LiteLLM issue summary - 2026-03-26 #33
Description
Open Issues from berriai/litellm
Retrieved on: 2026-03-26
Total issues displayed: 34
Issue List
#24639 - docs: add Supply Chain Security Audit guide
- Type: Pull Request
- Created: 2026-03-26T16:18:00Z
- Author: sayanget
- Link: docs: add Supply Chain Security Audit guide BerriAI/litellm#24639
- Description: Adds a dedicated 'Supply Chain Security Audit' page to the documentation following the March 2026 PyPI incident.
#24638 - fix: use redis_kwargs host/port in cache ping health check (#24636)
- Type: Pull Request
- Created: 2026-03-26T15:55:26Z
- Author: sjhddh
- Link: fix: use redis_kwargs host/port in cache ping health check (#24636) BerriAI/litellm#24638
- Description: Fixes #24636 - The /cache/ping endpoint was returning None for host and port.
#24637 - Fix overlong S3 logging object keys
- Type: Pull Request
- Created: 2026-03-26T15:35:17Z
- Author: raashish1601
- Link: Fix overlong S3 logging object keys BerriAI/litellm#24637
- Description: Fixes #24628 - Cap the final S3 filename component to a filesystem-safe length.
#24636 - Cache health check returns None for host and port despite working Redis connection
- Type: Bug
- Created: 2026-03-26T15:20:18Z
- Author: ThePlenkov
- Labels: bug, proxy
- Link: Cache health check returns None for host and port despite working Redis connection BerriAI/litellm#24636
- Description: The /cache/ping endpoint returns "None" for host and port in health_check_cache_params, even though Redis is working.
#24635 - feat(volcengine): add image generation support for Ark/Seedream
- Type: Pull Request
- Created: 2026-03-26T14:31:25Z
- Author: forrestIsRunning
- Link: feat(volcengine): add image generation support for Ark/Seedream BerriAI/litellm#24635
- Description: Adds first-class image generation support for VolcEngine (ByteDance Ark) Seedream models.
#24634 - fix(security): add SSRF protection to custom code guardrail HTTP primitives
- Type: Pull Request
- Created: 2026-03-26T14:20:22Z
- Author: 1Ckpwee
- Link: fix(security): add SSRF protection to custom code guardrail HTTP primitives BerriAI/litellm#24634
- Description: Adds SSRF protection to prevent requests to cloud metadata services and internal networks.
#24633 - Litellm fix opus tests
- Type: Pull Request
- Created: 2026-03-26T14:17:08Z
- Author: Sameerlite
- Link: Litellm fix opus tests BerriAI/litellm#24633
#24632 - Fix tests
- Type: Pull Request
- Created: 2026-03-26T14:13:21Z
- Author: Sameerlite
- Link: Fix tests BerriAI/litellm#24632
#24631 - feat(models): add openrouter/minimax/minimax-m2.7 model pricing
- Type: Pull Request
- Created: 2026-03-26T14:09:30Z
- Author: 1Ckpwee
- Link: feat(models): add openrouter/minimax/minimax-m2.7 model pricing BerriAI/litellm#24631
- Description: Adds pricing and context window information for the minimax-m2.7 model.
#24629 - feat(models): add openrouter/minimax/minimax-m2.7 to model pricing registry
- Type: Pull Request
- Created: 2026-03-26T13:11:27Z
- Author: Retr0-XD
- Link: feat(models): add openrouter/minimax/minimax-m2.7 to model pricing registry BerriAI/litellm#24629
- Description: Adds the openrouter/minimax/minimax-m2.7 model to pