerInfo": { - "name": "OpenAI", - "displayName": "OpenAI", - "slug": "openai", + "name": "Together", + "displayName": "Together", + "slug": "together", "baseUrl": "url", "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true + "retainsPrompts": false + } }, "headquarters": "US", "hasChatCompletions": true, "hasCompletions": true, "isAbortable": true, - "moderationRequired": true, - "group": "OpenAI", + "moderationRequired": false, + "group": "Together", "editors": [], "owners": [], "isMultipartSupported": true, - "statusPageUrl": "https://status.openai.com/", + "statusPageUrl": null, "byokEnabled": true, "isPrimaryProvider": true, "icon": { - "url": "/images/icons/OpenAI.svg", - "invertRequired": true + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256" } }, - "providerDisplayName": "OpenAI", - "providerModelId": "gpt-4-0314", - "providerGroup": "OpenAI", - "quantization": "unknown", + "providerDisplayName": "Together", + "providerModelId": "scb10x/scb10x-llama3-1-typhoon2-8b-instruct", + "providerGroup": "Together", + "quantization": null, "variant": "standard", "isFree": false, "canAbort": true, "maxPromptTokens": null, - "maxCompletionTokens": 4096, + "maxCompletionTokens": null, "maxPromptImages": null, "maxTokensPerImage": null, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "seed", + "top_k", + "repetition_penalty", "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" + "min_p", + "response_format" ], "isByok": false, - "moderationRequired": true, + "moderationRequired": false, "dataPolicy": { - "termsOfServiceUrl": "https://openai.com/policies/row-terms-of-use/", - "privacyPolicyUrl": "https://openai.com/policies/privacy-policy/", - "dataPolicyUrl": "https://platform.openai.com/docs/guides/your-data", + "termsOfServiceUrl": "https://www.together.ai/terms-of-service", + "privacyPolicyUrl": "https://www.together.ai/privacy", "paidModels": { "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, - "requiresUserIds": true, "training": false, - "retainsPrompts": true, - "retentionDays": 30 + "retainsPrompts": false }, "pricing": { - "prompt": "0.00003", - "completion": "0.00006", + "prompt": "0.00000018", + "completion": "0.00000018", "image": "0", "request": "0", "webSearch": "0", @@ -81232,7 +84231,7 @@ export default { "isHidden": false, "isDeranked": false, "isDisabled": false, - "supportsToolParameters": true, + "supportsToolParameters": false, "supportsReasoning": false, "supportsMultipart": true, "limitRpm": null, @@ -81240,29 +84239,32 @@ export default { "hasCompletions": true, "hasChatCompletions": true, "features": { + "supportedParameters": {}, "supportsDocumentUrl": null }, "providerRegion": null }, "sorting": { - "topWeekly": 314, - "newest": 316, - "throughputHighToLow": 208, - "latencyLowToHigh": 153, - "pricingLowToHigh": 313, - "pricingHighToLow": 6 + "topWeekly": 313, + "newest": 68, + "throughputHighToLow": 106, + "latencyLowToHigh": 77, + "pricingLowToHigh": 138, + "pricingHighToLow": 180 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { - "name": "OpenAI", - "slug": "openAi", - "quantization": "unknown", - "context": 8191, - "maxCompletionTokens": 4096, - "providerModelId": "gpt-4-0314", + "name": "Together", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://www.together.ai/&size=256", + "slug": "together", + "quantization": null, + "context": 8192, + "maxCompletionTokens": null, + "providerModelId": "scb10x/scb10x-llama3-1-typhoon2-8b-instruct", "pricing": { - "prompt": "0.00003", - "completion": "0.00006", + "prompt": "0.00000018", + "completion": "0.00000018", "image": "0", "request": "0", "webSearch": "0", @@ -81270,23 +84272,22 @@ export default { "discount": 0 }, "supportedParameters": [ - "tools", - "tool_choice", "max_tokens", "temperature", "top_p", "stop", "frequency_penalty", "presence_penalty", - "seed", + "top_k", + "repetition_penalty", "logit_bias", - "logprobs", - "top_logprobs", - "response_format", - "structured_outputs" + "min_p", + "response_format" ], - "inputCost": 30, - "outputCost": 60 + "inputCost": 0.18, + "outputCost": 0.18, + "throughput": 71.177, + "latency": 609 } ] }, @@ -81453,16 +84454,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 315, + "topWeekly": 314, "newest": 52, - "throughputHighToLow": 271, - "latencyLowToHigh": 190, - "pricingLowToHigh": 206, - "pricingHighToLow": 108 + "throughputHighToLow": 277, + "latencyLowToHigh": 228, + "pricingLowToHigh": 205, + "pricingHighToLow": 109 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [ { "name": "Featherless", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://featherless.ai/&size=256", "slug": "featherless", "quantization": null, "context": 4096, @@ -81490,199 +84493,9 @@ export default { "seed" ], "inputCost": 0.8, - "outputCost": 1.2 - } - ] - }, - { - "slug": "anthropic/claude-2", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-11-22T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Anthropic: Claude v2 (self-moderated)", - "shortName": "Claude v2 (self-moderated)", - "author": "anthropic", - "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.", - "modelVersionGroupId": null, - "contextLength": 200000, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Claude", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "anthropic/claude-2", - "reasoningConfig": null, - "features": {}, - "endpoint": { - "id": "258ddc79-e520-4e68-8069-ea6771698c5a", - "name": "Anthropic | anthropic/claude-2:beta", - "contextLength": 200000, - "model": { - "slug": "anthropic/claude-2", - "hfSlug": null, - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-11-22T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Anthropic: Claude v2", - "shortName": "Claude v2", - "author": "anthropic", - "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.", - "modelVersionGroupId": null, - "contextLength": 200000, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Claude", - "instructType": null, - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "anthropic/claude-2", - "reasoningConfig": null, - "features": {} - }, - "modelVariantSlug": "anthropic/claude-2:beta", - "modelVariantPermaslug": "anthropic/claude-2:beta", - "providerName": "Anthropic", - "providerInfo": { - "name": "Anthropic", - "displayName": "Anthropic", - "slug": "anthropic", - "baseUrl": "url", - "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true - }, - "headquarters": "US", - "hasChatCompletions": true, - "hasCompletions": true, - "isAbortable": true, - "moderationRequired": true, - "group": "Anthropic", - "editors": [], - "owners": [], - "isMultipartSupported": true, - "statusPageUrl": "https://status.anthropic.com/", - "byokEnabled": true, - "isPrimaryProvider": true, - "icon": { - "url": "/images/icons/Anthropic.svg" - } - }, - "providerDisplayName": "Anthropic", - "providerModelId": "claude-2.1", - "providerGroup": "Anthropic", - "quantization": "unknown", - "variant": "beta", - "isFree": false, - "canAbort": true, - "maxPromptTokens": null, - "maxCompletionTokens": 4096, - "maxPromptImages": null, - "maxTokensPerImage": null, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "top_k", - "stop" - ], - "isByok": false, - "moderationRequired": false, - "dataPolicy": { - "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", - "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", - "paidModels": { - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "requiresUserIds": true, - "training": false, - "retainsPrompts": true, - "retentionDays": 30 - }, - "pricing": { - "prompt": "0.000008", - "completion": "0.000024", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "variablePricings": [], - "isHidden": false, - "isDeranked": false, - "isDisabled": false, - "supportsToolParameters": false, - "supportsReasoning": false, - "supportsMultipart": true, - "limitRpm": null, - "limitRpd": null, - "hasCompletions": true, - "hasChatCompletions": true, - "features": { - "supportsDocumentUrl": null - }, - "providerRegion": null - }, - "sorting": { - "topWeekly": 283, - "newest": 294, - "throughputHighToLow": 293, - "latencyLowToHigh": 235, - "pricingLowToHigh": 296, - "pricingHighToLow": 19 - }, - "providers": [ - { - "name": "Anthropic", - "slug": "anthropic", - "quantization": "unknown", - "context": 200000, - "maxCompletionTokens": 4096, - "providerModelId": "claude-2.1", - "pricing": { - "prompt": "0.000008", - "completion": "0.000024", - "image": "0", - "request": "0", - "webSearch": "0", - "internalReasoning": "0", - "discount": 0 - }, - "supportedParameters": [ - "max_tokens", - "temperature", - "top_p", - "top_k", - "stop" - ], - "inputCost": 8, - "outputCost": 24 + "outputCost": 1.2, + "throughput": 21.052, + "latency": 1506 } ] }, @@ -81838,16 +84651,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 317, - "newest": 192, - "throughputHighToLow": 217, - "latencyLowToHigh": 276, + "topWeekly": 315, + "newest": 191, + "throughputHighToLow": 190, + "latencyLowToHigh": 278, "pricingLowToHigh": 262, "pricingHighToLow": 55 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://inflection.ai/\\u0026size=256", "providers": [ { "name": "Inflection", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://inflection.ai/&size=256", "slug": "inflection", "quantization": "unknown", "context": 8000, @@ -81869,7 +84684,9 @@ export default { "stop" ], "inputCost": 2.5, - "outputCost": 10 + "outputCost": 10, + "throughput": 45.045, + "latency": 2822 } ] }, @@ -82041,16 +84858,18 @@ export default { "providerRegion": null }, "sorting": { - "topWeekly": 318, + "topWeekly": 316, "newest": 308, - "throughputHighToLow": 245, - "latencyLowToHigh": 169, + "throughputHighToLow": 242, + "latencyLowToHigh": 174, "pricingLowToHigh": 316, "pricingHighToLow": 3 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [ { "name": "OpenAI", + "icon": "https://openrouter.ai/images/icons/OpenAI.svg", "slug": "openAi", "quantization": "unknown", "context": 32767, @@ -82082,22 +84901,24 @@ export default { "structured_outputs" ], "inputCost": 60, - "outputCost": 120 + "outputCost": 120, + "throughput": 31.436, + "latency": 1077 } ] }, { - "slug": "ai21/jamba-1-5-large", + "slug": "anthropic/claude-2", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-23T00:00:00+00:00", + "createdAt": "2023-11-22T00:00:00+00:00", "hfUpdatedAt": null, - "name": "AI21: Jamba 1.5 Large", - "shortName": "Jamba 1.5 Large", - "author": "ai21", - "description": "Jamba 1.5 Large is part of AI21's new family of open models, offering superior speed, efficiency, and quality.\n\nIt features a 256K effective context window, the longest among open models, enabling improved performance on tasks like document summarization and analysis.\n\nBuilt on a novel SSM-Transformer architecture, it outperforms larger models like Llama 3.1 70B on benchmarks while maintaining resource efficiency.\n\nRead their [announcement](https://www.ai21.com/blog/announcing-jamba-model-family) to learn more.", + "name": "Anthropic: Claude v2 (self-moderated)", + "shortName": "Claude v2 (self-moderated)", + "author": "anthropic", + "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.", "modelVersionGroupId": null, - "contextLength": 256000, + "contextLength": 200000, "inputModalities": [ "text" ], @@ -82105,37 +84926,193 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "Claude", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "ai21/jamba-1-5-large", + "permaslug": "anthropic/claude-2", "reasoningConfig": null, "features": {}, - "endpoint": null, + "endpoint": { + "id": "258ddc79-e520-4e68-8069-ea6771698c5a", + "name": "Anthropic | anthropic/claude-2:beta", + "contextLength": 200000, + "model": { + "slug": "anthropic/claude-2", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2023-11-22T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Anthropic: Claude v2", + "shortName": "Claude v2", + "author": "anthropic", + "description": "Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.", + "modelVersionGroupId": null, + "contextLength": 200000, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Claude", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "anthropic/claude-2", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "anthropic/claude-2:beta", + "modelVariantPermaslug": "anthropic/claude-2:beta", + "providerName": "Anthropic", + "providerInfo": { + "name": "Anthropic", + "displayName": "Anthropic", + "slug": "anthropic", + "baseUrl": "url", + "dataPolicy": { + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true + }, + "headquarters": "US", + "hasChatCompletions": true, + "hasCompletions": true, + "isAbortable": true, + "moderationRequired": true, + "group": "Anthropic", + "editors": [], + "owners": [], + "isMultipartSupported": true, + "statusPageUrl": "https://status.anthropic.com/", + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "/images/icons/Anthropic.svg" + } + }, + "providerDisplayName": "Anthropic", + "providerModelId": "claude-2.1", + "providerGroup": "Anthropic", + "quantization": "unknown", + "variant": "beta", + "isFree": false, + "canAbort": true, + "maxPromptTokens": null, + "maxCompletionTokens": 4096, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "termsOfServiceUrl": "https://www.anthropic.com/legal/commercial-terms", + "privacyPolicyUrl": "https://www.anthropic.com/legal/privacy", + "paidModels": { + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "requiresUserIds": true, + "training": false, + "retainsPrompts": true, + "retentionDays": 30 + }, + "pricing": { + "prompt": "0.000008", + "completion": "0.000024", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": true, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": true, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null + }, + "providerRegion": null + }, "sorting": { - "topWeekly": 319, - "newest": 339, - "throughputHighToLow": 339, - "latencyLowToHigh": 339, - "pricingLowToHigh": 339, - "pricingHighToLow": 341 + "topWeekly": 282, + "newest": 296, + "throughputHighToLow": 291, + "latencyLowToHigh": 202, + "pricingLowToHigh": 296, + "pricingHighToLow": 19 }, - "providers": [] + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", + "providers": [ + { + "name": "Anthropic", + "icon": "https://openrouter.ai/images/icons/Anthropic.svg", + "slug": "anthropic", + "quantization": "unknown", + "context": 200000, + "maxCompletionTokens": 4096, + "providerModelId": "claude-2.1", + "pricing": { + "prompt": "0.000008", + "completion": "0.000024", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "top_k", + "stop" + ], + "inputCost": 8, + "outputCost": 24, + "throughput": 16.037, + "latency": 1567 + } + ] }, { - "slug": "ai21/jamba-1-5-mini", + "slug": "ai21/jamba-instruct", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-23T00:00:00+00:00", + "createdAt": "2024-06-25T00:00:00+00:00", "hfUpdatedAt": null, - "name": "AI21: Jamba 1.5 Mini", - "shortName": "Jamba 1.5 Mini", + "name": "AI21: Jamba Instruct", + "shortName": "Jamba Instruct", "author": "ai21", - "description": "Jamba 1.5 Mini is the world's first production-grade Mamba-based model, combining SSM and Transformer architectures for a 256K context window and high efficiency.\n\nIt works with 9 languages and can handle various writing and analysis tasks as well as or better than similar small models.\n\nThis model uses less computer memory and works faster with longer texts than previous designs.\n\nRead their [announcement](https://www.ai21.com/blog/announcing-jamba-model-family) to learn more.", + "description": "The Jamba-Instruct model, introduced by AI21 Labs, is an instruction-tuned variant of their hybrid SSM-Transformer Jamba model, specifically optimized for enterprise applications.\n\n- 256K Context Window: It can process extensive information, equivalent to a 400-page novel, which is beneficial for tasks involving large documents such as financial reports or legal documents\n- Safety and Accuracy: Jamba-Instruct is designed with enhanced safety features to ensure secure deployment in enterprise environments, reducing the risk and cost of implementation\n\nRead their [announcement](https://www.ai21.com/blog/announcing-jamba) to learn more.\n\nJamba has a knowledge cutoff of February 2024.", "modelVersionGroupId": null, "contextLength": 256000, "inputModalities": [ @@ -82152,19 +85129,163 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "ai21/jamba-1-5-mini", + "permaslug": "ai21/jamba-instruct", "reasoningConfig": null, "features": {}, - "endpoint": null, + "endpoint": { + "id": "4bdfe9f1-ba9f-4da1-ae1f-904bc6a7cd84", + "name": "AI21 | ai21/jamba-instruct", + "contextLength": 256000, + "model": { + "slug": "ai21/jamba-instruct", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-06-25T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "AI21: Jamba Instruct", + "shortName": "Jamba Instruct", + "author": "ai21", + "description": "The Jamba-Instruct model, introduced by AI21 Labs, is an instruction-tuned variant of their hybrid SSM-Transformer Jamba model, specifically optimized for enterprise applications.\n\n- 256K Context Window: It can process extensive information, equivalent to a 400-page novel, which is beneficial for tasks involving large documents such as financial reports or legal documents\n- Safety and Accuracy: Jamba-Instruct is designed with enhanced safety features to ensure secure deployment in enterprise environments, reducing the risk and cost of implementation\n\nRead their [announcement](https://www.ai21.com/blog/announcing-jamba) to learn more.\n\nJamba has a knowledge cutoff of February 2024.", + "modelVersionGroupId": null, + "contextLength": 256000, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Other", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "ai21/jamba-instruct", + "reasoningConfig": null, + "features": {} + }, + "modelVariantSlug": "ai21/jamba-instruct", + "modelVariantPermaslug": "ai21/jamba-instruct", + "providerName": "AI21", + "providerInfo": { + "name": "AI21", + "displayName": "AI21", + "slug": "ai21", + "baseUrl": "url", + "dataPolicy": { + "privacyPolicyUrl": "https://www.ai21.com/privacy-policy/", + "termsOfServiceUrl": "https://www.ai21.com/terms-of-service/", + "paidModels": { + "training": false + } + }, + "headquarters": "IL", + "hasChatCompletions": true, + "hasCompletions": false, + "isAbortable": false, + "moderationRequired": false, + "group": "AI21", + "editors": [], + "owners": [], + "isMultipartSupported": false, + "statusPageUrl": "https://status.ai21.com/", + "byokEnabled": true, + "isPrimaryProvider": true, + "icon": { + "url": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://ai21.com/&size=256" + } + }, + "providerDisplayName": "AI21", + "providerModelId": "jamba-instruct", + "providerGroup": "AI21", + "quantization": "unknown", + "variant": "standard", + "isFree": false, + "canAbort": false, + "maxPromptTokens": null, + "maxCompletionTokens": 4096, + "maxPromptImages": null, + "maxTokensPerImage": null, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop" + ], + "isByok": false, + "moderationRequired": false, + "dataPolicy": { + "privacyPolicyUrl": "https://www.ai21.com/privacy-policy/", + "termsOfServiceUrl": "https://www.ai21.com/terms-of-service/", + "paidModels": { + "training": false + }, + "training": false + }, + "pricing": { + "prompt": "0.0000005", + "completion": "0.0000007", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "variablePricings": [], + "isHidden": false, + "isDeranked": false, + "isDisabled": false, + "supportsToolParameters": false, + "supportsReasoning": false, + "supportsMultipart": false, + "limitRpm": null, + "limitRpd": null, + "hasCompletions": false, + "hasChatCompletions": true, + "features": { + "supportsDocumentUrl": null + }, + "providerRegion": null + }, "sorting": { - "topWeekly": 320, - "newest": 341, - "throughputHighToLow": 341, - "latencyLowToHigh": 341, - "pricingLowToHigh": 341, - "pricingHighToLow": 340 + "topWeekly": 318, + "newest": 243, + "throughputHighToLow": 311, + "latencyLowToHigh": 318, + "pricingLowToHigh": 181, + "pricingHighToLow": 137 }, - "providers": [] + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai21.com/\\u0026size=256", + "providers": [ + { + "name": "AI21", + "icon": "https://t0.gstatic.com/faviconV2?client=SOCIAL&type=FAVICON&fallback_opts=TYPE,SIZE,URL&url=https://ai21.com/&size=256", + "slug": "ai21", + "quantization": "unknown", + "context": 256000, + "maxCompletionTokens": 4096, + "providerModelId": "jamba-instruct", + "pricing": { + "prompt": "0.0000005", + "completion": "0.0000007", + "image": "0", + "request": "0", + "webSearch": "0", + "internalReasoning": "0", + "discount": 0 + }, + "supportedParameters": [ + "max_tokens", + "temperature", + "top_p", + "stop" + ], + "inputCost": 0.5, + "outputCost": 0.7 + } + ] }, { "slug": "openrouter/optimus-alpha", @@ -82198,13 +85319,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 321, + "topWeekly": 319, "newest": 319, "throughputHighToLow": 319, "latencyLowToHigh": 319, "pricingLowToHigh": 319, "pricingHighToLow": 319 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://openrouter.ai/\\u0026size=256", "providers": [] }, { @@ -82238,13 +85360,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 322, + "topWeekly": 320, "newest": 320, "throughputHighToLow": 320, "latencyLowToHigh": 320, "pricingLowToHigh": 320, "pricingHighToLow": 320 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nvidia.com/\\u0026size=256", "providers": [] }, { @@ -82278,13 +85401,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 323, + "topWeekly": 321, "newest": 321, "throughputHighToLow": 321, "latencyLowToHigh": 321, "pricingLowToHigh": 321, "pricingHighToLow": 321 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -82319,13 +85443,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 324, + "topWeekly": 322, "newest": 322, "throughputHighToLow": 322, "latencyLowToHigh": 322, "pricingLowToHigh": 322, "pricingHighToLow": 322 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://openrouter.ai/\\u0026size=256", "providers": [] }, { @@ -82360,13 +85485,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 325, + "topWeekly": 323, "newest": 323, "throughputHighToLow": 323, "latencyLowToHigh": 323, "pricingLowToHigh": 323, "pricingHighToLow": 323 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://allenai.org/\\u0026size=256", "providers": [] }, { @@ -82411,13 +85537,14 @@ export default { }, "endpoint": null, "sorting": { - "topWeekly": 326, + "topWeekly": 324, "newest": 324, "throughputHighToLow": 324, "latencyLowToHigh": 324, "pricingLowToHigh": 324, "pricingHighToLow": 324 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -82451,13 +85578,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 327, + "topWeekly": 325, "newest": 325, "throughputHighToLow": 325, "latencyLowToHigh": 325, "pricingLowToHigh": 325, "pricingHighToLow": 325 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://allenai.org/\\u0026size=256", "providers": [] }, { @@ -82491,13 +85619,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 328, + "topWeekly": 326, "newest": 326, "throughputHighToLow": 326, "latencyLowToHigh": 326, "pricingLowToHigh": 326, "pricingHighToLow": 326 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -82531,13 +85660,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 329, + "topWeekly": 327, "newest": 327, "throughputHighToLow": 327, "latencyLowToHigh": 327, "pricingLowToHigh": 327, "pricingHighToLow": 327 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [] }, { @@ -82571,13 +85701,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 330, + "topWeekly": 328, "newest": 328, "throughputHighToLow": 328, "latencyLowToHigh": 328, "pricingLowToHigh": 328, "pricingHighToLow": 328 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://allenai.org/\\u0026size=256", "providers": [] }, { @@ -82611,13 +85742,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 331, + "topWeekly": 329, "newest": 329, "throughputHighToLow": 329, "latencyLowToHigh": 329, "pricingLowToHigh": 329, "pricingHighToLow": 329 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -82655,13 +85787,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 332, + "topWeekly": 330, "newest": 330, "throughputHighToLow": 330, "latencyLowToHigh": 330, "pricingLowToHigh": 330, "pricingHighToLow": 330 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -82696,13 +85829,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 333, + "topWeekly": 331, "newest": 331, "throughputHighToLow": 331, "latencyLowToHigh": 331, "pricingLowToHigh": 331, "pricingHighToLow": 331 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [] }, { @@ -82737,13 +85871,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 334, + "topWeekly": 332, "newest": 332, "throughputHighToLow": 332, "latencyLowToHigh": 332, "pricingLowToHigh": 332, "pricingHighToLow": 332 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [] }, { @@ -82777,13 +85912,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 335, + "topWeekly": 333, "newest": 333, "throughputHighToLow": 333, "latencyLowToHigh": 333, "pricingLowToHigh": 333, "pricingHighToLow": 333 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://x.ai/\\u0026size=256", "providers": [] }, { @@ -82817,13 +85953,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 336, + "topWeekly": 334, "newest": 334, "throughputHighToLow": 334, "latencyLowToHigh": 334, "pricingLowToHigh": 334, "pricingHighToLow": 334 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://x.ai/\\u0026size=256", "providers": [] }, { @@ -82861,13 +85998,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 337, + "topWeekly": 335, "newest": 335, "throughputHighToLow": 335, "latencyLowToHigh": 335, "pricingLowToHigh": 335, "pricingHighToLow": 335 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -82901,13 +86039,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 338, + "topWeekly": 336, "newest": 336, "throughputHighToLow": 336, "latencyLowToHigh": 336, "pricingLowToHigh": 336, "pricingHighToLow": 336 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -82942,13 +86081,14 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 339, + "topWeekly": 337, "newest": 337, "throughputHighToLow": 337, "latencyLowToHigh": 337, "pricingLowToHigh": 337, "pricingHighToLow": 337 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [] }, { @@ -82986,13 +86126,55 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 340, + "topWeekly": 338, "newest": 338, "throughputHighToLow": 338, "latencyLowToHigh": 338, "pricingLowToHigh": 338, "pricingHighToLow": 338 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://api.lynn.app/\\u0026size=256", + "providers": [] + }, + { + "slug": "ai21/jamba-1-5-large", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-08-23T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "AI21: Jamba 1.5 Large", + "shortName": "Jamba 1.5 Large", + "author": "ai21", + "description": "Jamba 1.5 Large is part of AI21's new family of open models, offering superior speed, efficiency, and quality.\n\nIt features a 256K effective context window, the longest among open models, enabling improved performance on tasks like document summarization and analysis.\n\nBuilt on a novel SSM-Transformer architecture, it outperforms larger models like Llama 3.1 70B on benchmarks while maintaining resource efficiency.\n\nRead their [announcement](https://www.ai21.com/blog/announcing-jamba-model-family) to learn more.", + "modelVersionGroupId": null, + "contextLength": 256000, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Other", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "ai21/jamba-1-5-large", + "reasoningConfig": null, + "features": {}, + "endpoint": null, + "sorting": { + "topWeekly": 339, + "newest": 341, + "throughputHighToLow": 339, + "latencyLowToHigh": 341, + "pricingLowToHigh": 339, + "pricingHighToLow": 341 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai21.com/\\u0026size=256", "providers": [] }, { @@ -83030,27 +86212,28 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 341, - "newest": 340, + "topWeekly": 340, + "newest": 339, "throughputHighToLow": 340, - "latencyLowToHigh": 340, + "latencyLowToHigh": 339, "pricingLowToHigh": 340, "pricingHighToLow": 339 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { - "slug": "01-ai/yi-large-turbo", + "slug": "ai21/jamba-1-5-mini", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2024-08-02T00:00:00+00:00", + "createdAt": "2024-08-23T00:00:00+00:00", "hfUpdatedAt": null, - "name": "01.AI: Yi Large Turbo", - "shortName": "Yi Large Turbo", - "author": "01-ai", - "description": "The Yi Large Turbo model is a High Performance and Cost-Effectiveness model offering powerful capabilities at a competitive price.\n\nIt's ideal for a wide range of scenarios, including complex inference and high-quality text generation.\n\nCheck out the [launch announcement](https://01-ai.github.io/blog/01.ai-yi-large-llm-launch) to learn more.", + "name": "AI21: Jamba 1.5 Mini", + "shortName": "Jamba 1.5 Mini", + "author": "ai21", + "description": "Jamba 1.5 Mini is the world's first production-grade Mamba-based model, combining SSM and Transformer architectures for a 256K context window and high efficiency.\n\nIt works with 9 languages and can handle various writing and analysis tasks as well as or better than similar small models.\n\nThis model uses less computer memory and works faster with longer texts than previous designs.\n\nRead their [announcement](https://www.ai21.com/blog/announcing-jamba-model-family) to learn more.", "modelVersionGroupId": null, - "contextLength": 4096, + "contextLength": 256000, "inputModalities": [ "text" ], @@ -83058,25 +86241,26 @@ export default { "text" ], "hasTextOutput": true, - "group": "Yi", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "01-ai/yi-large-turbo", + "permaslug": "ai21/jamba-1-5-mini", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { - "topWeekly": 342, - "newest": 343, - "throughputHighToLow": 343, - "latencyLowToHigh": 343, - "pricingLowToHigh": 343, - "pricingHighToLow": 342 + "topWeekly": 341, + "newest": 340, + "throughputHighToLow": 341, + "latencyLowToHigh": 340, + "pricingLowToHigh": 341, + "pricingHighToLow": 340 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai21.com/\\u0026size=256", "providers": [] }, { @@ -83110,13 +86294,55 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 343, - "newest": 342, + "topWeekly": 342, + "newest": 343, "throughputHighToLow": 342, - "latencyLowToHigh": 342, + "latencyLowToHigh": 343, "pricingLowToHigh": 342, "pricingHighToLow": 343 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [] + }, + { + "slug": "01-ai/yi-large-turbo", + "hfSlug": null, + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2024-08-02T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "01.AI: Yi Large Turbo", + "shortName": "Yi Large Turbo", + "author": "01-ai", + "description": "The Yi Large Turbo model is a High Performance and Cost-Effectiveness model offering powerful capabilities at a competitive price.\n\nIt's ideal for a wide range of scenarios, including complex inference and high-quality text generation.\n\nCheck out the [launch announcement](https://01-ai.github.io/blog/01.ai-yi-large-llm-launch) to learn more.", + "modelVersionGroupId": null, + "contextLength": 4096, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Yi", + "instructType": null, + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "01-ai/yi-large-turbo", + "reasoningConfig": null, + "features": {}, + "endpoint": null, + "sorting": { + "topWeekly": 343, + "newest": 342, + "throughputHighToLow": 343, + "latencyLowToHigh": 342, + "pricingLowToHigh": 343, + "pricingHighToLow": 342 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -83158,6 +86384,7 @@ export default { "pricingLowToHigh": 344, "pricingHighToLow": 344 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -83199,6 +86426,7 @@ export default { "pricingLowToHigh": 345, "pricingHighToLow": 345 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [] }, { @@ -83243,6 +86471,7 @@ export default { "pricingLowToHigh": 346, "pricingHighToLow": 346 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -83287,6 +86516,7 @@ export default { "pricingLowToHigh": 347, "pricingHighToLow": 347 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [] }, { @@ -83331,6 +86561,7 @@ export default { "pricingLowToHigh": 348, "pricingHighToLow": 348 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nousresearch.com/\\u0026size=256", "providers": [] }, { @@ -83374,6 +86605,7 @@ export default { "pricingLowToHigh": 349, "pricingHighToLow": 349 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -83419,6 +86651,7 @@ export default { "pricingLowToHigh": 350, "pricingHighToLow": 350 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nvidia.com/\\u0026size=256", "providers": [] }, { @@ -83462,6 +86695,7 @@ export default { "pricingLowToHigh": 351, "pricingHighToLow": 351 }, + "authorIcon": "https://openrouter.ai/images/icons/Microsoft.svg", "providers": [] }, { @@ -83505,6 +86739,7 @@ export default { "pricingLowToHigh": 352, "pricingHighToLow": 352 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -83547,20 +86782,21 @@ export default { "pricingLowToHigh": 353, "pricingHighToLow": 353 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { - "slug": "perplexity/llama-3-sonar-large-32k-online", + "slug": "perplexity/llama-3-sonar-small-32k-chat", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-05-14T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Perplexity: Llama3 Sonar 70B Online", - "shortName": "Llama3 Sonar 70B Online", + "name": "Perplexity: Llama3 Sonar 8B", + "shortName": "Llama3 Sonar 8B", "author": "perplexity", - "description": "Llama3 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3-sonar-large-32k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", - "modelVersionGroupId": "e0349fb1-b84a-4f4f-9023-e7f1c1e73b33", - "contextLength": 28000, + "description": "Llama3 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is a normal offline LLM, but the [online version](/models/perplexity/llama-3-sonar-small-32k-online) of this model has Internet access.", + "modelVersionGroupId": "9e4cc81b-5f14-4987-9e40-74db1c86ecda", + "contextLength": 32768, "inputModalities": [ "text" ], @@ -83575,32 +86811,33 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "perplexity/llama-3-sonar-large-32k-online", + "permaslug": "perplexity/llama-3-sonar-small-32k-chat", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 354, - "newest": 355, - "throughputHighToLow": 355, - "latencyLowToHigh": 355, - "pricingLowToHigh": 355, - "pricingHighToLow": 354 + "newest": 356, + "throughputHighToLow": 354, + "latencyLowToHigh": 356, + "pricingLowToHigh": 354, + "pricingHighToLow": 356 }, + "authorIcon": "https://openrouter.ai/images/icons/Perplexity.svg", "providers": [] }, { - "slug": "deepseek/deepseek-chat-v2.5", - "hfSlug": "deepseek-ai/DeepSeek-V2.5", + "slug": "perplexity/llama-3-sonar-large-32k-online", + "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-05-14T00:00:00+00:00", "hfUpdatedAt": null, - "name": "DeepSeek V2.5", - "shortName": "DeepSeek V2.5", - "author": "deepseek", - "description": "DeepSeek-V2.5 is an upgraded version that combines DeepSeek-V2-Chat and DeepSeek-Coder-V2-Instruct. The new model integrates the general and coding abilities of the two previous versions. For model details, please visit [DeepSeek-V2 page](https://github.com/deepseek-ai/DeepSeek-V2) for more information.", - "modelVersionGroupId": null, - "contextLength": 128000, + "name": "Perplexity: Llama3 Sonar 70B Online", + "shortName": "Llama3 Sonar 70B Online", + "author": "perplexity", + "description": "Llama3 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3-sonar-large-32k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", + "modelVersionGroupId": "e0349fb1-b84a-4f4f-9023-e7f1c1e73b33", + "contextLength": 28000, "inputModalities": [ "text" ], @@ -83608,39 +86845,40 @@ export default { "text" ], "hasTextOutput": true, - "group": "Other", + "group": "Llama3", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "deepseek/deepseek-chat", + "permaslug": "perplexity/llama-3-sonar-large-32k-online", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 355, - "newest": 358, - "throughputHighToLow": 358, - "latencyLowToHigh": 358, - "pricingLowToHigh": 358, - "pricingHighToLow": 355 + "newest": 354, + "throughputHighToLow": 355, + "latencyLowToHigh": 354, + "pricingLowToHigh": 355, + "pricingHighToLow": 354 }, + "authorIcon": "https://openrouter.ai/images/icons/Perplexity.svg", "providers": [] }, { - "slug": "perplexity/llama-3-sonar-small-32k-chat", + "slug": "perplexity/llama-3-sonar-small-32k-online", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-05-14T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Perplexity: Llama3 Sonar 8B", - "shortName": "Llama3 Sonar 8B", + "name": "Perplexity: Llama3 Sonar 8B Online", + "shortName": "Llama3 Sonar 8B Online", "author": "perplexity", - "description": "Llama3 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is a normal offline LLM, but the [online version](/models/perplexity/llama-3-sonar-small-32k-online) of this model has Internet access.", + "description": "Llama3 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3-sonar-small-32k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", "modelVersionGroupId": "9e4cc81b-5f14-4987-9e40-74db1c86ecda", - "contextLength": 32768, + "contextLength": 28000, "inputModalities": [ "text" ], @@ -83655,32 +86893,33 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "perplexity/llama-3-sonar-small-32k-chat", + "permaslug": "perplexity/llama-3-sonar-small-32k-online", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 356, - "newest": 354, - "throughputHighToLow": 354, - "latencyLowToHigh": 354, - "pricingLowToHigh": 354, - "pricingHighToLow": 356 + "newest": 357, + "throughputHighToLow": 356, + "latencyLowToHigh": 357, + "pricingLowToHigh": 356, + "pricingHighToLow": 357 }, + "authorIcon": "https://openrouter.ai/images/icons/Perplexity.svg", "providers": [] }, { - "slug": "perplexity/llama-3-sonar-small-32k-online", + "slug": "perplexity/llama-3-sonar-large-32k-chat", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-05-14T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Perplexity: Llama3 Sonar 8B Online", - "shortName": "Llama3 Sonar 8B Online", + "name": "Perplexity: Llama3 Sonar 70B", + "shortName": "Llama3 Sonar 70B", "author": "perplexity", - "description": "Llama3 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is the online version of the [offline chat model](/models/perplexity/llama-3-sonar-small-32k-chat). It is focused on delivering helpful, up-to-date, and factual responses. #online", - "modelVersionGroupId": "9e4cc81b-5f14-4987-9e40-74db1c86ecda", - "contextLength": 28000, + "description": "Llama3 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is a normal offline LLM, but the [online version](/models/perplexity/llama-3-sonar-large-32k-online) of this model has Internet access.", + "modelVersionGroupId": "e0349fb1-b84a-4f4f-9023-e7f1c1e73b33", + "contextLength": 32768, "inputModalities": [ "text" ], @@ -83695,32 +86934,33 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "perplexity/llama-3-sonar-small-32k-online", + "permaslug": "perplexity/llama-3-sonar-large-32k-chat", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 357, - "newest": 356, - "throughputHighToLow": 356, - "latencyLowToHigh": 356, - "pricingLowToHigh": 356, - "pricingHighToLow": 357 + "newest": 358, + "throughputHighToLow": 357, + "latencyLowToHigh": 358, + "pricingLowToHigh": 357, + "pricingHighToLow": 358 }, + "authorIcon": "https://openrouter.ai/images/icons/Perplexity.svg", "providers": [] }, { - "slug": "perplexity/llama-3-sonar-large-32k-chat", - "hfSlug": null, + "slug": "deepseek/deepseek-chat-v2.5", + "hfSlug": "deepseek-ai/DeepSeek-V2.5", "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-05-14T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Perplexity: Llama3 Sonar 70B", - "shortName": "Llama3 Sonar 70B", - "author": "perplexity", - "description": "Llama3 Sonar is Perplexity's latest model family. It surpasses their earlier Sonar models in cost-efficiency, speed, and performance.\n\nThis is a normal offline LLM, but the [online version](/models/perplexity/llama-3-sonar-large-32k-online) of this model has Internet access.", - "modelVersionGroupId": "e0349fb1-b84a-4f4f-9023-e7f1c1e73b33", - "contextLength": 32768, + "name": "DeepSeek V2.5", + "shortName": "DeepSeek V2.5", + "author": "deepseek", + "description": "DeepSeek-V2.5 is an upgraded version that combines DeepSeek-V2-Chat and DeepSeek-Coder-V2-Instruct. The new model integrates the general and coding abilities of the two previous versions. For model details, please visit [DeepSeek-V2 page](https://github.com/deepseek-ai/DeepSeek-V2) for more information.", + "modelVersionGroupId": null, + "contextLength": 128000, "inputModalities": [ "text" ], @@ -83728,38 +86968,39 @@ export default { "text" ], "hasTextOutput": true, - "group": "Llama3", + "group": "Other", "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "perplexity/llama-3-sonar-large-32k-chat", + "permaslug": "deepseek/deepseek-chat", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 358, - "newest": 357, - "throughputHighToLow": 357, - "latencyLowToHigh": 357, - "pricingLowToHigh": 357, - "pricingHighToLow": 358 + "newest": 355, + "throughputHighToLow": 358, + "latencyLowToHigh": 355, + "pricingLowToHigh": 358, + "pricingHighToLow": 355 }, + "authorIcon": "https://openrouter.ai/images/icons/DeepSeek.png", "providers": [] }, { - "slug": "meta-llama/llama-3-8b", - "hfSlug": "meta-llama/Meta-Llama-3-8b", + "slug": "meta-llama/llama-3-70b", + "hfSlug": "meta-llama/Meta-Llama-3-70b", "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-05-13T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3 8B (Base)", - "shortName": "Llama 3 8B (Base)", + "name": "Meta: Llama 3 70B (Base)", + "shortName": "Llama 3 70B (Base)", "author": "meta-llama", - "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This is the base 8B pre-trained version.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "modelVersionGroupId": "803c32ed-9861-4abf-b5da-7d9c9e6dcf04", + "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This is the base 70B pre-trained version.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "modelVersionGroupId": "397604e2-45fa-454e-a85d-9921f5138747", "contextLength": 8192, "inputModalities": [ "text" @@ -83779,31 +87020,32 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3-8b", + "permaslug": "meta-llama/llama-3-70b", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 359, "newest": 360, - "throughputHighToLow": 360, + "throughputHighToLow": 359, "latencyLowToHigh": 360, - "pricingLowToHigh": 360, - "pricingHighToLow": 359 + "pricingLowToHigh": 359, + "pricingHighToLow": 360 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [] }, { - "slug": "meta-llama/llama-3-70b", - "hfSlug": "meta-llama/Meta-Llama-3-70b", + "slug": "meta-llama/llama-3-8b", + "hfSlug": "meta-llama/Meta-Llama-3-8b", "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-05-13T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Meta: Llama 3 70B (Base)", - "shortName": "Llama 3 70B (Base)", + "name": "Meta: Llama 3 8B (Base)", + "shortName": "Llama 3 8B (Base)", "author": "meta-llama", - "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This is the base 70B pre-trained version.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", - "modelVersionGroupId": "397604e2-45fa-454e-a85d-9921f5138747", + "description": "Meta's latest class of model (Llama 3) launched with a variety of sizes & flavors. This is the base 8B pre-trained version.\n\nIt has demonstrated strong performance compared to leading closed-source models in human evaluations.\n\nTo read more about the model release, [click here](https://ai.meta.com/blog/meta-llama-3/). Usage of this model is subject to [Meta's Acceptable Use Policy](https://llama.meta.com/llama3/use-policy/).", + "modelVersionGroupId": "803c32ed-9861-4abf-b5da-7d9c9e6dcf04", "contextLength": 8192, "inputModalities": [ "text" @@ -83823,18 +87065,19 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "meta-llama/llama-3-70b", + "permaslug": "meta-llama/llama-3-8b", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 360, "newest": 359, - "throughputHighToLow": 359, + "throughputHighToLow": 360, "latencyLowToHigh": 359, - "pricingLowToHigh": 359, - "pricingHighToLow": 360 + "pricingLowToHigh": 360, + "pricingHighToLow": 359 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [] }, { @@ -83880,18 +87123,19 @@ export default { "pricingLowToHigh": 361, "pricingHighToLow": 361 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { - "slug": "qwen/qwen-110b-chat", - "hfSlug": "Qwen/Qwen1.5-110B-Chat", + "slug": "qwen/qwen-4b-chat", + "hfSlug": "Qwen/Qwen1.5-4B-Chat", "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-05-09T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen 1.5 110B Chat", - "shortName": "Qwen 1.5 110B Chat", + "name": "Qwen 1.5 4B Chat", + "shortName": "Qwen 1.5 4B Chat", "author": "qwen", - "description": "Qwen1.5 110B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:\n\n- Significant performance improvement in human preference for chat models\n- Multilingual support of both base and chat models\n- Stable support of 32K context length for models of all sizes\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "description": "Qwen1.5 4B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:\n\n- Significant performance improvement in human preference for chat models\n- Multilingual support of both base and chat models\n- Stable support of 32K context length for models of all sizes\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ @@ -83912,30 +87156,31 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-110b-chat", + "permaslug": "qwen/qwen-4b-chat", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 362, - "newest": 363, - "throughputHighToLow": 363, - "latencyLowToHigh": 363, - "pricingLowToHigh": 363, - "pricingHighToLow": 362 + "newest": 367, + "throughputHighToLow": 362, + "latencyLowToHigh": 367, + "pricingLowToHigh": 362, + "pricingHighToLow": 367 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [] }, { - "slug": "qwen/qwen-72b-chat", - "hfSlug": "Qwen/Qwen1.5-72B-Chat", + "slug": "qwen/qwen-110b-chat", + "hfSlug": "Qwen/Qwen1.5-110B-Chat", "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-05-09T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen 1.5 72B Chat", - "shortName": "Qwen 1.5 72B Chat", + "name": "Qwen 1.5 110B Chat", + "shortName": "Qwen 1.5 110B Chat", "author": "qwen", - "description": "Qwen1.5 72B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:\n\n- Significant performance improvement in human preference for chat models\n- Multilingual support of both base and chat models\n- Stable support of 32K context length for models of all sizes\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "description": "Qwen1.5 110B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:\n\n- Significant performance improvement in human preference for chat models\n- Multilingual support of both base and chat models\n- Stable support of 32K context length for models of all sizes\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ @@ -83956,30 +87201,31 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-72b-chat", + "permaslug": "qwen/qwen-110b-chat", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 363, - "newest": 366, - "throughputHighToLow": 366, - "latencyLowToHigh": 366, - "pricingLowToHigh": 366, - "pricingHighToLow": 363 + "newest": 362, + "throughputHighToLow": 363, + "latencyLowToHigh": 362, + "pricingLowToHigh": 363, + "pricingHighToLow": 362 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [] }, { - "slug": "qwen/qwen-32b-chat", - "hfSlug": "Qwen/Qwen1.5-32B-Chat", + "slug": "qwen/qwen-14b-chat", + "hfSlug": "Qwen/Qwen1.5-14B-Chat", "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-05-09T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen 1.5 32B Chat", - "shortName": "Qwen 1.5 32B Chat", + "name": "Qwen 1.5 14B Chat", + "shortName": "Qwen 1.5 14B Chat", "author": "qwen", - "description": "Qwen1.5 32B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:\n\n- Significant performance improvement in human preference for chat models\n- Multilingual support of both base and chat models\n- Stable support of 32K context length for models of all sizes\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "description": "Qwen1.5 14B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:\n\n- Significant performance improvement in human preference for chat models\n- Multilingual support of both base and chat models\n- Stable support of 32K context length for models of all sizes\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ @@ -84000,30 +87246,31 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-32b-chat", + "permaslug": "qwen/qwen-14b-chat", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 364, - "newest": 367, - "throughputHighToLow": 367, - "latencyLowToHigh": 367, - "pricingLowToHigh": 367, - "pricingHighToLow": 364 + "newest": 365, + "throughputHighToLow": 364, + "latencyLowToHigh": 365, + "pricingLowToHigh": 364, + "pricingHighToLow": 365 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [] }, { - "slug": "qwen/qwen-14b-chat", - "hfSlug": "Qwen/Qwen1.5-14B-Chat", + "slug": "qwen/qwen-7b-chat", + "hfSlug": "Qwen/Qwen1.5-7B-Chat", "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-05-09T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen 1.5 14B Chat", - "shortName": "Qwen 1.5 14B Chat", + "name": "Qwen 1.5 7B Chat", + "shortName": "Qwen 1.5 7B Chat", "author": "qwen", - "description": "Qwen1.5 14B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:\n\n- Significant performance improvement in human preference for chat models\n- Multilingual support of both base and chat models\n- Stable support of 32K context length for models of all sizes\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "description": "Qwen1.5 7B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:\n\n- Significant performance improvement in human preference for chat models\n- Multilingual support of both base and chat models\n- Stable support of 32K context length for models of all sizes\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ @@ -84044,30 +87291,31 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-14b-chat", + "permaslug": "qwen/qwen-7b-chat", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 365, - "newest": 364, - "throughputHighToLow": 364, - "latencyLowToHigh": 364, - "pricingLowToHigh": 364, - "pricingHighToLow": 365 + "newest": 366, + "throughputHighToLow": 365, + "latencyLowToHigh": 366, + "pricingLowToHigh": 365, + "pricingHighToLow": 366 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [] }, { - "slug": "qwen/qwen-7b-chat", - "hfSlug": "Qwen/Qwen1.5-7B-Chat", + "slug": "qwen/qwen-72b-chat", + "hfSlug": "Qwen/Qwen1.5-72B-Chat", "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-05-09T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen 1.5 7B Chat", - "shortName": "Qwen 1.5 7B Chat", + "name": "Qwen 1.5 72B Chat", + "shortName": "Qwen 1.5 72B Chat", "author": "qwen", - "description": "Qwen1.5 7B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:\n\n- Significant performance improvement in human preference for chat models\n- Multilingual support of both base and chat models\n- Stable support of 32K context length for models of all sizes\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "description": "Qwen1.5 72B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:\n\n- Significant performance improvement in human preference for chat models\n- Multilingual support of both base and chat models\n- Stable support of 32K context length for models of all sizes\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ @@ -84088,30 +87336,31 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-7b-chat", + "permaslug": "qwen/qwen-72b-chat", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 366, - "newest": 365, - "throughputHighToLow": 365, - "latencyLowToHigh": 365, - "pricingLowToHigh": 365, - "pricingHighToLow": 366 + "newest": 363, + "throughputHighToLow": 366, + "latencyLowToHigh": 363, + "pricingLowToHigh": 366, + "pricingHighToLow": 363 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [] }, { - "slug": "qwen/qwen-4b-chat", - "hfSlug": "Qwen/Qwen1.5-4B-Chat", + "slug": "qwen/qwen-32b-chat", + "hfSlug": "Qwen/Qwen1.5-32B-Chat", "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-05-09T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Qwen 1.5 4B Chat", - "shortName": "Qwen 1.5 4B Chat", + "name": "Qwen 1.5 32B Chat", + "shortName": "Qwen 1.5 32B Chat", "author": "qwen", - "description": "Qwen1.5 4B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:\n\n- Significant performance improvement in human preference for chat models\n- Multilingual support of both base and chat models\n- Stable support of 32K context length for models of all sizes\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", + "description": "Qwen1.5 32B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:\n\n- Significant performance improvement in human preference for chat models\n- Multilingual support of both base and chat models\n- Stable support of 32K context length for models of all sizes\n\nFor more details, see this [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).\n\nUsage of this model is subject to [Tongyi Qianwen LICENSE AGREEMENT](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE).", "modelVersionGroupId": null, "contextLength": 32768, "inputModalities": [ @@ -84132,18 +87381,19 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "qwen/qwen-4b-chat", + "permaslug": "qwen/qwen-32b-chat", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 367, - "newest": 362, - "throughputHighToLow": 362, - "latencyLowToHigh": 362, - "pricingLowToHigh": 362, - "pricingHighToLow": 367 + "newest": 364, + "throughputHighToLow": 367, + "latencyLowToHigh": 364, + "pricingLowToHigh": 367, + "pricingHighToLow": 364 }, + "authorIcon": "https://openrouter.ai/images/icons/Qwen.png", "providers": [] }, { @@ -84188,6 +87438,7 @@ export default { "pricingLowToHigh": 368, "pricingHighToLow": 368 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -84232,6 +87483,7 @@ export default { "pricingLowToHigh": 369, "pricingHighToLow": 369 }, + "authorIcon": "https://openrouter.ai/images/icons/Fireworks.png", "providers": [] }, { @@ -84275,6 +87527,7 @@ export default { "pricingLowToHigh": 370, "pricingHighToLow": 370 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://api.lynn.app/\\u0026size=256", "providers": [] }, { @@ -84318,6 +87571,7 @@ export default { "pricingLowToHigh": 371, "pricingHighToLow": 371 }, + "authorIcon": "https://openrouter.ai/images/icons/Microsoft.svg", "providers": [] }, { @@ -84361,6 +87615,7 @@ export default { "pricingLowToHigh": 372, "pricingHighToLow": 372 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -84405,6 +87660,7 @@ export default { "pricingLowToHigh": 373, "pricingHighToLow": 373 }, + "authorIcon": "https://openrouter.ai/images/icons/Mistral.png", "providers": [] }, { @@ -84449,6 +87705,7 @@ export default { "pricingLowToHigh": 374, "pricingHighToLow": 374 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://databricks.com/\\u0026size=256", "providers": [] }, { @@ -84493,6 +87750,7 @@ export default { "pricingLowToHigh": 375, "pricingHighToLow": 375 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [] }, { @@ -84537,6 +87795,7 @@ export default { "pricingLowToHigh": 376, "pricingHighToLow": 376 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nousresearch.com/\\u0026size=256", "providers": [] }, { @@ -84580,6 +87839,7 @@ export default { "pricingLowToHigh": 377, "pricingHighToLow": 377 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [] }, { @@ -84623,6 +87883,7 @@ export default { "pricingLowToHigh": 378, "pricingHighToLow": 378 }, + "authorIcon": "https://openrouter.ai/images/icons/Recursal.jpg", "providers": [] }, { @@ -84663,6 +87924,7 @@ export default { "pricingLowToHigh": 379, "pricingHighToLow": 379 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -84707,6 +87969,7 @@ export default { "pricingLowToHigh": 380, "pricingHighToLow": 380 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nousresearch.com/\\u0026size=256", "providers": [] }, { @@ -84750,6 +88013,7 @@ export default { "pricingLowToHigh": 381, "pricingHighToLow": 381 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -84793,20 +88057,21 @@ export default { "pricingLowToHigh": 382, "pricingHighToLow": 382 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { - "slug": "nousresearch/nous-hermes-yi-34b", - "hfSlug": "NousResearch/Nous-Hermes-2-Yi-34B", + "slug": "neversleep/noromaid-mixtral-8x7b-instruct", + "hfSlug": "NeverSleep/Noromaid-v0.1-mixtral-8x7b-Instruct-v3", "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-01-02T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Nous: Hermes 2 Yi 34B", - "shortName": "Hermes 2 Yi 34B", - "author": "nousresearch", - "description": "Nous Hermes 2 Yi 34B was trained on 1,000,000 entries of primarily GPT-4 generated data, as well as other high quality data from open datasets across the AI landscape.\n\nNous-Hermes 2 on Yi 34B outperforms all Nous-Hermes & Open-Hermes models of the past, achieving new heights in all benchmarks for a Nous Research LLM as well as surpassing many popular finetunes.", + "name": "Noromaid Mixtral 8x7B Instruct", + "shortName": "Noromaid Mixtral 8x7B Instruct", + "author": "neversleep", + "description": "This model was trained for 8h(v1) + 8h(v2) + 12h(v3) on customized modified datasets, focusing on RP, uncensoring, and a modified version of the Alpaca prompting (that was already used in LimaRP), which should be at the same conversational level as ChatLM or Llama2-Chat without adding any additional special tokens.", "modelVersionGroupId": null, - "contextLength": 4096, + "contextLength": 8000, "inputModalities": [ "text" ], @@ -84814,43 +88079,43 @@ export default { "text" ], "hasTextOutput": true, - "group": "Yi", - "instructType": "chatml", + "group": "Mistral", + "instructType": "alpaca-modif", "defaultSystem": null, "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" + "###", + "" ], "hidden": false, "router": null, - "warningMessage": null, - "permaslug": "nousresearch/nous-hermes-yi-34b", + "warningMessage": "Due to low usage, this model is deprecated and will be removed from the API on July 22, 2024.", + "permaslug": "neversleep/noromaid-mixtral-8x7b-instruct", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 383, "newest": 384, - "throughputHighToLow": 384, + "throughputHighToLow": 383, "latencyLowToHigh": 384, - "pricingLowToHigh": 384, - "pricingHighToLow": 383 + "pricingLowToHigh": 383, + "pricingHighToLow": 384 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { - "slug": "neversleep/noromaid-mixtral-8x7b-instruct", - "hfSlug": "NeverSleep/Noromaid-v0.1-mixtral-8x7b-Instruct-v3", + "slug": "nousresearch/nous-hermes-yi-34b", + "hfSlug": "NousResearch/Nous-Hermes-2-Yi-34B", "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2024-01-02T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Noromaid Mixtral 8x7B Instruct", - "shortName": "Noromaid Mixtral 8x7B Instruct", - "author": "neversleep", - "description": "This model was trained for 8h(v1) + 8h(v2) + 12h(v3) on customized modified datasets, focusing on RP, uncensoring, and a modified version of the Alpaca prompting (that was already used in LimaRP), which should be at the same conversational level as ChatLM or Llama2-Chat without adding any additional special tokens.", + "name": "Nous: Hermes 2 Yi 34B", + "shortName": "Hermes 2 Yi 34B", + "author": "nousresearch", + "description": "Nous Hermes 2 Yi 34B was trained on 1,000,000 entries of primarily GPT-4 generated data, as well as other high quality data from open datasets across the AI landscape.\n\nNous-Hermes 2 on Yi 34B outperforms all Nous-Hermes & Open-Hermes models of the past, achieving new heights in all benchmarks for a Nous Research LLM as well as surpassing many popular finetunes.", "modelVersionGroupId": null, - "contextLength": 8000, + "contextLength": 4096, "inputModalities": [ "text" ], @@ -84858,28 +88123,30 @@ export default { "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": "alpaca-modif", + "group": "Yi", + "instructType": "chatml", "defaultSystem": null, "defaultStops": [ - "###", - "" + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" ], "hidden": false, "router": null, - "warningMessage": "Due to low usage, this model is deprecated and will be removed from the API on July 22, 2024.", - "permaslug": "neversleep/noromaid-mixtral-8x7b-instruct", + "warningMessage": null, + "permaslug": "nousresearch/nous-hermes-yi-34b", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 384, "newest": 383, - "throughputHighToLow": 383, + "throughputHighToLow": 384, "latencyLowToHigh": 383, - "pricingLowToHigh": 383, - "pricingHighToLow": 384 + "pricingLowToHigh": 384, + "pricingHighToLow": 383 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nousresearch.com/\\u0026size=256", "providers": [] }, { @@ -84924,6 +88191,7 @@ export default { "pricingLowToHigh": 385, "pricingHighToLow": 385 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -84967,6 +88235,7 @@ export default { "pricingLowToHigh": 386, "pricingHighToLow": 386 }, + "authorIcon": "https://openrouter.ai/images/icons/Recursal.jpg", "providers": [] }, { @@ -85010,6 +88279,7 @@ export default { "pricingLowToHigh": 387, "pricingHighToLow": 387 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -85053,6 +88323,7 @@ export default { "pricingLowToHigh": 388, "pricingHighToLow": 388 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -85093,6 +88364,7 @@ export default { "pricingLowToHigh": 389, "pricingHighToLow": 389 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -85136,130 +88408,49 @@ export default { "pricingLowToHigh": 390, "pricingHighToLow": 390 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { - "slug": "01-ai/yi-34b", - "hfSlug": "01-ai/Yi-34B", + "slug": "nousresearch/nous-hermes-2-vision-7b", + "hfSlug": "NousResearch/Nous-Hermes-2-Vision-Alpha", "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2023-12-07T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Yi 34B (base)", - "shortName": "Yi 34B (base)", - "author": "01-ai", - "description": "The Yi series models are large language models trained from scratch by developers at [01.AI](https://01.ai/). This is the base 34B parameter model.", + "name": "Nous: Hermes 2 Vision 7B (alpha)", + "shortName": "Hermes 2 Vision 7B (alpha)", + "author": "nousresearch", + "description": "This vision-language model builds on innovations from the popular [OpenHermes-2.5](/models/teknium/openhermes-2.5-mistral-7b) model, by Teknium. It adds vision support, and is trained on a custom dataset enriched with function calling\n\nThis project is led by [qnguyen3](https://twitter.com/stablequan) and [teknium](https://twitter.com/Teknium1).\n\n#multimodal", "modelVersionGroupId": null, "contextLength": 4096, "inputModalities": [ - "text" + "text", + "image" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Yi", - "instructType": "none", + "group": "Mistral", + "instructType": null, "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "01-ai/yi-34b", + "permaslug": "nousresearch/nous-hermes-2-vision-7b", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 391, - "newest": 393, - "throughputHighToLow": 393, - "latencyLowToHigh": 393, - "pricingLowToHigh": 393, - "pricingHighToLow": 391 - }, - "providers": [] - }, - { - "slug": "01-ai/yi-34b-chat", - "hfSlug": "01-ai/Yi-34B-Chat", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-12-07T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Yi 34B Chat", - "shortName": "Yi 34B Chat", - "author": "01-ai", - "description": "The Yi series models are large language models trained from scratch by developers at [01.AI](https://01.ai/). This 34B parameter model has been instruct-tuned for chat.", - "modelVersionGroupId": null, - "contextLength": 4096, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Yi", - "instructType": "chatml", - "defaultSystem": null, - "defaultStops": [ - "<|im_start|>", - "<|im_end|>", - "<|endoftext|>" - ], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "01-ai/yi-34b-chat", - "reasoningConfig": null, - "features": {}, - "endpoint": null, - "sorting": { - "topWeekly": 392, "newest": 395, - "throughputHighToLow": 395, + "throughputHighToLow": 391, "latencyLowToHigh": 395, - "pricingLowToHigh": 395, - "pricingHighToLow": 392 - }, - "providers": [] - }, - { - "slug": "01-ai/yi-6b", - "hfSlug": "01-ai/Yi-6B", - "updatedAt": "2025-03-28T03:20:30.853469+00:00", - "createdAt": "2023-12-07T00:00:00+00:00", - "hfUpdatedAt": null, - "name": "Yi 6B (base)", - "shortName": "Yi 6B (base)", - "author": "01-ai", - "description": "The Yi series models are large language models trained from scratch by developers at [01.AI](https://01.ai/). This is the base 6B parameter model.", - "modelVersionGroupId": null, - "contextLength": 4096, - "inputModalities": [ - "text" - ], - "outputModalities": [ - "text" - ], - "hasTextOutput": true, - "group": "Yi", - "instructType": "none", - "defaultSystem": null, - "defaultStops": [], - "hidden": false, - "router": null, - "warningMessage": null, - "permaslug": "01-ai/yi-6b", - "reasoningConfig": null, - "features": {}, - "endpoint": null, - "sorting": { - "topWeekly": 393, - "newest": 394, - "throughputHighToLow": 394, - "latencyLowToHigh": 394, - "pricingLowToHigh": 394, - "pricingHighToLow": 393 + "pricingLowToHigh": 391, + "pricingHighToLow": 395 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nousresearch.com/\\u0026size=256", "providers": [] }, { @@ -85296,54 +88487,141 @@ export default { "features": {}, "endpoint": null, "sorting": { - "topWeekly": 394, - "newest": 392, + "topWeekly": 392, + "newest": 394, "throughputHighToLow": 392, - "latencyLowToHigh": 392, + "latencyLowToHigh": 394, "pricingLowToHigh": 392, "pricingHighToLow": 394 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { - "slug": "nousresearch/nous-hermes-2-vision-7b", - "hfSlug": "NousResearch/Nous-Hermes-2-Vision-Alpha", + "slug": "01-ai/yi-34b", + "hfSlug": "01-ai/Yi-34B", "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2023-12-07T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Nous: Hermes 2 Vision 7B (alpha)", - "shortName": "Hermes 2 Vision 7B (alpha)", - "author": "nousresearch", - "description": "This vision-language model builds on innovations from the popular [OpenHermes-2.5](/models/teknium/openhermes-2.5-mistral-7b) model, by Teknium. It adds vision support, and is trained on a custom dataset enriched with function calling\n\nThis project is led by [qnguyen3](https://twitter.com/stablequan) and [teknium](https://twitter.com/Teknium1).\n\n#multimodal", + "name": "Yi 34B (base)", + "shortName": "Yi 34B (base)", + "author": "01-ai", + "description": "The Yi series models are large language models trained from scratch by developers at [01.AI](https://01.ai/). This is the base 34B parameter model.", "modelVersionGroupId": null, "contextLength": 4096, "inputModalities": [ - "text", - "image" + "text" ], "outputModalities": [ "text" ], "hasTextOutput": true, - "group": "Mistral", - "instructType": null, + "group": "Yi", + "instructType": "none", "defaultSystem": null, "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "nousresearch/nous-hermes-2-vision-7b", + "permaslug": "01-ai/yi-34b", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { - "topWeekly": 395, + "topWeekly": 393, "newest": 391, - "throughputHighToLow": 391, + "throughputHighToLow": 393, "latencyLowToHigh": 391, - "pricingLowToHigh": 391, - "pricingHighToLow": 395 + "pricingLowToHigh": 393, + "pricingHighToLow": 391 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [] + }, + { + "slug": "01-ai/yi-6b", + "hfSlug": "01-ai/Yi-6B", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2023-12-07T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Yi 6B (base)", + "shortName": "Yi 6B (base)", + "author": "01-ai", + "description": "The Yi series models are large language models trained from scratch by developers at [01.AI](https://01.ai/). This is the base 6B parameter model.", + "modelVersionGroupId": null, + "contextLength": 4096, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Yi", + "instructType": "none", + "defaultSystem": null, + "defaultStops": [], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "01-ai/yi-6b", + "reasoningConfig": null, + "features": {}, + "endpoint": null, + "sorting": { + "topWeekly": 394, + "newest": 393, + "throughputHighToLow": 394, + "latencyLowToHigh": 393, + "pricingLowToHigh": 394, + "pricingHighToLow": 393 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", + "providers": [] + }, + { + "slug": "01-ai/yi-34b-chat", + "hfSlug": "01-ai/Yi-34B-Chat", + "updatedAt": "2025-03-28T03:20:30.853469+00:00", + "createdAt": "2023-12-07T00:00:00+00:00", + "hfUpdatedAt": null, + "name": "Yi 34B Chat", + "shortName": "Yi 34B Chat", + "author": "01-ai", + "description": "The Yi series models are large language models trained from scratch by developers at [01.AI](https://01.ai/). This 34B parameter model has been instruct-tuned for chat.", + "modelVersionGroupId": null, + "contextLength": 4096, + "inputModalities": [ + "text" + ], + "outputModalities": [ + "text" + ], + "hasTextOutput": true, + "group": "Yi", + "instructType": "chatml", + "defaultSystem": null, + "defaultStops": [ + "<|im_start|>", + "<|im_end|>", + "<|endoftext|>" + ], + "hidden": false, + "router": null, + "warningMessage": null, + "permaslug": "01-ai/yi-34b-chat", + "reasoningConfig": null, + "features": {}, + "endpoint": null, + "sorting": { + "topWeekly": 395, + "newest": 392, + "throughputHighToLow": 395, + "latencyLowToHigh": 392, + "pricingLowToHigh": 395, + "pricingHighToLow": 392 + }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -85387,6 +88665,7 @@ export default { "pricingLowToHigh": 396, "pricingHighToLow": 396 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://openrouter.ai/\\u0026size=256", "providers": [] }, { @@ -85430,6 +88709,7 @@ export default { "pricingLowToHigh": 397, "pricingHighToLow": 397 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nousresearch.com/\\u0026size=256", "providers": [] }, { @@ -85473,6 +88753,7 @@ export default { "pricingLowToHigh": 398, "pricingHighToLow": 398 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -85515,6 +88796,7 @@ export default { "pricingLowToHigh": 399, "pricingHighToLow": 399 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -85558,6 +88840,7 @@ export default { "pricingLowToHigh": 400, "pricingHighToLow": 400 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -85600,6 +88883,7 @@ export default { "pricingLowToHigh": 401, "pricingHighToLow": 401 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [] }, { @@ -85644,6 +88928,7 @@ export default { "pricingLowToHigh": 402, "pricingHighToLow": 402 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -85685,6 +88970,7 @@ export default { "pricingLowToHigh": 403, "pricingHighToLow": 403 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -85728,6 +89014,7 @@ export default { "pricingLowToHigh": 404, "pricingHighToLow": 404 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nousresearch.com/\\u0026size=256", "providers": [] }, { @@ -85769,6 +89056,7 @@ export default { "pricingLowToHigh": 405, "pricingHighToLow": 405 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [] }, { @@ -85812,6 +89100,7 @@ export default { "pricingLowToHigh": 406, "pricingHighToLow": 406 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -85852,18 +89141,19 @@ export default { "pricingLowToHigh": 407, "pricingHighToLow": 407 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://openrouter.ai/\\u0026size=256", "providers": [] }, { - "slug": "google/palm-2-chat-bison-32k", + "slug": "google/palm-2-codechat-bison-32k", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2023-11-03T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Google: PaLM 2 Chat 32k", - "shortName": "PaLM 2 Chat 32k", + "name": "Google: PaLM 2 Code Chat 32k", + "shortName": "PaLM 2 Code Chat 32k", "author": "google", - "description": "PaLM 2 is a language model by Google with improved multilingual, reasoning and coding capabilities.", + "description": "PaLM 2 fine-tuned for chatbot conversations that help with code-related questions.", "modelVersionGroupId": null, "contextLength": 32760, "inputModalities": [ @@ -85880,30 +89170,31 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "google/palm-2-chat-bison-32k", + "permaslug": "google/palm-2-codechat-bison-32k", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 408, "newest": 409, - "throughputHighToLow": 409, + "throughputHighToLow": 408, "latencyLowToHigh": 409, - "pricingLowToHigh": 409, - "pricingHighToLow": 408 + "pricingLowToHigh": 408, + "pricingHighToLow": 409 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [] }, { - "slug": "google/palm-2-codechat-bison-32k", + "slug": "google/palm-2-chat-bison-32k", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2023-11-03T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Google: PaLM 2 Code Chat 32k", - "shortName": "PaLM 2 Code Chat 32k", + "name": "Google: PaLM 2 Chat 32k", + "shortName": "PaLM 2 Chat 32k", "author": "google", - "description": "PaLM 2 fine-tuned for chatbot conversations that help with code-related questions.", + "description": "PaLM 2 is a language model by Google with improved multilingual, reasoning and coding capabilities.", "modelVersionGroupId": null, "contextLength": 32760, "inputModalities": [ @@ -85920,18 +89211,19 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "google/palm-2-codechat-bison-32k", + "permaslug": "google/palm-2-chat-bison-32k", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 409, "newest": 408, - "throughputHighToLow": 408, + "throughputHighToLow": 409, "latencyLowToHigh": 408, - "pricingLowToHigh": 408, - "pricingHighToLow": 409 + "pricingLowToHigh": 409, + "pricingHighToLow": 408 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [] }, { @@ -85976,6 +89268,7 @@ export default { "pricingLowToHigh": 410, "pricingHighToLow": 410 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -86020,6 +89313,7 @@ export default { "pricingLowToHigh": 411, "pricingHighToLow": 411 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -86063,6 +89357,7 @@ export default { "pricingLowToHigh": 412, "pricingHighToLow": 412 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nousresearch.com/\\u0026size=256", "providers": [] }, { @@ -86106,6 +89401,7 @@ export default { "pricingLowToHigh": 413, "pricingHighToLow": 413 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -86149,6 +89445,7 @@ export default { "pricingLowToHigh": 414, "pricingHighToLow": 414 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -86192,18 +89489,19 @@ export default { "pricingLowToHigh": 415, "pricingHighToLow": 415 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [] }, { - "slug": "phind/phind-codellama-34b", - "hfSlug": "Phind/Phind-CodeLlama-34B-v2", + "slug": "nousresearch/nous-hermes-llama2-13b", + "hfSlug": "NousResearch/Nous-Hermes-Llama2-13b", "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2023-08-20T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Phind: CodeLlama 34B v2", - "shortName": "CodeLlama 34B v2", - "author": "phind", - "description": "A fine-tune of CodeLlama-34B on an internal dataset that helps it exceed GPT-4 on some benchmarks, including HumanEval.", + "name": "Nous: Hermes 13B", + "shortName": "Hermes 13B", + "author": "nousresearch", + "description": "A state-of-the-art language model fine-tuned on over 300k instructions by Nous Research, with Teknium and Emozilla leading the fine tuning process.", "modelVersionGroupId": null, "contextLength": 4096, "inputModalities": [ @@ -86223,30 +89521,31 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "phind/phind-codellama-34b", + "permaslug": "nousresearch/nous-hermes-llama2-13b", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 416, "newest": 417, - "throughputHighToLow": 417, + "throughputHighToLow": 416, "latencyLowToHigh": 417, - "pricingLowToHigh": 417, - "pricingHighToLow": 416 + "pricingLowToHigh": 416, + "pricingHighToLow": 417 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://nousresearch.com/\\u0026size=256", "providers": [] }, { - "slug": "nousresearch/nous-hermes-llama2-13b", - "hfSlug": "NousResearch/Nous-Hermes-Llama2-13b", + "slug": "phind/phind-codellama-34b", + "hfSlug": "Phind/Phind-CodeLlama-34B-v2", "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2023-08-20T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Nous: Hermes 13B", - "shortName": "Hermes 13B", - "author": "nousresearch", - "description": "A state-of-the-art language model fine-tuned on over 300k instructions by Nous Research, with Teknium and Emozilla leading the fine tuning process.", + "name": "Phind: CodeLlama 34B v2", + "shortName": "CodeLlama 34B v2", + "author": "phind", + "description": "A fine-tune of CodeLlama-34B on an internal dataset that helps it exceed GPT-4 on some benchmarks, including HumanEval.", "modelVersionGroupId": null, "contextLength": 4096, "inputModalities": [ @@ -86266,18 +89565,19 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "nousresearch/nous-hermes-llama2-13b", + "permaslug": "phind/phind-codellama-34b", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 417, "newest": 416, - "throughputHighToLow": 416, + "throughputHighToLow": 417, "latencyLowToHigh": 416, - "pricingLowToHigh": 416, - "pricingHighToLow": 417 + "pricingLowToHigh": 417, + "pricingHighToLow": 416 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { @@ -86321,16 +89621,17 @@ export default { "pricingLowToHigh": 418, "pricingHighToLow": 418 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://huggingface.co/\\u0026size=256", "providers": [] }, { - "slug": "anthropic/claude-instant-1", + "slug": "anthropic/claude-1", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2023-07-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude Instant v1", - "shortName": "Claude Instant v1", + "name": "Anthropic: Claude v1", + "shortName": "Claude v1", "author": "anthropic", "description": "Anthropic's model for low-latency, high throughput text generation. Supports hundreds of pages of text.", "modelVersionGroupId": null, @@ -86343,34 +89644,37 @@ export default { ], "hasTextOutput": true, "group": "Claude", - "instructType": null, + "instructType": "claude", "defaultSystem": null, - "defaultStops": [], + "defaultStops": [ + "\n\nHuman:" + ], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-instant-1", + "permaslug": "anthropic/claude-1", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 419, - "newest": 422, - "throughputHighToLow": 422, - "latencyLowToHigh": 422, - "pricingLowToHigh": 422, - "pricingHighToLow": 419 + "newest": 420, + "throughputHighToLow": 419, + "latencyLowToHigh": 420, + "pricingLowToHigh": 419, + "pricingHighToLow": 420 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [] }, { - "slug": "anthropic/claude-1", + "slug": "anthropic/claude-instant-1.0", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2023-07-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude v1", - "shortName": "Claude v1", + "name": "Anthropic: Claude Instant v1.0", + "shortName": "Claude Instant v1.0", "author": "anthropic", "description": "Anthropic's model for low-latency, high throughput text generation. Supports hundreds of pages of text.", "modelVersionGroupId": null, @@ -86391,18 +89695,19 @@ export default { "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-1", + "permaslug": "anthropic/claude-instant-1.0", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 420, - "newest": 419, - "throughputHighToLow": 419, - "latencyLowToHigh": 419, - "pricingLowToHigh": 419, - "pricingHighToLow": 420 + "newest": 422, + "throughputHighToLow": 420, + "latencyLowToHigh": 422, + "pricingLowToHigh": 420, + "pricingHighToLow": 422 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [] }, { @@ -86445,16 +89750,17 @@ export default { "pricingLowToHigh": 421, "pricingHighToLow": 421 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [] }, { - "slug": "anthropic/claude-instant-1.0", + "slug": "anthropic/claude-instant-1", "hfSlug": null, "updatedAt": "2025-03-28T03:20:30.853469+00:00", "createdAt": "2023-07-28T00:00:00+00:00", "hfUpdatedAt": null, - "name": "Anthropic: Claude Instant v1.0", - "shortName": "Claude Instant v1.0", + "name": "Anthropic: Claude Instant v1", + "shortName": "Claude Instant v1", "author": "anthropic", "description": "Anthropic's model for low-latency, high throughput text generation. Supports hundreds of pages of text.", "modelVersionGroupId": null, @@ -86467,26 +89773,25 @@ export default { ], "hasTextOutput": true, "group": "Claude", - "instructType": "claude", + "instructType": null, "defaultSystem": null, - "defaultStops": [ - "\n\nHuman:" - ], + "defaultStops": [], "hidden": false, "router": null, "warningMessage": null, - "permaslug": "anthropic/claude-instant-1.0", + "permaslug": "anthropic/claude-instant-1", "reasoningConfig": null, "features": {}, "endpoint": null, "sorting": { "topWeekly": 422, - "newest": 420, - "throughputHighToLow": 420, - "latencyLowToHigh": 420, - "pricingLowToHigh": 420, - "pricingHighToLow": 422 + "newest": 419, + "throughputHighToLow": 422, + "latencyLowToHigh": 419, + "pricingLowToHigh": 422, + "pricingHighToLow": 419 }, + "authorIcon": "https://openrouter.ai/images/icons/Anthropic.svg", "providers": [] }, { @@ -86527,6 +89832,7 @@ export default { "pricingLowToHigh": 423, "pricingHighToLow": 423 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [] }, { @@ -86567,6 +89873,7 @@ export default { "pricingLowToHigh": 424, "pricingHighToLow": 424 }, + "authorIcon": "https://openrouter.ai/images/icons/GoogleGemini.svg", "providers": [] }, { @@ -86610,6 +89917,7 @@ export default { "pricingLowToHigh": 425, "pricingHighToLow": 425 }, + "authorIcon": "https://t0.gstatic.com/faviconV2?client=SOCIAL\\u0026type=FAVICON\\u0026fallback_opts=TYPE,SIZE,URL\\u0026url=https://ai.meta.com/\\u0026size=256", "providers": [] }, { @@ -86650,6 +89958,7 @@ export default { "pricingLowToHigh": 426, "pricingHighToLow": 426 }, + "authorIcon": "https://openrouter.ai/images/icons/OpenAI.svg", "providers": [] } ] diff --git a/pkgs/language-models/src/parser.ts b/pkgs/language-models/src/parser.ts index e4d72a9ca..f188ebeeb 100644 --- a/pkgs/language-models/src/parser.ts +++ b/pkgs/language-models/src/parser.ts @@ -389,7 +389,12 @@ export function filterModels( }) .reduce((p: number, n: number) => (p ? p : n), 0) - let sortingStrategy = orderBy(parsed?.priorities?.map((f) => `provider.${f}`) || []) + const reversedFields = [ + 'throughput', + 'cost' + ] + + let sortingStrategy = orderBy(parsed?.priorities?.map((f) => `${reversedFields.includes(f) ? '-' : '' }provider.${f}`) || []) // Re-join back on model, replacing the providers with the filtered providers return { @@ -430,7 +435,7 @@ export function getModel(modelIdentifier: string, augments: Record=18'} - peerDependencies: - zod: ^3.0.0 - '@ai-sdk/anthropic@1.2.11': resolution: {integrity: sha512-lZLcEMh8MXY4NVSrN/7DyI2rnid8k7cn/30nMmd3bwJrnIsOuIuuFvY8f0nj+pFcTi6AYK7ujLdqW5dQVz1YQw==} engines: {node: '>=18'} @@ -1597,24 +1595,12 @@ packages: peerDependencies: zod: ^3.0.0 - '@ai-sdk/google-vertex@2.2.15': - resolution: {integrity: sha512-XTl0dQ1rvLjhrkifSy/483qw3O7vCI6H2b4aAJnzQMfy0vzczMXmvQFS5RA8KmnO+YvsKTuZwBM2xRCNvKw1oQ==} - engines: {node: '>=18'} - peerDependencies: - zod: ^3.0.0 - '@ai-sdk/google-vertex@2.2.22': resolution: {integrity: sha512-qVprZs+DGCVbywLN19AGekgrqu5XEQUm+XPA/lNA4w3qATa81nphj4PiO7ilnNnWJ+YBDcbdZHx+/b1cMOVXIQ==} engines: {node: '>=18'} peerDependencies: zod: ^3.0.0 - '@ai-sdk/google@1.2.11': - resolution: {integrity: sha512-gjGcxKcRri/Jbkujs9nVwP4qOW5GI4rYQ6vQ17uLAvGMo3qnwr26Q2KUqUWuVHQYtboXVSrxC/Kb6sm3hE5WUQ==} - engines: {node: '>=18'} - peerDependencies: - zod: ^3.0.0 - '@ai-sdk/google@1.2.18': resolution: {integrity: sha512-8B70+i+uB12Ae6Sn6B9Oc6W0W/XorGgc88Nx0pyUrcxFOdytHBaAVhTPqYsO3LLClfjYN8pQ9GMxd5cpGEnUcA==} engines: {node: '>=18'} @@ -1699,16 +1685,6 @@ packages: resolution: {integrity: sha512-qZMxYJ0qqX/RfnuIaab+zp8UAeJn/ygXXAffR5I4N0n1IrvA6qBsjc8hXLmBiMV2zoXlifkacF7sEFnYnjBcqg==} engines: {node: '>=18'} - '@ai-sdk/react@1.2.11': - resolution: {integrity: sha512-+kPqLkJ3TWP6czaJPV+vzAKSUcKQ1598BUrcLHt56sH99+LhmIIW3ylZp0OfC3O6TR3eO1Lt0Yzw4R0mK6g9Gw==} - engines: {node: '>=18'} - peerDependencies: - react: ^18 || ^19 || ^19.0.0-rc - zod: ^3.23.8 - peerDependenciesMeta: - zod: - optional: true - '@ai-sdk/react@1.2.12': resolution: {integrity: sha512-jK1IZZ22evPZoQW3vlkZ7wvjYGYF+tRBKXtrcolduIkQ/m/sOAVcVeVDUDvh1T91xCnWCdUGCPZg2avZ90mv3g==} engines: {node: '>=18'} @@ -1719,12 +1695,6 @@ packages: zod: optional: true - '@ai-sdk/ui-utils@1.2.10': - resolution: {integrity: sha512-GUj+LBoAlRQF1dL/M49jtufGqtLOMApxTpCmVjoRpIPt/dFALVL9RfqfvxwztyIwbK+IxGzcYjSGRsrWrj+86g==} - engines: {node: '>=18'} - peerDependencies: - zod: ^3.23.8 - '@ai-sdk/ui-utils@1.2.11': resolution: {integrity: sha512-3zcwCc8ezzFlwp3ZD15wAPjf2Au4s3vAbKsXQVyhxODHcmu0iyPO2Eua6D/vicq/AUm/BAo60r97O6HU+EI0+w==} engines: {node: '>=18'} @@ -5270,19 +5240,6 @@ packages: '@types/react-dom': optional: true - '@radix-ui/react-dialog@1.1.7': - resolution: {integrity: sha512-EIdma8C0C/I6kL6sO02avaCRqi3fmWJpxH6mqbVScorW6nNktzKJT/le7VPho3o/7wCsyRg3z0+Q+Obr0Gy/VQ==} - peerDependencies: - '@types/react': '*' - '@types/react-dom': '*' - react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc - react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc - peerDependenciesMeta: - '@types/react': - optional: true - '@types/react-dom': - optional: true - '@radix-ui/react-direction@1.1.1': resolution: {integrity: sha512-1UEWRX6jnOA2y4H5WczZ44gOOjTEmlqv1uNW4GAJEO5+bauCBhv8snY65Iw5/VOS/ghKN9gr2KjnLKxrsvoMVw==} peerDependencies: @@ -5292,19 +5249,6 @@ packages: '@types/react': optional: true - '@radix-ui/react-dismissable-layer@1.1.6': - resolution: {integrity: sha512-7gpgMT2gyKym9Jz2ZhlRXSg2y6cNQIK8d/cqBZ0RBCaps8pFryCWXiUKI+uHGFrhMrbGUP7U6PWgiXzIxoyF3Q==} - peerDependencies: - '@types/react': '*' - '@types/react-dom': '*' - react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc - react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc - peerDependenciesMeta: - '@types/react': - optional: true - '@types/react-dom': - optional: true - '@radix-ui/react-dismissable-layer@1.1.7': resolution: {integrity: sha512-j5+WBUdhccJsmH5/H0K6RncjDtoALSEr6jbkaZu+bjw6hOPOhHycr6vEUujl+HBK8kjUfWcoCJXxP6e4lUlMZw==} peerDependencies: @@ -5353,19 +5297,6 @@ packages: '@types/react': optional: true - '@radix-ui/react-focus-scope@1.1.3': - resolution: {integrity: sha512-4XaDlq0bPt7oJwR+0k0clCiCO/7lO7NKZTAaJBYxDNQT/vj4ig0/UvctrRscZaFREpRvUTkpKR96ov1e6jptQg==} - peerDependencies: - '@types/react': '*' - '@types/react-dom': '*' - react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc - react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc - peerDependenciesMeta: - '@types/react': - optional: true - '@types/react-dom': - optional: true - '@radix-ui/react-focus-scope@1.1.4': resolution: {integrity: sha512-r2annK27lIW5w9Ho5NyQgqs0MmgZSTIKXWpVCJaLC1q2kZrZkcqnmHkCHMEmv8XLvsLlurKMPT+kbKkRkm/xVA==} peerDependencies: @@ -5466,19 +5397,6 @@ packages: '@types/react-dom': optional: true - '@radix-ui/react-portal@1.1.5': - resolution: {integrity: sha512-ps/67ZqsFm+Mb6lSPJpfhRLrVL2i2fntgCmGMqqth4eaGUf+knAuuRtWVJrNjUhExgmdRqftSgzpf0DF0n6yXA==} - peerDependencies: - '@types/react': '*' - '@types/react-dom': '*' - react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc - react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc - peerDependenciesMeta: - '@types/react': - optional: true - '@types/react-dom': - optional: true - '@radix-ui/react-portal@1.1.6': resolution: {integrity: sha512-XmsIl2z1n/TsYFLIdYam2rmFwf9OC/Sh2avkbmVMDuBZIe7hSpM0cYnWPAo7nHOVx8zTuwDZGByfcqLdnzp3Vw==} peerDependencies: @@ -7477,14 +7395,14 @@ packages: '@tanstack/query-devtools@5.76.0': resolution: {integrity: sha512-1p92nqOBPYVqVDU0Ua5nzHenC6EGZNrLnB2OZphYw8CNA1exuvI97FVgIKON7Uug3uQqvH/QY8suUKpQo8qHNQ==} - '@tanstack/react-query-devtools@5.76.0': - resolution: {integrity: sha512-RoyRzH5XJB//OhAdzQTutesw9uHyNZroLp/I7NDAQf8OVJKTTcoaYBmaw5pmB2e3bVdgqFu6nHFZUr5j5qBdZw==} + '@tanstack/react-query-devtools@5.76.1': + resolution: {integrity: sha512-LFVWgk/VtXPkerNLfYIeuGHh0Aim/k9PFGA+JxLdRaUiroQ4j4eoEqBrUpQ1Pd/KXoG4AB9vVE/M6PUQ9vwxBQ==} peerDependencies: - '@tanstack/react-query': ^5.76.0 + '@tanstack/react-query': ^5.76.1 react: ^18 || ^19 - '@tanstack/react-query@5.76.0': - resolution: {integrity: sha512-dZLYzVuUFZJkenxd8o01oyFimeLBmSkaUviPHuDzXe7LSLO4WTTx92jwJlNUXOOHzg6t0XknklZ15cjhYNSDjA==} + '@tanstack/react-query@5.76.1': + resolution: {integrity: sha512-YxdLZVGN4QkT5YT1HKZQWiIlcgauIXEIsMOTSjvyD5wLYK8YVvKZUPAysMqossFJJfDpJW3pFn7WNZuPOqq+fw==} peerDependencies: react: ^18 || ^19 @@ -8195,6 +8113,11 @@ packages: '@vitest/spy@3.1.1': resolution: {integrity: sha512-+EmrUOOXbKzLkTDwlsc/xrwOlPDXyVk3Z6P6K4oiCndxz7YLpp/0R0UsWVOKT0IXWjjBJuSMk6D27qipaupcvQ==} + '@vitest/ui@3.1.1': + resolution: {integrity: sha512-2HpiRIYg3dlvAJBV9RtsVswFgUSJK4Sv7QhpxoP0eBGkYwzGIKP34PjaV00AULQi9Ovl6LGyZfsetxDWY5BQdQ==} + peerDependencies: + vitest: 3.1.1 + '@vitest/utils@1.6.1': resolution: {integrity: sha512-jOrrUvXM4Av9ZWiG1EajNto0u96kWAhJ1LmPmJhXXQx/32MecEKd10pOLYgS2BQx1TgkGhloPU1ArDW2vvaY6g==} @@ -8442,8 +8365,8 @@ packages: agents@0.0.47: resolution: {integrity: sha512-DiNRi4cps9FEh67j0CntlV4hHy4vkJgRaZtGmAUdf1yslj0TUvx89liS2wAVYKAyy3HgFiMxvELGLV4nWNs+jw==} - agents@0.0.87: - resolution: {integrity: sha512-nanFC1FoByZqAIeCXta7rBR21GmUd3fulkiMLpg0jnDOzs3i0KfH7Rlr41DQ2nP4i7cTYv2VZhjWti0RQdpIPg==} + agents@0.0.88: + resolution: {integrity: sha512-ysQ3LMONALxx1VL0X/Tq49EzfuuLj1SHfztCCIfAX8hVtYw2Jr9rytnM94lZQ29yFFqmcIKWR+D13j/l7Cyf8A==} peerDependencies: react: '*' @@ -8487,16 +8410,6 @@ packages: vue: optional: true - ai@4.3.13: - resolution: {integrity: sha512-cC5HXItuOwGykSMacCPzNp6+NMTxeuTjOenztVgSJhdC9Z4OrzBxwkyeDAf4h1QP938ZFi7IBdq3u4lxVoVmvw==} - engines: {node: '>=18'} - peerDependencies: - react: ^18 || ^19 || ^19.0.0-rc - zod: ^3.23.8 - peerDependenciesMeta: - react: - optional: true - ai@4.3.15: resolution: {integrity: sha512-TYKRzbWg6mx/pmTadlAEIhuQtzfHUV0BbLY72+zkovXwq/9xhcH24IlQmkyBpElK6/4ArS0dHdOOtR1jOPVwtg==} engines: {node: '>=18'} @@ -15984,6 +15897,10 @@ packages: resolution: {integrity: sha512-94Bdh3cC2PKrbgSOUqTiGPWVZeSiXfKOVZNJniWoqrWrRkB1CJzBU3NEbiTsPcYy1lDsANA/THzS+9WBiy5nfQ==} engines: {node: '>= 10'} + sirv@3.0.1: + resolution: {integrity: sha512-FoqMu0NCGBLCcAkS1qA+XJIQTR6/JHfQXl+uGteNCQ76T91DMUjPa9xfmeqMY3z80nLSg9yQmNjK0Px6RWsH/A==} + engines: {node: '>=18'} + sisteransi@1.0.5: resolution: {integrity: sha512-bLGGlR1QxBcynn2d5YmDX4MGjlZvy2MRBDRNHLJ8VI6l6+9FUiyTFNJ0IveOSP0bcXgVDPRcfGqA0pjaqUpfVg==} @@ -17953,51 +17870,39 @@ snapshots: '@adobe/css-tools@4.4.2': {} - '@ai-sdk/amazon-bedrock@1.1.6(zod@3.24.3)': + '@ai-sdk/amazon-bedrock@1.1.6(zod@3.24.4)': dependencies: '@ai-sdk/provider': 1.0.7 - '@ai-sdk/provider-utils': 2.1.6(zod@3.24.3) + '@ai-sdk/provider-utils': 2.1.6(zod@3.24.4) '@aws-sdk/client-bedrock-runtime': 3.810.0 - zod: 3.24.3 + zod: 3.24.4 transitivePeerDependencies: - aws-crt - '@ai-sdk/anthropic@1.2.10(zod@3.24.3)': - dependencies: - '@ai-sdk/provider': 1.1.3 - '@ai-sdk/provider-utils': 2.2.7(zod@3.24.3) - zod: 3.24.3 - '@ai-sdk/anthropic@1.2.11(zod@3.24.3)': dependencies: '@ai-sdk/provider': 1.1.3 '@ai-sdk/provider-utils': 2.2.8(zod@3.24.3) zod: 3.24.3 - '@ai-sdk/deepseek@0.2.13(zod@3.24.3)': + '@ai-sdk/anthropic@1.2.11(zod@3.24.4)': dependencies: - '@ai-sdk/openai-compatible': 0.2.13(zod@3.24.3) '@ai-sdk/provider': 1.1.3 - '@ai-sdk/provider-utils': 2.2.7(zod@3.24.3) - zod: 3.24.3 + '@ai-sdk/provider-utils': 2.2.8(zod@3.24.4) + zod: 3.24.4 - '@ai-sdk/elevenlabs@0.0.2(zod@3.24.3)': + '@ai-sdk/deepseek@0.2.13(zod@3.24.4)': dependencies: + '@ai-sdk/openai-compatible': 0.2.13(zod@3.24.4) '@ai-sdk/provider': 1.1.3 - '@ai-sdk/provider-utils': 2.2.7(zod@3.24.3) - zod: 3.24.3 + '@ai-sdk/provider-utils': 2.2.7(zod@3.24.4) + zod: 3.24.4 - '@ai-sdk/google-vertex@2.2.15(encoding@0.1.13)(zod@3.24.3)': + '@ai-sdk/elevenlabs@0.0.2(zod@3.24.4)': dependencies: - '@ai-sdk/anthropic': 1.2.10(zod@3.24.3) - '@ai-sdk/google': 1.2.11(zod@3.24.3) '@ai-sdk/provider': 1.1.3 - '@ai-sdk/provider-utils': 2.2.7(zod@3.24.3) - google-auth-library: 9.15.1(encoding@0.1.13) - zod: 3.24.3 - transitivePeerDependencies: - - encoding - - supports-color + '@ai-sdk/provider-utils': 2.2.7(zod@3.24.4) + zod: 3.24.4 '@ai-sdk/google-vertex@2.2.22(encoding@0.1.13)(zod@3.24.3)': dependencies: @@ -18011,11 +17916,17 @@ snapshots: - encoding - supports-color - '@ai-sdk/google@1.2.11(zod@3.24.3)': + '@ai-sdk/google-vertex@2.2.22(encoding@0.1.13)(zod@3.24.4)': dependencies: + '@ai-sdk/anthropic': 1.2.11(zod@3.24.4) + '@ai-sdk/google': 1.2.18(zod@3.24.4) '@ai-sdk/provider': 1.1.3 - '@ai-sdk/provider-utils': 2.2.7(zod@3.24.3) - zod: 3.24.3 + '@ai-sdk/provider-utils': 2.2.8(zod@3.24.4) + google-auth-library: 9.15.1(encoding@0.1.13) + zod: 3.24.4 + transitivePeerDependencies: + - encoding + - supports-color '@ai-sdk/google@1.2.18(zod@3.24.3)': dependencies: @@ -18023,11 +17934,17 @@ snapshots: '@ai-sdk/provider-utils': 2.2.8(zod@3.24.3) zod: 3.24.3 - '@ai-sdk/groq@1.2.9(zod@3.24.3)': + '@ai-sdk/google@1.2.18(zod@3.24.4)': dependencies: '@ai-sdk/provider': 1.1.3 - '@ai-sdk/provider-utils': 2.2.8(zod@3.24.3) - zod: 3.24.3 + '@ai-sdk/provider-utils': 2.2.8(zod@3.24.4) + zod: 3.24.4 + + '@ai-sdk/groq@1.2.9(zod@3.24.4)': + dependencies: + '@ai-sdk/provider': 1.1.3 + '@ai-sdk/provider-utils': 2.2.8(zod@3.24.4) + zod: 3.24.4 '@ai-sdk/openai-compatible@0.0.9(zod@3.24.4)': dependencies: @@ -18035,17 +17952,17 @@ snapshots: '@ai-sdk/provider-utils': 2.0.4(zod@3.24.4) zod: 3.24.4 - '@ai-sdk/openai-compatible@0.2.13(zod@3.24.3)': + '@ai-sdk/openai-compatible@0.2.13(zod@3.24.4)': dependencies: '@ai-sdk/provider': 1.1.3 - '@ai-sdk/provider-utils': 2.2.7(zod@3.24.3) - zod: 3.24.3 + '@ai-sdk/provider-utils': 2.2.7(zod@3.24.4) + zod: 3.24.4 - '@ai-sdk/openai-compatible@0.2.14(zod@3.24.3)': + '@ai-sdk/openai-compatible@0.2.14(zod@3.24.4)': dependencies: '@ai-sdk/provider': 1.1.3 - '@ai-sdk/provider-utils': 2.2.8(zod@3.24.3) - zod: 3.24.3 + '@ai-sdk/provider-utils': 2.2.8(zod@3.24.4) + zod: 3.24.4 '@ai-sdk/openai@1.3.16(zod@3.24.3)': dependencies: @@ -18059,11 +17976,11 @@ snapshots: '@ai-sdk/provider-utils': 2.2.7(zod@3.24.4) zod: 3.24.4 - '@ai-sdk/perplexity@1.1.9(zod@3.24.3)': + '@ai-sdk/perplexity@1.1.9(zod@3.24.4)': dependencies: '@ai-sdk/provider': 1.1.3 - '@ai-sdk/provider-utils': 2.2.8(zod@3.24.3) - zod: 3.24.3 + '@ai-sdk/provider-utils': 2.2.8(zod@3.24.4) + zod: 3.24.4 '@ai-sdk/provider-utils@2.0.4(zod@3.24.4)': dependencies: @@ -18074,14 +17991,14 @@ snapshots: optionalDependencies: zod: 3.24.4 - '@ai-sdk/provider-utils@2.1.6(zod@3.24.3)': + '@ai-sdk/provider-utils@2.1.6(zod@3.24.4)': dependencies: '@ai-sdk/provider': 1.0.7 eventsource-parser: 3.0.1 nanoid: 3.3.11 secure-json-parse: 2.7.0 optionalDependencies: - zod: 3.24.3 + zod: 3.24.4 '@ai-sdk/provider-utils@2.2.7(zod@3.24.3)': dependencies: @@ -18123,25 +18040,15 @@ snapshots: dependencies: json-schema: 0.4.0 - '@ai-sdk/react@1.2.11(react@18.3.1)(zod@3.24.3)': + '@ai-sdk/react@1.2.12(react@18.3.1)(zod@3.24.4)': dependencies: - '@ai-sdk/provider-utils': 2.2.7(zod@3.24.3) - '@ai-sdk/ui-utils': 1.2.10(zod@3.24.3) + '@ai-sdk/provider-utils': 2.2.8(zod@3.24.4) + '@ai-sdk/ui-utils': 1.2.11(zod@3.24.4) react: 18.3.1 swr: 2.3.3(react@18.3.1) throttleit: 2.1.0 optionalDependencies: - zod: 3.24.3 - - '@ai-sdk/react@1.2.11(react@19.1.0)(zod@3.24.3)': - dependencies: - '@ai-sdk/provider-utils': 2.2.7(zod@3.24.3) - '@ai-sdk/ui-utils': 1.2.10(zod@3.24.3) - react: 19.1.0 - swr: 2.3.3(react@19.1.0) - throttleit: 2.1.0 - optionalDependencies: - zod: 3.24.3 + zod: 3.24.4 '@ai-sdk/react@1.2.12(react@19.1.0)(zod@3.24.3)': dependencies: @@ -18163,13 +18070,6 @@ snapshots: optionalDependencies: zod: 3.24.4 - '@ai-sdk/ui-utils@1.2.10(zod@3.24.3)': - dependencies: - '@ai-sdk/provider': 1.1.3 - '@ai-sdk/provider-utils': 2.2.7(zod@3.24.3) - zod: 3.24.3 - zod-to-json-schema: 3.24.5(zod@3.24.3) - '@ai-sdk/ui-utils@1.2.11(zod@3.24.3)': dependencies: '@ai-sdk/provider': 1.1.3 @@ -18184,12 +18084,12 @@ snapshots: zod: 3.24.4 zod-to-json-schema: 3.24.5(zod@3.24.4) - '@ai-sdk/xai@1.2.16(zod@3.24.3)': + '@ai-sdk/xai@1.2.16(zod@3.24.4)': dependencies: - '@ai-sdk/openai-compatible': 0.2.14(zod@3.24.3) + '@ai-sdk/openai-compatible': 0.2.14(zod@3.24.4) '@ai-sdk/provider': 1.1.3 - '@ai-sdk/provider-utils': 2.2.8(zod@3.24.3) - zod: 3.24.3 + '@ai-sdk/provider-utils': 2.2.8(zod@3.24.4) + zod: 3.24.4 '@alloc/quick-lru@5.2.0': {} @@ -18255,15 +18155,15 @@ snapshots: '@csstools/css-tokenizer': 3.0.3 lru-cache: 10.4.3 - '@asteasolutions/zod-to-openapi@6.4.0(zod@3.24.3)': + '@asteasolutions/zod-to-openapi@6.4.0(zod@3.24.4)': dependencies: openapi3-ts: 4.4.0 - zod: 3.24.3 + zod: 3.24.4 - '@asteasolutions/zod-to-openapi@7.3.0(zod@3.24.3)': + '@asteasolutions/zod-to-openapi@7.3.0(zod@3.24.4)': dependencies: openapi3-ts: 4.4.0 - zod: 3.24.3 + zod: 3.24.4 '@auth/core@0.37.2(nodemailer@6.9.16)': dependencies: @@ -19327,15 +19227,15 @@ snapshots: '@braintrust/core@0.0.84': dependencies: - '@asteasolutions/zod-to-openapi': 6.4.0(zod@3.24.3) + '@asteasolutions/zod-to-openapi': 6.4.0(zod@3.24.4) uuid: 9.0.1 - zod: 3.24.3 + zod: 3.24.4 '@browserbasehq/sdk@1.5.0': dependencies: playwright-core: 1.51.1 puppeteer-core: 22.15.0 - zod: 3.24.3 + zod: 3.24.4 transitivePeerDependencies: - bare-buffer - bufferutil @@ -19354,17 +19254,17 @@ snapshots: transitivePeerDependencies: - encoding - '@browserbasehq/stagehand@1.14.0(@playwright/test@1.51.1)(deepmerge@4.3.1)(dotenv@16.5.0)(encoding@0.1.13)(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3))(zod@3.24.3)': + '@browserbasehq/stagehand@1.14.0(@playwright/test@1.51.1)(deepmerge@4.3.1)(dotenv@16.5.0)(encoding@0.1.13)(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4))(zod@3.24.4)': dependencies: '@anthropic-ai/sdk': 0.27.3(encoding@0.1.13) '@browserbasehq/sdk': 2.5.0(encoding@0.1.13) '@playwright/test': 1.51.1 deepmerge: 4.3.1 dotenv: 16.5.0 - openai: 4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3) + openai: 4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4) ws: 8.18.1 - zod: 3.24.3 - zod-to-json-schema: 3.24.5(zod@3.24.3) + zod: 3.24.4 + zod-to-json-schema: 3.24.5(zod@3.24.4) transitivePeerDependencies: - bufferutil - encoding @@ -19434,7 +19334,7 @@ snapshots: optionalDependencies: workerd: 1.20250424.0 - '@cloudflare/vitest-pool-workers@0.8.22(@cloudflare/workers-types@4.20250415.0)(@vitest/runner@3.1.1)(@vitest/snapshot@3.1.1)(vitest@1.6.1(@types/node@22.14.0)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0))': + '@cloudflare/vitest-pool-workers@0.8.22(@cloudflare/workers-types@4.20250415.0)(@vitest/runner@3.1.1)(@vitest/snapshot@3.1.1)(vitest@1.6.1(@types/node@22.14.0)(@vitest/ui@3.1.1)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0))': dependencies: '@vitest/runner': 3.1.1 '@vitest/snapshot': 3.1.1 @@ -19443,9 +19343,9 @@ snapshots: devalue: 4.3.3 miniflare: 4.20250424.1 semver: 7.7.1 - vitest: 1.6.1(@types/node@22.14.0)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0) + vitest: 1.6.1(@types/node@22.14.0)(@vitest/ui@3.1.1)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0) wrangler: 4.13.2(@cloudflare/workers-types@4.20250415.0) - zod: 3.24.3 + zod: 3.24.4 transitivePeerDependencies: - '@cloudflare/workers-types' - bufferutil @@ -20644,10 +20544,10 @@ snapshots: dependencies: axios: 1.8.4 - '@hono/zod-validator@0.1.11(hono@4.7.6)(zod@3.24.3)': + '@hono/zod-validator@0.1.11(hono@4.7.6)(zod@3.24.4)': dependencies: hono: 4.7.6 - zod: 3.24.3 + zod: 3.24.4 '@hookform/resolvers@5.0.1(react-hook-form@7.56.1(react@19.1.0))': dependencies: @@ -21150,6 +21050,23 @@ snapshots: transitivePeerDependencies: - openai + '@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4))': + dependencies: + '@cfworker/json-schema': 4.1.1 + ansi-styles: 5.2.0 + camelcase: 6.3.0 + decamelize: 1.2.0 + js-tiktoken: 1.0.20 + langsmith: 0.3.21(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)) + mustache: 4.2.0 + p-queue: 6.6.2 + p-retry: 4.6.2 + uuid: 10.0.0 + zod: 3.24.4 + zod-to-json-schema: 3.24.5(zod@3.24.4) + transitivePeerDependencies: + - openai + '@langchain/openai@0.5.5(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)))(encoding@0.1.13)(ws@8.18.1)': dependencies: '@langchain/core': 0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)) @@ -21161,11 +21078,27 @@ snapshots: - encoding - ws + '@langchain/openai@0.5.5(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)))(encoding@0.1.13)(ws@8.18.1)': + dependencies: + '@langchain/core': 0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)) + js-tiktoken: 1.0.20 + openai: 4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4) + zod: 3.24.4 + zod-to-json-schema: 3.24.5(zod@3.24.4) + transitivePeerDependencies: + - encoding + - ws + '@langchain/textsplitters@0.1.0(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)))': dependencies: '@langchain/core': 0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)) js-tiktoken: 1.0.20 + '@langchain/textsplitters@0.1.0(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)))': + dependencies: + '@langchain/core': 0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)) + js-tiktoken: 1.0.20 + '@lexical/clipboard@0.28.0': dependencies: '@lexical/html': 0.28.0 @@ -21629,11 +21562,11 @@ snapshots: '@tybys/wasm-util': 0.9.0 optional: true - '@next-safe-action/adapter-react-hook-form@1.0.14(@hookform/resolvers@5.0.1(react-hook-form@7.56.1(react@19.1.0)))(next-safe-action@7.10.8(next@15.3.1(@babel/core@7.26.10)(@opentelemetry/api@1.9.0)(@playwright/test@1.51.1)(babel-plugin-macros@3.1.0)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)(sass@1.77.4))(react-dom@19.1.0(react@19.1.0))(react@19.1.0)(zod@3.24.3))(next@15.3.1(@babel/core@7.26.10)(@opentelemetry/api@1.9.0)(@playwright/test@1.51.1)(babel-plugin-macros@3.1.0)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)(sass@1.77.4))(react-dom@19.1.0(react@19.1.0))(react-hook-form@7.56.1(react@19.1.0))(react@19.1.0)': + '@next-safe-action/adapter-react-hook-form@1.0.14(@hookform/resolvers@5.0.1(react-hook-form@7.56.1(react@19.1.0)))(next-safe-action@7.10.8(next@15.3.1(@babel/core@7.26.10)(@opentelemetry/api@1.9.0)(@playwright/test@1.51.1)(babel-plugin-macros@3.1.0)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)(sass@1.77.4))(react-dom@19.1.0(react@19.1.0))(react@19.1.0)(zod@3.24.4))(next@15.3.1(@babel/core@7.26.10)(@opentelemetry/api@1.9.0)(@playwright/test@1.51.1)(babel-plugin-macros@3.1.0)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)(sass@1.77.4))(react-dom@19.1.0(react@19.1.0))(react-hook-form@7.56.1(react@19.1.0))(react@19.1.0)': dependencies: '@hookform/resolvers': 5.0.1(react-hook-form@7.56.1(react@19.1.0)) next: 15.3.1(@babel/core@7.26.10)(@opentelemetry/api@1.9.0)(@playwright/test@1.51.1)(babel-plugin-macros@3.1.0)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)(sass@1.77.4) - next-safe-action: 7.10.8(next@15.3.1(@babel/core@7.26.10)(@opentelemetry/api@1.9.0)(@playwright/test@1.51.1)(babel-plugin-macros@3.1.0)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)(sass@1.77.4))(react-dom@19.1.0(react@19.1.0))(react@19.1.0)(zod@3.24.3) + next-safe-action: 7.10.8(next@15.3.1(@babel/core@7.26.10)(@opentelemetry/api@1.9.0)(@playwright/test@1.51.1)(babel-plugin-macros@3.1.0)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)(sass@1.77.4))(react-dom@19.1.0(react@19.1.0))(react@19.1.0)(zod@3.24.4) react: 19.1.0 react-dom: 19.1.0(react@19.1.0) react-hook-form: 7.56.1(react@19.1.0) @@ -22592,47 +22525,12 @@ snapshots: '@types/react': 18.3.20 '@types/react-dom': 19.0.4(@types/react@18.3.20) - '@radix-ui/react-dialog@1.1.7(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)': - dependencies: - '@radix-ui/primitive': 1.1.2 - '@radix-ui/react-compose-refs': 1.1.2(@types/react@18.3.20)(react@19.1.0) - '@radix-ui/react-context': 1.1.2(@types/react@18.3.20)(react@19.1.0) - '@radix-ui/react-dismissable-layer': 1.1.6(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0) - '@radix-ui/react-focus-guards': 1.1.2(@types/react@18.3.20)(react@19.1.0) - '@radix-ui/react-focus-scope': 1.1.3(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0) - '@radix-ui/react-id': 1.1.1(@types/react@18.3.20)(react@19.1.0) - '@radix-ui/react-portal': 1.1.5(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0) - '@radix-ui/react-presence': 1.1.3(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0) - '@radix-ui/react-primitive': 2.0.3(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0) - '@radix-ui/react-slot': 1.2.0(@types/react@18.3.20)(react@19.1.0) - '@radix-ui/react-use-controllable-state': 1.1.1(@types/react@18.3.20)(react@19.1.0) - aria-hidden: 1.2.4 - react: 19.1.0 - react-dom: 19.1.0(react@19.1.0) - react-remove-scroll: 2.6.3(@types/react@18.3.20)(react@19.1.0) - optionalDependencies: - '@types/react': 18.3.20 - '@types/react-dom': 19.0.4(@types/react@18.3.20) - '@radix-ui/react-direction@1.1.1(@types/react@18.3.20)(react@19.1.0)': dependencies: react: 19.1.0 optionalDependencies: '@types/react': 18.3.20 - '@radix-ui/react-dismissable-layer@1.1.6(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)': - dependencies: - '@radix-ui/primitive': 1.1.2 - '@radix-ui/react-compose-refs': 1.1.2(@types/react@18.3.20)(react@19.1.0) - '@radix-ui/react-primitive': 2.0.3(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0) - '@radix-ui/react-use-callback-ref': 1.1.1(@types/react@18.3.20)(react@19.1.0) - '@radix-ui/react-use-escape-keydown': 1.1.1(@types/react@18.3.20)(react@19.1.0) - react: 19.1.0 - react-dom: 19.1.0(react@19.1.0) - optionalDependencies: - '@types/react': 18.3.20 - '@types/react-dom': 19.0.4(@types/react@18.3.20) - '@radix-ui/react-dismissable-layer@1.1.7(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)': dependencies: '@radix-ui/primitive': 1.1.2 @@ -22680,17 +22578,6 @@ snapshots: optionalDependencies: '@types/react': 18.3.20 - '@radix-ui/react-focus-scope@1.1.3(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)': - dependencies: - '@radix-ui/react-compose-refs': 1.1.2(@types/react@18.3.20)(react@19.1.0) - '@radix-ui/react-primitive': 2.0.3(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0) - '@radix-ui/react-use-callback-ref': 1.1.1(@types/react@18.3.20)(react@19.1.0) - react: 19.1.0 - react-dom: 19.1.0(react@19.1.0) - optionalDependencies: - '@types/react': 18.3.20 - '@types/react-dom': 19.0.4(@types/react@18.3.20) - '@radix-ui/react-focus-scope@1.1.4(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)': dependencies: '@radix-ui/react-compose-refs': 1.1.2(@types/react@18.3.20)(react@19.1.0) @@ -22814,16 +22701,6 @@ snapshots: '@types/react': 18.3.20 '@types/react-dom': 19.0.4(@types/react@18.3.20) - '@radix-ui/react-portal@1.1.5(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)': - dependencies: - '@radix-ui/react-primitive': 2.0.3(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0) - '@radix-ui/react-use-layout-effect': 1.1.1(@types/react@18.3.20)(react@19.1.0) - react: 19.1.0 - react-dom: 19.1.0(react@19.1.0) - optionalDependencies: - '@types/react': 18.3.20 - '@types/react-dom': 19.0.4(@types/react@18.3.20) - '@radix-ui/react-portal@1.1.6(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)': dependencies: '@radix-ui/react-primitive': 2.1.0(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0) @@ -23684,7 +23561,7 @@ snapshots: github-slugger: 2.0.0 nanoid: 5.1.5 vue: 3.5.13(typescript@5.7.3) - zod: 3.24.3 + zod: 3.24.4 transitivePeerDependencies: - '@hyperjump/browser' - '@vue/composition-api' @@ -23836,7 +23713,7 @@ snapshots: '@unhead/schema': 1.11.20 nanoid: 5.1.5 type-fest: 4.40.1 - zod: 3.24.3 + zod: 3.24.4 '@scalar/types@0.1.7': dependencies: @@ -23844,7 +23721,7 @@ snapshots: '@unhead/schema': 1.11.20 nanoid: 5.1.5 type-fest: 4.40.0 - zod: 3.24.3 + zod: 3.24.4 '@scalar/use-codemirror@0.11.92(typescript@5.7.3)': dependencies: @@ -25598,13 +25475,13 @@ snapshots: '@tanstack/query-devtools@5.76.0': {} - '@tanstack/react-query-devtools@5.76.0(@tanstack/react-query@5.76.0(react@19.1.0))(react@19.1.0)': + '@tanstack/react-query-devtools@5.76.1(@tanstack/react-query@5.76.1(react@19.1.0))(react@19.1.0)': dependencies: '@tanstack/query-devtools': 5.76.0 - '@tanstack/react-query': 5.76.0(react@19.1.0) + '@tanstack/react-query': 5.76.1(react@19.1.0) react: 19.1.0 - '@tanstack/react-query@5.76.0(react@19.1.0)': + '@tanstack/react-query@5.76.1(react@19.1.0)': dependencies: '@tanstack/query-core': 5.76.0 react: 19.1.0 @@ -26389,7 +26266,7 @@ snapshots: transitivePeerDependencies: - supports-color - '@vitest/coverage-istanbul@3.1.3(vitest@3.1.1(@types/debug@4.1.12)(@types/node@22.14.0)(jiti@2.4.2)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0)(tsx@4.19.3)(yaml@2.7.1))': + '@vitest/coverage-istanbul@3.1.3(vitest@3.1.1)': dependencies: '@istanbuljs/schema': 0.1.3 debug: 4.4.0(supports-color@8.1.1) @@ -26401,7 +26278,7 @@ snapshots: magicast: 0.3.5 test-exclude: 7.0.1 tinyrainbow: 2.0.0 - vitest: 3.1.1(@types/debug@4.1.12)(@types/node@22.14.0)(jiti@2.4.2)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0)(tsx@4.19.3)(yaml@2.7.1) + vitest: 3.1.1(@types/debug@4.1.12)(@types/node@22.14.0)(@vitest/ui@3.1.1)(jiti@2.4.2)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0)(tsx@4.19.3)(yaml@2.7.1) transitivePeerDependencies: - supports-color @@ -26470,6 +26347,17 @@ snapshots: dependencies: tinyspy: 3.0.2 + '@vitest/ui@3.1.1(vitest@3.1.1)': + dependencies: + '@vitest/utils': 3.1.1 + fflate: 0.8.2 + flatted: 3.3.3 + pathe: 2.0.3 + sirv: 3.0.1 + tinyglobby: 0.2.13 + tinyrainbow: 2.0.0 + vitest: 3.1.1(@types/debug@4.1.12)(@types/node@22.14.0)(@vitest/ui@3.1.1)(jiti@2.4.2)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0)(tsx@4.19.3)(yaml@2.7.1) + '@vitest/utils@1.6.1': dependencies: diff-sequences: 29.6.3 @@ -26803,7 +26691,7 @@ snapshots: transitivePeerDependencies: - '@cloudflare/workers-types' - agents@0.0.87(@cloudflare/workers-types@4.20250415.0)(react@19.1.0): + agents@0.0.88(@cloudflare/workers-types@4.20250415.0)(react@19.1.0): dependencies: '@modelcontextprotocol/sdk': 1.11.2 ai: 4.3.15(react@19.1.0)(zod@3.24.4) @@ -26827,14 +26715,14 @@ snapshots: clean-stack: 5.2.0 indent-string: 5.0.0 - ai-functions@0.2.19(@aws-sdk/credential-providers@3.787.0)(encoding@0.1.13)(socks@2.8.4)(ws@8.18.1)(zod@3.24.3): + ai-functions@0.2.19(@aws-sdk/credential-providers@3.787.0)(encoding@0.1.13)(socks@2.8.4)(ws@8.18.1)(zod@3.24.4): dependencies: js-yaml: 4.1.0 json-schema-to-ts: 3.1.1 kafkajs: 2.2.4 lodash-es: 4.17.21 mongodb: 6.15.0(@aws-sdk/credential-providers@3.787.0)(socks@2.8.4) - openai: 4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3) + openai: 4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4) p-queue: 7.4.1 partial-json-parser: 1.2.2 xstate: 5.19.2 @@ -26860,18 +26748,18 @@ snapshots: transitivePeerDependencies: - react - ai-providers@0.1.0(encoding@0.1.13)(react@19.1.0)(zod@3.24.3): + ai-providers@0.1.0(encoding@0.1.13)(react@19.1.0)(zod@3.24.4): dependencies: - '@ai-sdk/amazon-bedrock': 1.1.6(zod@3.24.3) - '@ai-sdk/anthropic': 1.2.10(zod@3.24.3) - '@ai-sdk/elevenlabs': 0.0.2(zod@3.24.3) - '@ai-sdk/google': 1.2.11(zod@3.24.3) - '@ai-sdk/google-vertex': 2.2.22(encoding@0.1.13)(zod@3.24.3) - '@ai-sdk/groq': 1.2.9(zod@3.24.3) - '@ai-sdk/openai': 1.3.16(zod@3.24.3) - '@ai-sdk/perplexity': 1.1.9(zod@3.24.3) - '@ai-sdk/xai': 1.2.16(zod@3.24.3) - ai: 4.3.15(react@19.1.0)(zod@3.24.3) + '@ai-sdk/amazon-bedrock': 1.1.6(zod@3.24.4) + '@ai-sdk/anthropic': 1.2.11(zod@3.24.4) + '@ai-sdk/elevenlabs': 0.0.2(zod@3.24.4) + '@ai-sdk/google': 1.2.18(zod@3.24.4) + '@ai-sdk/google-vertex': 2.2.22(encoding@0.1.13)(zod@3.24.4) + '@ai-sdk/groq': 1.2.9(zod@3.24.4) + '@ai-sdk/openai': 1.3.16(zod@3.24.4) + '@ai-sdk/perplexity': 1.1.9(zod@3.24.4) + '@ai-sdk/xai': 1.2.16(zod@3.24.4) + ai: 4.3.15(react@19.1.0)(zod@3.24.4) transitivePeerDependencies: - aws-crt - encoding @@ -26881,16 +26769,16 @@ snapshots: ai-workflows@1.1.0(@aws-sdk/credential-providers@3.787.0)(@mdx-js/mdx@3.1.0(acorn@8.14.1))(encoding@0.1.13)(socks@2.8.4)(web-streams-polyfill@4.1.0)(ws@8.18.1): dependencies: - '@ai-sdk/openai': 1.3.16(zod@3.24.3) - ai: 4.3.13(react@18.3.1)(zod@3.24.3) - ai-functions: 0.2.19(@aws-sdk/credential-providers@3.787.0)(encoding@0.1.13)(socks@2.8.4)(ws@8.18.1)(zod@3.24.3) + '@ai-sdk/openai': 1.3.16(zod@3.24.4) + ai: 4.3.15(react@18.3.1)(zod@3.24.4) + ai-functions: 0.2.19(@aws-sdk/credential-providers@3.787.0)(encoding@0.1.13)(socks@2.8.4)(ws@8.18.1)(zod@3.24.4) gray-matter: 4.0.3 mdxld: 0.1.3(@mdx-js/mdx@3.1.0(acorn@8.14.1))(web-streams-polyfill@4.1.0) - openai: 4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3) + openai: 4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4) react: 18.3.1 react-dom: 18.3.1(react@18.3.1) yaml: 2.7.1 - zod: 3.24.3 + zod: 3.24.4 transitivePeerDependencies: - '@aws-sdk/credential-providers' - '@mdx-js/mdx' @@ -26935,30 +26823,18 @@ snapshots: svelte: 4.2.19 vue: 3.5.13(typescript@5.8.3) - ai@4.3.13(react@18.3.1)(zod@3.24.3): + ai@4.3.15(react@18.3.1)(zod@3.24.4): dependencies: '@ai-sdk/provider': 1.1.3 - '@ai-sdk/provider-utils': 2.2.7(zod@3.24.3) - '@ai-sdk/react': 1.2.11(react@18.3.1)(zod@3.24.3) - '@ai-sdk/ui-utils': 1.2.10(zod@3.24.3) + '@ai-sdk/provider-utils': 2.2.8(zod@3.24.4) + '@ai-sdk/react': 1.2.12(react@18.3.1)(zod@3.24.4) + '@ai-sdk/ui-utils': 1.2.11(zod@3.24.4) '@opentelemetry/api': 1.9.0 jsondiffpatch: 0.6.0 - zod: 3.24.3 + zod: 3.24.4 optionalDependencies: react: 18.3.1 - ai@4.3.13(react@19.1.0)(zod@3.24.3): - dependencies: - '@ai-sdk/provider': 1.1.3 - '@ai-sdk/provider-utils': 2.2.7(zod@3.24.3) - '@ai-sdk/react': 1.2.11(react@19.1.0)(zod@3.24.3) - '@ai-sdk/ui-utils': 1.2.10(zod@3.24.3) - '@opentelemetry/api': 1.9.0 - jsondiffpatch: 0.6.0 - zod: 3.24.3 - optionalDependencies: - react: 19.1.0 - ai@4.3.15(react@19.1.0)(zod@3.24.3): dependencies: '@ai-sdk/provider': 1.1.3 @@ -27232,8 +27108,8 @@ snapshots: linear-sum-assignment: 1.0.7 mustache: 4.2.0 openai: 4.47.1(encoding@0.1.13) - zod: 3.24.3 - zod-to-json-schema: 3.24.5(zod@3.24.3) + zod: 3.24.4 + zod-to-json-schema: 3.24.5(zod@3.24.4) transitivePeerDependencies: - encoding @@ -27637,10 +27513,10 @@ snapshots: chanfana@2.8.0: dependencies: - '@asteasolutions/zod-to-openapi': 7.3.0(zod@3.24.3) + '@asteasolutions/zod-to-openapi': 7.3.0(zod@3.24.4) js-yaml: 4.1.0 openapi3-ts: 4.4.0 - zod: 3.24.3 + zod: 3.24.4 char-regex@1.0.2: {} @@ -27896,7 +27772,7 @@ snapshots: cmdk@1.1.1(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0): dependencies: '@radix-ui/react-compose-refs': 1.1.2(@types/react@18.3.20)(react@19.1.0) - '@radix-ui/react-dialog': 1.1.7(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0) + '@radix-ui/react-dialog': 1.1.11(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0) '@radix-ui/react-id': 1.1.1(@types/react@18.3.20)(react@19.1.0) '@radix-ui/react-primitive': 2.0.3(@types/react-dom@19.0.4(@types/react@18.3.20))(@types/react@18.3.20)(react-dom@19.1.0(react@19.1.0))(react@19.1.0) react: 19.1.0 @@ -27994,7 +27870,7 @@ snapshots: array-ify: 1.0.0 dot-prop: 5.3.0 - composio-core@0.5.31(@ai-sdk/openai@1.3.16(zod@3.24.3))(@cloudflare/workers-types@4.20250415.0)(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)))(@langchain/openai@0.5.5(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)))(encoding@0.1.13)(ws@8.18.1))(ai@4.3.13(react@19.1.0)(zod@3.24.3))(langchain@0.3.21(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)))(axios@1.8.4)(cheerio@1.0.0)(encoding@0.1.13)(handlebars@4.7.8)(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3))(ws@8.18.1))(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)): + composio-core@0.5.31(@ai-sdk/openai@1.3.16(zod@3.24.3))(@cloudflare/workers-types@4.20250415.0)(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)))(@langchain/openai@0.5.5(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)))(encoding@0.1.13)(ws@8.18.1))(ai@4.3.15(react@19.1.0)(zod@3.24.3))(langchain@0.3.21(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)))(axios@1.8.4)(cheerio@1.0.0)(encoding@0.1.13)(handlebars@4.7.8)(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3))(ws@8.18.1))(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)): dependencies: '@ai-sdk/openai': 1.3.16(zod@3.24.3) '@cloudflare/workers-types': 4.20250415.0 @@ -28002,7 +27878,7 @@ snapshots: '@hey-api/client-axios': 0.2.12(axios@1.8.4) '@langchain/core': 0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)) '@langchain/openai': 0.5.5(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)))(encoding@0.1.13)(ws@8.18.1) - ai: 4.3.13(react@19.1.0)(zod@3.24.3) + ai: 4.3.15(react@19.1.0)(zod@3.24.3) axios: 1.8.4 chalk: 4.1.2 cli-progress: 3.12.0 @@ -28019,23 +27895,23 @@ snapshots: transitivePeerDependencies: - debug - composio-core@0.5.31(@ai-sdk/openai@1.3.16(zod@3.24.3))(@cloudflare/workers-types@4.20250415.0)(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)))(@langchain/openai@0.5.5(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)))(encoding@0.1.13)(ws@8.18.1))(ai@4.3.13(react@19.1.0)(zod@3.24.3))(langchain@0.3.21(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)))(axios@1.9.0)(cheerio@1.0.0)(encoding@0.1.13)(handlebars@4.7.8)(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3))(ws@8.18.1))(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)): + composio-core@0.5.31(@ai-sdk/openai@1.3.16(zod@3.24.4))(@cloudflare/workers-types@4.20250415.0)(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)))(@langchain/openai@0.5.5(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)))(encoding@0.1.13)(ws@8.18.1))(ai@4.3.15(react@19.1.0)(zod@3.24.4))(langchain@0.3.21(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)))(axios@1.9.0)(cheerio@1.0.0)(encoding@0.1.13)(handlebars@4.7.8)(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4))(ws@8.18.1))(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)): dependencies: - '@ai-sdk/openai': 1.3.16(zod@3.24.3) + '@ai-sdk/openai': 1.3.16(zod@3.24.4) '@cloudflare/workers-types': 4.20250415.0 '@composio/mcp': 1.0.3-0 '@hey-api/client-axios': 0.2.12(axios@1.8.4) - '@langchain/core': 0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)) - '@langchain/openai': 0.5.5(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)))(encoding@0.1.13)(ws@8.18.1) - ai: 4.3.13(react@19.1.0)(zod@3.24.3) + '@langchain/core': 0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)) + '@langchain/openai': 0.5.5(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)))(encoding@0.1.13)(ws@8.18.1) + ai: 4.3.15(react@19.1.0)(zod@3.24.4) axios: 1.8.4 chalk: 4.1.2 cli-progress: 3.12.0 commander: 12.1.0 inquirer: 10.2.2 - langchain: 0.3.21(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)))(axios@1.9.0)(cheerio@1.0.0)(encoding@0.1.13)(handlebars@4.7.8)(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3))(ws@8.18.1) + langchain: 0.3.21(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)))(axios@1.9.0)(cheerio@1.0.0)(encoding@0.1.13)(handlebars@4.7.8)(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4))(ws@8.18.1) open: 8.4.2 - openai: 4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3) + openai: 4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4) pusher-js: 8.4.0-rc2 resolve-package-path: 4.0.3 uuid: 10.0.0 @@ -29341,7 +29217,7 @@ snapshots: eslint: 9.24.0(jiti@2.4.2) eslint-import-resolver-node: 0.3.9 eslint-import-resolver-typescript: 3.10.1(eslint-plugin-import@2.31.0(@typescript-eslint/parser@8.31.0(eslint@9.24.0(jiti@2.4.2))(typescript@5.8.3))(eslint@9.24.0(jiti@2.4.2)))(eslint@9.24.0(jiti@2.4.2)) - eslint-plugin-import: 2.31.0(@typescript-eslint/parser@8.31.0(eslint@9.24.0(jiti@2.4.2))(typescript@5.8.3))(eslint-import-resolver-typescript@3.10.1(eslint-plugin-import@2.31.0(@typescript-eslint/parser@8.31.0(eslint@9.24.0(jiti@2.4.2))(typescript@5.8.3))(eslint@9.24.0(jiti@2.4.2)))(eslint@9.24.0(jiti@2.4.2)))(eslint@9.24.0(jiti@2.4.2)) + eslint-plugin-import: 2.31.0(@typescript-eslint/parser@8.31.0(eslint@9.24.0(jiti@2.4.2))(typescript@5.8.3))(eslint-import-resolver-typescript@3.10.1)(eslint@9.24.0(jiti@2.4.2)) eslint-plugin-jsx-a11y: 6.10.2(eslint@9.24.0(jiti@2.4.2)) eslint-plugin-react: 7.37.5(eslint@9.24.0(jiti@2.4.2)) eslint-plugin-react-hooks: 5.2.0(eslint@9.24.0(jiti@2.4.2)) @@ -29371,7 +29247,7 @@ snapshots: tinyglobby: 0.2.13 unrs-resolver: 1.6.2 optionalDependencies: - eslint-plugin-import: 2.31.0(@typescript-eslint/parser@8.31.0(eslint@9.24.0(jiti@2.4.2))(typescript@5.8.3))(eslint-import-resolver-typescript@3.10.1(eslint-plugin-import@2.31.0(@typescript-eslint/parser@8.31.0(eslint@9.24.0(jiti@2.4.2))(typescript@5.8.3))(eslint@9.24.0(jiti@2.4.2)))(eslint@9.24.0(jiti@2.4.2)))(eslint@9.24.0(jiti@2.4.2)) + eslint-plugin-import: 2.31.0(@typescript-eslint/parser@8.31.0(eslint@9.24.0(jiti@2.4.2))(typescript@5.8.3))(eslint-import-resolver-typescript@3.10.1)(eslint@9.24.0(jiti@2.4.2)) transitivePeerDependencies: - supports-color @@ -29441,7 +29317,7 @@ snapshots: - eslint-import-resolver-webpack - supports-color - eslint-plugin-import@2.31.0(@typescript-eslint/parser@8.31.0(eslint@9.24.0(jiti@2.4.2))(typescript@5.8.3))(eslint-import-resolver-typescript@3.10.1(eslint-plugin-import@2.31.0(@typescript-eslint/parser@8.31.0(eslint@9.24.0(jiti@2.4.2))(typescript@5.8.3))(eslint@9.24.0(jiti@2.4.2)))(eslint@9.24.0(jiti@2.4.2)))(eslint@9.24.0(jiti@2.4.2)): + eslint-plugin-import@2.31.0(@typescript-eslint/parser@8.31.0(eslint@9.24.0(jiti@2.4.2))(typescript@5.8.3))(eslint-import-resolver-typescript@3.10.1)(eslint@9.24.0(jiti@2.4.2)): dependencies: '@rtsao/scc': 1.1.0 array-includes: 3.1.8 @@ -31712,7 +31588,7 @@ snapshots: lodash: 4.17.21 minimist: 1.2.8 prettier: 3.5.3 - tinyglobby: 0.2.12 + tinyglobby: 0.2.13 json-schema-traverse@0.4.1: {} @@ -31856,15 +31732,15 @@ snapshots: - openai - ws - langchain@0.3.21(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)))(axios@1.9.0)(cheerio@1.0.0)(encoding@0.1.13)(handlebars@4.7.8)(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3))(ws@8.18.1): + langchain@0.3.21(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)))(axios@1.9.0)(cheerio@1.0.0)(encoding@0.1.13)(handlebars@4.7.8)(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4))(ws@8.18.1): dependencies: - '@langchain/core': 0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)) - '@langchain/openai': 0.5.5(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)))(encoding@0.1.13)(ws@8.18.1) - '@langchain/textsplitters': 0.1.0(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3))) + '@langchain/core': 0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)) + '@langchain/openai': 0.5.5(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)))(encoding@0.1.13)(ws@8.18.1) + '@langchain/textsplitters': 0.1.0(@langchain/core@0.3.44(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4))) js-tiktoken: 1.0.20 js-yaml: 4.1.0 jsonpointer: 5.0.1 - langsmith: 0.3.21(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3)) + langsmith: 0.3.21(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)) openapi-types: 12.1.3 p-retry: 4.6.2 uuid: 10.0.0 @@ -31900,6 +31776,18 @@ snapshots: optionalDependencies: openai: 4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.3) + langsmith@0.3.21(openai@4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4)): + dependencies: + '@types/uuid': 10.0.0 + chalk: 4.1.2 + console-table-printer: 2.12.1 + p-queue: 6.6.2 + p-retry: 4.6.2 + semver: 7.7.1 + uuid: 10.0.0 + optionalDependencies: + openai: 4.94.0(encoding@0.1.13)(ws@8.18.1)(zod@3.24.4) + language-subtag-registry@0.3.23: {} language-tags@1.0.9: @@ -33587,13 +33475,13 @@ snapshots: - acorn - supports-color - next-safe-action@7.10.8(next@15.3.1(@babel/core@7.26.10)(@opentelemetry/api@1.9.0)(@playwright/test@1.51.1)(babel-plugin-macros@3.1.0)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)(sass@1.77.4))(react-dom@19.1.0(react@19.1.0))(react@19.1.0)(zod@3.24.3): + next-safe-action@7.10.8(next@15.3.1(@babel/core@7.26.10)(@opentelemetry/api@1.9.0)(@playwright/test@1.51.1)(babel-plugin-macros@3.1.0)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)(sass@1.77.4))(react-dom@19.1.0(react@19.1.0))(react@19.1.0)(zod@3.24.4): dependencies: next: 15.3.1(@babel/core@7.26.10)(@opentelemetry/api@1.9.0)(@playwright/test@1.51.1)(babel-plugin-macros@3.1.0)(react-dom@19.1.0(react@19.1.0))(react@19.1.0)(sass@1.77.4) react: 19.1.0 react-dom: 19.1.0(react@19.1.0) optionalDependencies: - zod: 3.24.3 + zod: 3.24.4 next-themes@0.4.6(react-dom@19.1.0(react@19.1.0))(react@19.1.0): dependencies: @@ -33675,8 +33563,8 @@ snapshots: react-compiler-runtime: 0.0.0-experimental-22c6e49-20241219(react@19.1.0) react-dom: 19.1.0(react@19.1.0) scroll-into-view-if-needed: 3.1.0 - zod: 3.24.3 - zod-validation-error: 3.4.0(zod@3.24.3) + zod: 3.24.4 + zod-validation-error: 3.4.0(zod@3.24.4) zustand: 5.0.3(@types/react@18.3.20)(react@19.1.0)(use-sync-external-store@1.5.0(react@19.1.0)) transitivePeerDependencies: - '@types/react' @@ -33724,8 +33612,8 @@ snapshots: unist-util-visit: 5.0.0 unist-util-visit-children: 3.0.0 yaml: 2.7.1 - zod: 3.24.3 - zod-validation-error: 3.4.0(zod@3.24.3) + zod: 3.24.4 + zod-validation-error: 3.4.0(zod@3.24.4) transitivePeerDependencies: - acorn - supports-color @@ -36255,6 +36143,12 @@ snapshots: mrmime: 2.0.1 totalist: 3.0.1 + sirv@3.0.1: + dependencies: + '@polka/url': 1.0.0-next.29 + mrmime: 2.0.1 + totalist: 3.0.1 + sisteransi@1.0.5: {} skin-tone@2.0.0: @@ -37840,19 +37734,19 @@ snapshots: tsx: 4.19.3 yaml: 2.7.1 - vitest-environment-miniflare@2.14.4(vitest@1.6.1(@types/node@22.14.0)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0)): + vitest-environment-miniflare@2.14.4(vitest@1.6.1(@types/node@22.14.0)(@vitest/ui@3.1.1)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0)): dependencies: '@miniflare/queues': 2.14.4 '@miniflare/runner-vm': 2.14.4 '@miniflare/shared': 2.14.4 '@miniflare/shared-test-environment': 2.14.4 undici: 5.28.4 - vitest: 1.6.1(@types/node@22.14.0)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0) + vitest: 1.6.1(@types/node@22.14.0)(@vitest/ui@3.1.1)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0) transitivePeerDependencies: - bufferutil - utf-8-validate - vitest@1.6.1(@types/node@22.14.0)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0): + vitest@1.6.1(@types/node@22.14.0)(@vitest/ui@3.1.1)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0): dependencies: '@vitest/expect': 1.6.1 '@vitest/runner': 1.6.1 @@ -37876,6 +37770,7 @@ snapshots: why-is-node-running: 2.3.0 optionalDependencies: '@types/node': 22.14.0 + '@vitest/ui': 3.1.1(vitest@3.1.1) jsdom: 26.1.0 transitivePeerDependencies: - less @@ -37887,7 +37782,7 @@ snapshots: - supports-color - terser - vitest@3.1.1(@types/debug@4.1.12)(@types/node@20.17.30)(jiti@2.4.2)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0)(tsx@4.19.3)(yaml@2.7.1): + vitest@3.1.1(@types/debug@4.1.12)(@types/node@20.17.30)(@vitest/ui@3.1.1)(jiti@2.4.2)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0)(tsx@4.19.3)(yaml@2.7.1): dependencies: '@vitest/expect': 3.1.1 '@vitest/mocker': 3.1.1(vite@6.2.6(@types/node@22.14.0)(jiti@2.4.2)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0)(tsx@4.19.3)(yaml@2.7.1)) @@ -37912,6 +37807,7 @@ snapshots: optionalDependencies: '@types/debug': 4.1.12 '@types/node': 20.17.30 + '@vitest/ui': 3.1.1(vitest@3.1.1) jsdom: 26.1.0 transitivePeerDependencies: - jiti @@ -37927,7 +37823,7 @@ snapshots: - tsx - yaml - vitest@3.1.1(@types/debug@4.1.12)(@types/node@22.14.0)(jiti@2.4.2)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0)(tsx@4.19.3)(yaml@2.7.1): + vitest@3.1.1(@types/debug@4.1.12)(@types/node@22.14.0)(@vitest/ui@3.1.1)(jiti@2.4.2)(jsdom@26.1.0)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0)(tsx@4.19.3)(yaml@2.7.1): dependencies: '@vitest/expect': 3.1.1 '@vitest/mocker': 3.1.1(vite@6.2.6(@types/node@22.14.0)(jiti@2.4.2)(lightningcss@1.29.2)(sass@1.77.4)(terser@5.39.0)(tsx@4.19.3)(yaml@2.7.1)) @@ -37952,6 +37848,7 @@ snapshots: optionalDependencies: '@types/debug': 4.1.12 '@types/node': 22.14.0 + '@vitest/ui': 3.1.1(vitest@3.1.1) jsdom: 26.1.0 transitivePeerDependencies: - jiti @@ -38437,9 +38334,9 @@ snapshots: dependencies: zod: 3.24.4 - zod-validation-error@3.4.0(zod@3.24.3): + zod-validation-error@3.4.0(zod@3.24.4): dependencies: - zod: 3.24.3 + zod: 3.24.4 zod@3.22.3: {} diff --git a/scripts/standardize-sdk-packages.js b/scripts/standardize-sdk-packages.js index bbd5a3693..384001457 100755 --- a/scripts/standardize-sdk-packages.js +++ b/scripts/standardize-sdk-packages.js @@ -47,15 +47,19 @@ for (const dir of sdkDirs) { modified = true } - if (dir === 'apis.do' && packageJson.bin && packageJson.bin.apis) { - const distSrcPath = path.join(sdksDir, dir, 'dist', 'src') - const distBinPath = path.join(distSrcPath, 'bin.js') + if (packageJson.bin) { + for (const [binName, binPath] of Object.entries(packageJson.bin)) { + const distPath = path.join(sdksDir, dir, 'dist'); + const binFile = binPath.replace('./dist/', ''); + const distBinPath = path.join(distPath, binFile); - if (fs.existsSync(distSrcPath) && fs.existsSync(distBinPath)) { - if (packageJson.bin.apis !== './dist/src/bin.js') { - console.log(`Fixing bin path for apis.do`) - packageJson.bin.apis = './dist/src/bin.js' - modified = true + if (fs.existsSync(distPath)) { + if (!fs.existsSync(distBinPath)) { + console.log(`Binary file ${binFile} is missing for ${dir}`); + } else if (!fs.statSync(distBinPath).mode & 0o111) { + console.log(`Setting executable permissions for ${binFile} in ${dir}`); + fs.chmodSync(distBinPath, '755'); + } } } } diff --git a/sdks/actions.do/src/constants.ts b/sdks/actions.do/src/constants.ts index c2731877e..8e39ba6a7 100644 --- a/sdks/actions.do/src/constants.ts +++ b/sdks/actions.do/src/constants.ts @@ -1,10 +1,2 @@ -/** - * Generated Action constants - * DO NOT EDIT DIRECTLY - This file is generated at build time - */ - -export const ACTION_NAMES = [ - "github.createIssue" -] as const; - +export const ACTION_NAMES = [] as const; export type ActionName = (typeof ACTION_NAMES)[number]; diff --git a/sdks/integrations.do/tsup.config.ts b/sdks/integrations.do/tsup.config.ts index 50f0be47d..815e46706 100644 --- a/sdks/integrations.do/tsup.config.ts +++ b/sdks/integrations.do/tsup.config.ts @@ -2,6 +2,9 @@ import { defineConfig } from 'tsup' export default defineConfig({ entry: ['index.ts', 'types.ts', 'bin.ts'], + banner: { + js: '#!/usr/bin/env node' + }, format: ['esm'], dts: true, clean: true, diff --git a/sdks/llm.do/package.json b/sdks/llm.do/package.json index 9fe23b8a4..0d713d324 100644 --- a/sdks/llm.do/package.json +++ b/sdks/llm.do/package.json @@ -9,5 +9,8 @@ "keywords": [], "author": "", "license": "ISC", - "packageManager": "pnpm@10.8.1" + "packageManager": "pnpm@10.8.1", + "devDependencies": { + "@vitest/ui": "3.1.1" + } } diff --git a/sdks/llm.do/src/index.ts b/sdks/llm.do/src/index.ts index 60d8d1a10..1d007ee9b 100644 --- a/sdks/llm.do/src/index.ts +++ b/sdks/llm.do/src/index.ts @@ -27,6 +27,7 @@ interface LLMProviderConstructorOptions { errorToMessage: (error: any) => string } defaultObjectGenerationMode: 'tool' + fetch: (url: string, init: RequestInit) => Promise } // TODO: Ask Nathan about the route. @@ -80,9 +81,28 @@ export const createLLMProvider = (options: LLMProviderOptions) => { errorToMessage: (error) => error.message, }, defaultObjectGenerationMode: 'tool', - } + fetch: async (url, init) => { + // Mixin our model options into the body. + // By default, AI SDK wont let us add properties that are not inside the OpenAI schema. + // So we're doing this to use our superset standard. + + const newBody = { + ...JSON.parse(init?.body as string ?? '{}'), + modelOptions: settings + } + + const response = await fetch( + url, + { + ...init, + body: JSON.stringify(newBody) + } + ) - // Use explicit cast to any to bypass excessive type checking + return response + } + } + return new OpenAICompatibleChatLanguageModel(modelId, settings ?? {}, providerOptions as any) } } diff --git a/sdks/llm.do/test/provider.test.ts b/sdks/llm.do/test/provider.test.ts index 8d88fa326..de0f84993 100644 --- a/sdks/llm.do/test/provider.test.ts +++ b/sdks/llm.do/test/provider.test.ts @@ -1,13 +1,88 @@ import { describe, it, expect } from 'vitest' -import { generateText, tool } from 'ai' +import { generateText, tool, embed, generateObject } from 'ai' import { createLLMProvider } from '../src' +import { getModel } from '@/pkgs/language-models' import { z } from 'zod' const llm = createLLMProvider({ baseURL: `${ process.env.NEXT_PREVIEW_URL ?? 'http://localhost:3000' }/llm` }) -describe('llm.do provider', () => { +const geminiToolFixPrompt = ' Do not ask for arguments to a tool, use your best judgement. If you are unsure, return null.' + +describe('llm.do Chat Completions 💭', () => { + // Basic functionality tests + it('should support basic text generation', async () => { + const result = await generateText({ + model: llm('gemini'), + prompt: 'Respond with a short greeting' + }) + + expect(result.text).toBeTruthy() + }) + + it('should route to a specific provider', async () => { + const result = await generateText({ + model: llm( + 'qwen3-32b', + { + providerPriorities: ['cost'] + } + ), + prompt: 'Respond with a short greeting' + }) + + const model = getModel( + 'qwen3-32b', + { + providerPriorities: ['cost'], + + } + ) + + expect(result.text).toBeTruthy() + // @ts-expect-error - body is not typed + expect(result.response.body?.provider.name).toBe(model.provider.name) + }) + + // Structured outputs + it('should support structured outputs', async () => { + const result = await generateObject({ + model: llm('gemini'), + prompt: 'Respond with a short greeting', + schema: z.object({ + greeting: z.string() + }) + }) + + expect(result.object.greeting).toBeTruthy() + }) + + // Currently broken inside AI SDK. + it.skip('should support structured outputs with tools', async () => { + const result = await generateObject({ + model: llm('gemini'), + prompt: 'Use the greeting tool to generate a greeting to return. Person name must be "Connor"' + geminiToolFixPrompt, + schema: z.object({ + greeting: z.string() + }), + tools: { + greetingTool: tool({ + description: 'A tool that returns a greeting', + parameters: z.object({ + personName: z.string() + }), + execute: async (args) => { + return `Hello, ${args.personName}!` + } + }) + } + }) + + expect(result.object.greeting).toBeTruthy() + }) + + // Simple tool tests it('should use external tools', async () => { const result = await generateText({ model: llm('gemini(testTool)'), @@ -22,7 +97,7 @@ describe('llm.do provider', () => { await generateText({ model: llm('gemini'), - prompt: 'Return the testingTool output. Dont ask for the arguments, use your best judgement.', + prompt: 'Return the testingTool output.' + geminiToolFixPrompt, tools: { testingTool: tool({ description: 'A tool that returns "Hello, World."', @@ -39,4 +114,105 @@ describe('llm.do provider', () => { expect(toolCallSuccessful).toBe(true) }) + + // Complex tool tests + it('should handle tools with complex parameter schemas', async () => { + let receivedInput = '' + + const response = await generateText({ + model: llm('gemini'), + prompt: 'Call the complexTool with a detailed message. Then return the result of the complexTool. You must use the complexTool.' + geminiToolFixPrompt, + tools: { + complexTool: tool({ + description: 'A tool that accepts complex parameters', + parameters: z.object({ + message: z.string().describe('A detailed message'), + options: z.object({ + priority: z.enum(['high', 'medium', 'low']).optional(), + tags: z.array(z.string()).optional() + }).optional() + }), + execute: async (params) => { + // Using type assertion to avoid type errors + const input = params as { message: string } + receivedInput = input.message + return `Processed: ${input.message}` + } + }) + }, + maxSteps: 3 + }) + + expect(receivedInput).toBeTruthy() + }) + + it('should handle both local tools and Composio tools', async () => { + const result = await generateText({ + model: llm('gpt-4.1(testTool)'), + prompt: 'Call the testTool with the argument "Hello, World.", and then pass that into the complexTool.', + tools: { + complexTool: tool({ + description: 'A tool that returns a string', + parameters: z.object({ + message: z.string() + }), + execute: async (args) => { + return '123 123 4321' + } + }) + } + }) + + expect(result.toolCalls.length).toBe(1) + }) + + it('should handle Composio tools', async () => { + const result = await generateText({ + model: llm('gpt-4.1(hackernews.getItemWithId)'), + prompt: 'Look up the article "43969827", and tell me the article title and url.' + }) + + expect(result.text).toBeTruthy() + expect(result.text.toLowerCase()).toContain('firefox') + }) + + it('should handle tool error conditions', async () => { + try { + // This should always throw. + await generateText({ + model: llm('gemini'), + prompt: 'Try to use the errorTool and handle any errors it returns.' + geminiToolFixPrompt, + tools: { + errorTool: tool({ + description: 'A tool that sometimes fails', + parameters: z.object({ + shouldFail: z.boolean().optional() + }), + execute: async (params) => { + // Using type assertion to avoid type errors + const input = params as { shouldFail?: boolean } + + throw new Error('Tool execution failed') + + return '' + } + }) + } + }) + } catch (error) { + expect(error).toBeDefined() + } + }) +}, 100_000) + +describe.skip('llm.do Embeddings API 🔍', () => { + it('should support basic text embedding', async () => { + const result = await embed({ + // @ts-expect-error - Embeddings are not supported on llm.do yet. + model: llm.embedding('gemini'), + value: ['Hello, world.'] + }) + + expect(result.embedding).toBeDefined() + }) }, 100_000) \ No newline at end of file diff --git a/sdks/sdk.do/package.json b/sdks/sdk.do/package.json index 1f9f002b8..1807a39b4 100644 --- a/sdks/sdk.do/package.json +++ b/sdks/sdk.do/package.json @@ -47,6 +47,7 @@ "scripts": { "build": "tsc", "build:packages": "tsc", + "postbuild": "chmod +x dist/bin.js", "test": "vitest run", "test:watch": "vitest", "dev": "tsc --watch", diff --git a/sdks/sdk.do/src/index.ts b/sdks/sdk.do/src/index.ts index 13a69e82a..758224260 100644 --- a/sdks/sdk.do/src/index.ts +++ b/sdks/sdk.do/src/index.ts @@ -1,3 +1,4 @@ export { API } from './client.js' export { CLI } from './cli.js' export * from './types.js' +export * from './bin.js' diff --git a/sdks/sdk.do/tsconfig.json b/sdks/sdk.do/tsconfig.json index 1a8d28f10..a6620fa02 100644 --- a/sdks/sdk.do/tsconfig.json +++ b/sdks/sdk.do/tsconfig.json @@ -2,8 +2,8 @@ "extends": "tsconfig/src/sdk.json", "compilerOptions": { "outDir": "dist", - "rootDir": "../.." + "rootDir": "." }, - "include": ["**/*.ts"], + "include": ["src/**/*.ts"], "exclude": ["node_modules", "dist", "**/*.test.ts"] } diff --git a/sdks/workflows.do/package.json b/sdks/workflows.do/package.json index f8f6413f5..9c62c6794 100644 --- a/sdks/workflows.do/package.json +++ b/sdks/workflows.do/package.json @@ -39,6 +39,7 @@ "scripts": { "build": "tsc", "build:packages": "tsc", + "postbuild": "chmod +x dist/index.js", "test": "vitest run" }, "dependencies": { diff --git a/tests/workers/api-tests/e2e/ORMwhitePaper.pdf b/tests/sdk/assets/ORMwhitePaper.pdf similarity index 100% rename from tests/workers/api-tests/e2e/ORMwhitePaper.pdf rename to tests/sdk/assets/ORMwhitePaper.pdf diff --git a/tests/sdk/assets/images/test-image.jpg b/tests/sdk/assets/images/test-image.jpg new file mode 100644 index 000000000..ef7ed0612 Binary files /dev/null and b/tests/sdk/assets/images/test-image.jpg differ diff --git a/tests/sdk/openai.test.ts b/tests/sdk/openai.test.ts new file mode 100644 index 000000000..2bc6a76e8 --- /dev/null +++ b/tests/sdk/openai.test.ts @@ -0,0 +1,620 @@ +import { generateText } from '@/sdks/gpt.do/dist' +import fs from 'fs' +import OpenAI from 'openai' +import type { ResponseInput } from 'openai/resources/responses/responses.mjs' +import { describe, expect, test } from 'vitest' + +const searchModels = ['gpt-4o-search-preview'] + +const models = ['openai/o4-mini', 'google/gemini-2.0-flash-001', 'anthropic/claude-3.7-sonnet'] + +const pdf = getPdfData('ORMwhitePaper.pdf') + +// Use to skip tests in development +const skipTests: ('pdf' | 'structured-outputs' | 'tools')[] = [] + +console.log( + `Skipping tests: ${skipTests.join(', ')}` +) + +const geminiToolFixPrompt = ' Do not ask for arguments to a tool, use your best judgement. If you are unsure, return null.' + +// Explicitly call out llm.do and OpenRouter key usage +const client = new OpenAI({ + apiKey: process.env.OPENROUTER_API_KEY, + // baseURL: 'https://openrouter.ai/api/v1', + // baseURL: 'http://llm.do/llm', + baseURL: 'http://localhost:3000/llm', + defaultHeaders: { + 'HTTP-Referer': 'https://workflows.do', // Site URL for rankings on openrouter.ai. + 'X-Title': 'Workflows.do Business-as-Code', // Site title for rankings on openrouter.ai. + }, +}) + +describe('OpenAI SDK Responses', () => { + const isMockKey = process.env.OPEN_ROUTER_API_KEY === 'mock-openrouter-key' + + test.fails.each(models)( + 'can create a response with %s', + async (model) => { + try { + const response = await client.responses.create({ input: 'Hello, world!', model }) + expect(response).toBeDefined() + + if (isMockKey) { + console.log(`Using mock key for ${model}, verifying response structure only`) + } else { + expect(response.error).toBeNull() + expect(response.id).toBeDefined() + expect(response.output_text).toMatch(/hello|hi/i) + console.log(`${model}: ${response.output_text}`) + } + } catch (error) { + if (isMockKey) { + const errorMessage = error instanceof Error ? error.message : String(error) + console.log(`Expected error with mock key for ${model}: ${errorMessage}`) + } else { + throw error + } + } + }, + 60000, + ) + + test.skipIf(skipTests.includes('pdf')).fails.each(models)( + 'can handle PDF input with %s', + async (model) => { + const input: ResponseInput = [ + { + role: 'user', + content: [ + { + type: 'input_file', + filename: 'ORMwhitePaper.pdf', + file_data: pdf, + }, + { + type: 'input_text', + text: 'Who is the author of the white paper?', + }, + ], + }, + ] + + const response = await client.responses.create({ model, input }) + console.log(`${model}: ${response.output_text}`) + expect(response).toBeDefined() + expect(response.error).toBeNull() + expect(response.id).toBeDefined() + expect(response.output_text).toContain('Halpin') + }, + 60000, + ) +}) + +describe('OpenAI SDK Chat Completions', () => { + const isMockKey = process.env.OPEN_ROUTER_API_KEY === 'mock-openrouter-key' + + test.each(models)( + 'can create a chat completion with %s', + async (model) => { + try { + const response = await client.chat.completions.create({ model, messages: [{ role: 'user', content: 'Hello, world!' }] }) + expect(response).toBeDefined() + + if (isMockKey) { + console.log(`Using mock key for ${model}, verifying response structure only`) + } else { + console.log(response) + expect(response.id).toBeDefined() + expect(response.choices[0].message.content).toMatch(/hello|hi/i) + console.log(`${model}: ${response.choices[0].message.content}`) + } + } catch (error) { + if (isMockKey) { + const errorMessage = error instanceof Error ? error.message : String(error) + console.log(`Expected error with mock key for ${model}: ${errorMessage}`) + } else { + throw error + } + } + }, + 60000, + ) + + test.skipIf(skipTests.includes('pdf')).each(models)( + 'can handle PDF input with %s', + async (model) => { + const response = await client.chat.completions.create({ + model, + messages: [ + { + role: 'user', + content: [ + { + type: 'file', + file: { + filename: 'ORMwhitePaper.pdf', + file_data: pdf, + }, + }, + { + type: 'text', + text: 'What are the steps of CSDP?', + }, + ], + }, + ], + }) + expect(response).toBeDefined() + expect(response.id).toBeDefined() + expect(response.choices[0].message.content).toContain('elementary fact') + console.log(`${model}: ${JSON.stringify(response.choices[0].message.content)}`) + }, + 60000, + ) + + test.skipIf(skipTests.includes('tools')).each(models)( + 'can use Composio tools using %s', + async (model) => { + const response = await client.chat.completions.create({ + model: `${model}(testTool)`, // Force the model to use the testTool + messages: [ + { + role: 'user', + content: 'You must return the result of the testTool. The message parameter must be "Hello, World.", dont ask for other parameters, do your best effort.', + }, + ], + }) + + expect(response).toBeDefined() + expect(response.id).toBeDefined() + expect(response.choices[0].message.content).toContain('Hello, World.') + }, + 60000, + ) + + test.skipIf(skipTests.includes('tools')).each(models)( + 'can use local tools using %s', + async (model) => { + const response = await client.chat.completions.create({ + model, + messages: [ + { + role: 'user', + content: 'You must return the result of the testTool. The message parameter must be "Hello, World.", dont ask for other parameters, do your best effort.', + }, + ], + tools: [ + { + type: 'function', + function: { + name: 'localTestingTool', + description: 'A testing tool.', + parameters: { type: 'object', properties: { message: { type: 'string', description: 'The message to return' } } }, + strict: true, + }, + }, + ], + }) + + console.log(response) + + expect(response).toBeDefined() + expect(response.id).toBeDefined() + expect(response.choices[0].message.content).toContain('Hello, World.') + }, + 60000, + ) + + test.skipIf(skipTests.includes('structured-outputs')).each(models)('can use structured outputs using %s', async (model) => { + const response = await client.chat.completions.create({ + model, + messages: [{ role: 'user', content: 'Hello, world!' }], + response_format: { type: 'json_schema', json_schema: { name: 'test', strict: true, schema: { type: 'object', properties: { test: { type: 'string' } } } } }, + }) + + expect(response).toBeDefined() + expect(response.id).toBeDefined() + expect(response.choices[0].message.content).toContain('{') + expect(response.choices[0].message.content).toMatch(/hello|hi/i) + console.log(`${model}: ${JSON.stringify(response.choices[0].message.content)}`) + }) +}) + +describe('OpenAI SDK Web Search', () => { + test.each(searchModels)( + 'can create a web search chat completion with %s', + async (model) => { + const response = await client.chat.completions.create({ + web_search_options: { + search_context_size: 'low', + user_location: { + type: 'approximate', + approximate: { city: 'South Saint Paul', country: 'US', region: 'Minnesota', timezone: 'America/Chicago' }, + }, + }, + model, + messages: [{ role: 'user', content: "What's the weather in South Saint Paul?" }], + }) + expect(response).toBeDefined() + console.log(`${model}: ${JSON.stringify(response.choices[0].message.content)}`) + expect(response.choices[0].message.content).toMatch(/°/) + }, + 60000, + ) +}) + +describe('OpenAI SDK Models', () => { + const isMockKey = process.env.OPEN_ROUTER_API_KEY === 'mock-openrouter-key' + + test.fails( + 'can list models', + async () => { + try { + const models = await client.models.list() + expect(models).toBeDefined() + + if (!isMockKey) { + expect(models.data.length).toBeGreaterThan(0) + } + } catch (error) { + if (isMockKey) { + const errorMessage = error instanceof Error ? error.message : String(error) + console.log(`Expected error with mock key for models.list: ${errorMessage}`) + } else { + throw error + } + } + }, + 60000, + ) +}) + +describe('OpenRouter Privacy and Logging', () => { + test.todo( + 'can create a chat completion with privacy enabled', + async () => { + const response = await client.chat.completions.create({ + model: models[0], + messages: [{ role: 'user', content: 'Hello, world!' }], + // @ts-expect-error OpenAI SDK does not support privacy feature + privacy: { + remove_metadata: true, + remove_pii: true, + store_message: false, + }, + }) + expect(response).toBeDefined() + expect(response.id).toBeDefined() + expect(response.choices[0].message.content).toBeTruthy() + console.log(`Privacy test: ${JSON.stringify(response.choices[0].message.content)}`) + }, + 60000, + ) +}) + +describe('OpenRouter Model Routing', () => { + test('can auto-route', async () => { + const response = await client.chat.completions.create({ + model: 'openrouter/auto', + messages: [{ role: 'user', content: 'Hello, world!' }], + }) + expect(response).toBeDefined() + expect(response.id).toBeDefined() + expect(response.model).not.toBe('openrouter/auto') + }, 60000) + + test('can route to specific models by name', async () => { + const response = await client.chat.completions.create({ + messages: [{ role: 'user', content: 'Hello, world!' }], + // @ts-expect-error OpenAI SDK does not support models array + models: models.slice(1, 2), + }) + expect(response).toBeDefined() + expect(response.id).toBeDefined() + expect(response.model).toBe(models[1]) + }, 60000) +}) + +describe('OpenRouter Provider Routing', () => { + test('can route to providers by sorting by price', async () => { + const response = await client.chat.completions.create({ + messages: [{ role: 'user', content: 'Hello, world!' }], + // @ts-expect-error OpenAI SDK does not support provider + provider: { + sort: 'price', + }, + }) + expect(response).toBeDefined() + expect(response.id).toBeDefined() + expect(response.model).not.toMatch(/openai/) + }, 60000) + + test('can route to specific providers', async () => { + const response = await client.chat.completions.create({ + model: 'mistralai/mistral-nemo', + messages: [{ role: 'user', content: 'Hello, world!' }], + // @ts-expect-error OpenAI SDK does not support provider + provider: { + only: ['Parasail'], + }, + }) + expect(response).toBeDefined() + expect(response.id).toBeDefined() + // @ts-expect-error OpenAI SDK does not support provider + expect(response.provider).toBe('Parasail') + }, 60000) +}) + +describe('OpenRouter Prompt Caching', () => { + test.skip.fails( + 'can create a chat completion with caching enabled', + async () => { + const firstResponse = await client.chat.completions.create({ + model: 'openai/o1-mini', + messages: [{ role: 'user', content: 'What is the capital of France?' }], + // @ts-expect-error OpenAI SDK does not support cache usage + usage: { include: true }, + }) + expect(firstResponse).toBeDefined() + + const secondResponse = await client.chat.completions.create({ + model: 'openai/o1-mini', + messages: [{ role: 'user', content: 'What is the capital of France?' }], + // @ts-expect-error OpenAI SDK does not support cache usage + usage: { include: true }, + }) + expect(secondResponse).toBeDefined() + + console.log(JSON.stringify(firstResponse, null, 2)) + console.log(JSON.stringify(secondResponse, null, 2)) + + expect(secondResponse.choices[0].message.content).toBe(firstResponse.choices[0].message.content) + expect(secondResponse.usage).toBeDefined() + expect(secondResponse.usage?.prompt_tokens_details).toBeDefined() + expect(secondResponse.usage?.prompt_tokens_details?.cached_tokens).toBeGreaterThan(0) + console.log(`Caching test: ${firstResponse.choices[0].message.content}`) + }, + 60000, + ) +}) + +describe('OpenRouter Structured Outputs', () => { + test.fails( + 'can create a chat completion with JSON schema', + async () => { + const response = await client.chat.completions.create({ + model: models[0], + messages: [{ role: 'user', content: 'List three French cities and their populations' }], + response_format: { + type: 'json_object', + }, + }) + expect(response).toBeDefined() + expect(response.choices[0].message.content).toBeTruthy() + const content = response.choices[0].message.content || '' + try { + if (content.includes('{')) { + const parsed = JSON.parse(content) + expect(parsed).toBeDefined() + if (parsed.cities) { + expect(parsed.cities.length).toBeGreaterThan(0) + } + console.log(`Structured output: ${content}`) + } + } catch (e) { + console.error('Failed to parse JSON response:', e) + } + }, + 60000, + ) +}) + +describe('OpenRouter Tool Calling', () => { + test( + 'can create a chat completion with tool calling', + async () => { + const response = await client.chat.completions.create({ + model: models[0], + messages: [ + { + role: 'user', + content: 'What is the weather in New York and London today?', + }, + ], + tools: [ + { + type: 'function', + function: { + name: 'get_weather', + description: 'Get the current weather in a given location', + parameters: { + type: 'object', + properties: { + location: { + type: 'string', + description: 'The city and state, e.g. San Francisco, CA', + }, + unit: { + type: 'string', + enum: ['celsius', 'fahrenheit'], + description: 'The unit of temperature to use', + }, + }, + required: ['location'], + }, + }, + }, + ], + tool_choice: 'auto', + }) + expect(response).toBeDefined() + expect(response.choices[0].message).toBeDefined() + if (response.choices[0].message.tool_calls) { + expect(response.choices[0].message.tool_calls.length).toBeGreaterThan(0) + expect(response.choices[0].message.tool_calls[0].function.name).toBe('get_weather') + console.log(`Tool calling: ${JSON.stringify(response.choices[0].message.tool_calls[0])}`) + } + }, + 60000, + ) +}) + +describe('OpenRouter Images & PDFs', () => { + const imagePath = `${__dirname}/assets/images/test-image.jpg` + const imageData = 'data:image/jpeg;base64,' + fs.readFileSync(imagePath, 'base64') + + test.fails( + 'can create a chat completion with image input', + async () => { + const response = await client.chat.completions.create({ + model: 'openai/gpt-4-vision-preview', + messages: [ + { + role: 'user', + content: [ + { + type: 'text', + text: 'What do you see in this image?', + }, + { + type: 'image_url', + image_url: { + url: imageData, + }, + }, + ], + }, + ], + }) + expect(response).toBeDefined() + expect(response.choices[0].message.content).toBeTruthy() + console.log(`Image input: ${response.choices[0].message.content}`) + }, + 60000, + ) +}) + +describe('OpenRouter Message Transforms', () => { + test.todo( + 'can create a chat completion with message transforms', + async () => { + const response = await client.chat.completions.create({ + model: models[0], + messages: [{ role: 'user', content: 'Hello, world!' }], + // @ts-expect-error OpenAI SDK does not support message transforms + transforms: [ + { + type: 'context_window', + target: 64000, + }, + ], + }) + expect(response).toBeDefined() + expect(response.choices[0].message.content).toBeTruthy() + console.log(`Message transforms: ${response.choices[0].message.content}`) + }, + 60000, + ) +}) + +describe('OpenRouter Uptime Optimization', () => { + test.todo( + 'can create a chat completion with uptime optimization', + async () => { + const response = await client.chat.completions.create({ + model: models[0], + messages: [{ role: 'user', content: 'Hello, world!' }], + // @ts-expect-error OpenAI SDK does not support uptime optimization + route_option: 'high_uptime', + }) + expect(response).toBeDefined() + expect(response.choices[0].message.content).toBeTruthy() + console.log(`Uptime optimization: ${response.choices[0].message.content}`) + }, + 60000, + ) +}) + +describe('OpenRouter Web Search', () => { + test.todo( + 'can create a chat completion with web search', + async () => { + const response = await client.chat.completions.create({ + model: searchModels[0], + messages: [{ role: 'user', content: 'What are the latest news about AI today?' }], + web_search_options: { + search_context_size: 'high', + user_location: { + type: 'approximate', + approximate: { + city: 'San Francisco', + country: 'US', + region: 'California', + timezone: 'America/Los_Angeles', + }, + }, + }, + }) + expect(response).toBeDefined() + expect(response.choices[0].message).toBeDefined() + const content = response.choices[0].message.content || '' + expect(content.length).toBeGreaterThan(0) + console.log(`Web search: ${content.substring(0, Math.min(content.length, 100))}...`) + }, + 60000, + ) +}) + +describe('OpenRouter Zero Completion Insurance', () => { + test.todo( + 'can create a chat completion with zero completion insurance', + async () => { + const response = await client.chat.completions.create({ + model: models[0], + messages: [{ role: 'user', content: 'Hello, world!' }], + // @ts-expect-error OpenAI SDK does not support zero completion insurance + zero_completion_insurance: true, + }) + expect(response).toBeDefined() + expect(response.choices[0].message.content).toBeTruthy() + console.log(`Zero completion insurance: ${response.choices[0].message.content}`) + }, + 60000, + ) +}) + +describe('OpenRouter Provisioning API Keys', () => { + test.fails( + 'can provision API keys for providers', + async () => { + // @ts-expect-error OpenAI SDK does not support provisioning API keys + const response = await client.keys.create({ + provider: 'openai', + key: 'sk-test-key', + name: 'Test OpenAI Key', + }) + expect(response).toBeDefined() + expect(response.id).toBeDefined() + console.log(`Provisioned key: ${response.id}`) + }, + 60000, + ) + + test.fails( + 'can list provisioned API keys', + async () => { + // @ts-expect-error OpenAI SDK does not support listing API keys + const response = await client.keys.list() + expect(response).toBeDefined() + expect(response.data).toBeDefined() + console.log(`Listed keys count: ${response.data.length}`) + }, + 60000, + ) +}) + +function getPdfData(filename: string) { + return 'data:application/pdf;base64,' + fs.readFileSync(`${__dirname}/assets/${filename}`, 'base64') +} diff --git a/tests/workers/api-tests/e2e/openai.test.ts b/tests/workers/api-tests/e2e/openai.test.ts deleted file mode 100644 index 845cabc05..000000000 --- a/tests/workers/api-tests/e2e/openai.test.ts +++ /dev/null @@ -1,146 +0,0 @@ -import fs from 'fs' -import OpenAI from 'openai' -import type { ResponseInput } from 'openai/resources/responses/responses.mjs' -import { describe, expect, test } from 'vitest' - -const models = ['openai/o4-mini', 'google/gemini-2.0-flash-001', 'anthropic/claude-3.7-sonnet'] - -// Explicitly call out llm.do and OpenRouter key usage -const client = new OpenAI({ - apiKey: process.env.OPENROUTER_API_KEY, - // baseURL: 'https://openrouter.ai/api/v1', - baseURL: 'https://llm.do/api/v1', - // baseURL: 'http://127.0.0.1:8787/api/v1', - defaultHeaders: { - 'HTTP-Referer': 'https://workflows.do', // Site URL for rankings on openrouter.ai. - 'X-Title': 'Workflows.do Business-as-Code', // Site title for rankings on openrouter.ai. - }, -}) - -describe.skip('OpenAI SDK Responses', () => { - const isMockKey = process.env.OPENROUTER_API_KEY === 'mock-openrouter-key' - - test.each(models)('can create a response with %s', async (model) => { - try { - const response = await client.responses.create({ input: 'Hello, world!', model }) - expect(response).toBeDefined() - - if (isMockKey) { - console.log(`Using mock key for ${model}, verifying response structure only`) - } else { - expect(response.error).toBeNull() - expect(response.id).toBeDefined() - expect(response.output_text).toMatch(/hello|hi/i) - console.log(`${model}: ${response.output_text}`) - } - } catch (error) { - if (isMockKey) { - const errorMessage = error instanceof Error ? error.message : String(error); - console.log(`Expected error with mock key for ${model}: ${errorMessage}`) - } else { - throw error - } - } - }) - - // Currently fails due to OpenRouter not supporting PDFs - test.fails.skip.each(models)( - 'can handle PDF input with %s', - async (model) => { - const input: ResponseInput = [ - { - role: 'user', - content: [ - { - type: 'input_file', - filename: 'ORMwhitePaper.pdf', - file_data: 'data:application/pdf;base64,' + fs.readFileSync(`${__dirname}/ORMwhitePaper.pdf`, 'base64'), - }, - { - type: 'input_text', - text: 'What are the steps of CSDP?', - }, - ], - }, - ] - - const response = await client.responses.create({ model, input }) - console.log(`${model}: ${response.output_text}`) - expect(response).toBeDefined() - expect(response.error).toBeNull() - expect(response.id).toBeDefined() - expect(response.output_text).toContain('elementary fact') - }, - 60000, - ) -}) - -describe('OpenAI SDK Chat Completions', () => { - const isMockKey = process.env.OPENROUTER_API_KEY === 'mock-openrouter-key'; - - test.each(models)('can create a chat completion with %s', async (model) => { - try { - const response = await client.chat.completions.create({ model, messages: [{ role: 'user', content: 'Hello, world!' }] }) - expect(response).toBeDefined() - - if (isMockKey) { - console.log(`Using mock key for ${model}, verifying response structure only`) - } else { - expect(response.id).toBeDefined() - expect(response.choices[0].message.content).toMatch(/hello|hi/i) - console.log(`${model}: ${response.choices[0].message.content}`) - } - } catch (error) { - if (isMockKey) { - const errorMessage = error instanceof Error ? error.message : String(error); - console.log(`Expected error with mock key for ${model}: ${errorMessage}`) - } else { - throw error - } - } - }) - - // Currently fails due to OpenRouter not supporting PDFs - test.fails.skip.each(models)('can handle PDF input with %s', async (model) => { - const response = await client.chat.completions.create({ model, messages: [{ role: 'user', content: 'Hello, world!' }] }) - expect(response).toBeDefined() - expect(response.id).toBeDefined() - expect(response.choices[0].message.content).toContain('elementary fact') - console.log(`${model}: ${JSON.stringify(response.choices[0].message.content)}`) - }) - - // test.each(models)('can use structured outputs using %s', async (model) => { - // const response = await client.chat.completions.create({ - // model, - // messages: [{ role: 'user', content: 'Hello, world!' }], - // response_format: { type: 'json_schema', json_schema: { name: 'test', strict: true, schema: { type: 'object', properties: { test: { type: 'string' } } } } }, - // }) - // expect(response).toBeDefined() - // expect(response.id).toBeDefined() - // expect(response.choices[0].message.content).toContain('{') - // expect(response.choices[0].message.content).toMatch(/hello|hi/i) - // console.log(`${model}: ${JSON.stringify(response.choices[0].message.content)}`) - // }) -}) - -describe('OpenAI SDK Models', () => { - const isMockKey = process.env.OPENROUTER_API_KEY === 'mock-openrouter-key'; - - test.fails('can list models', async () => { - try { - const models = await client.models.list() - expect(models).toBeDefined() - - if (!isMockKey) { - expect(models.data.length).toBeGreaterThan(0) - } - } catch (error) { - if (isMockKey) { - const errorMessage = error instanceof Error ? error.message : String(error); - console.log(`Expected error with mock key for models.list: ${errorMessage}`) - } else { - throw error - } - } - }) -}) diff --git a/tests/workers/api-tests/e2e/vitest.config.ts b/tests/workers/api-tests/e2e/vitest.config.ts deleted file mode 100644 index 949c80986..000000000 --- a/tests/workers/api-tests/e2e/vitest.config.ts +++ /dev/null @@ -1,21 +0,0 @@ -import { defineConfig } from 'vitest/config' -import { resolve } from 'path' - -export default defineConfig({ - test: { - environment: 'node', - setupFiles: ['dotenv/config'], - }, - resolve: { - alias: { - '@payload-config': resolve(__dirname, '../../../../payload.config.ts'), - '@': resolve(__dirname, '../../../..'), - '@/lib': resolve(__dirname, '../../../../lib'), - '@/.velite': resolve(__dirname, '../../../../.velite'), - '@/app': resolve(__dirname, '../../../../app'), - '@/collections': resolve(__dirname, '../../../../collections'), - '@/tasks': resolve(__dirname, '../../../../tasks'), - '@/scripts': resolve(__dirname, '../../../../scripts'), - }, - }, -}) diff --git a/tests/workers/api-tests/package.json b/tests/workers/api-tests/package.json deleted file mode 100644 index cbab400d0..000000000 --- a/tests/workers/api-tests/package.json +++ /dev/null @@ -1,19 +0,0 @@ -{ - "name": "api-tests", - "version": "1.0.0", - "description": "", - "scripts": { - "test": "vitest" - }, - "keywords": [], - "author": "", - "license": "ISC", - "packageManager": "pnpm@10.7.0", - "dependencies": { - "openai": "^4.90.0" - }, - "devDependencies": { - "dotenv": "^16.4.7", - "vitest": "^3.0.9" - } -} diff --git a/tests/workers/api-tests/vitest.config.ts b/tests/workers/api-tests/vitest.config.ts deleted file mode 100644 index d3b0d2d39..000000000 --- a/tests/workers/api-tests/vitest.config.ts +++ /dev/null @@ -1,21 +0,0 @@ -import { defineConfig } from 'vitest/config' -import { resolve } from 'path' - -export default defineConfig({ - test: { - environment: 'node', - setupFiles: ['dotenv/config'], - }, - resolve: { - alias: { - '@payload-config': resolve(__dirname, '../../../payload.config.ts'), - '@': resolve(__dirname, '../../..'), - '@/lib': resolve(__dirname, '../../../lib'), - '@/.velite': resolve(__dirname, '../../../.velite'), - '@/app': resolve(__dirname, '../../../app'), - '@/collections': resolve(__dirname, '../../../collections'), - '@/tasks': resolve(__dirname, '../../../tasks'), - '@/scripts': resolve(__dirname, '../../../scripts'), - }, - }, -})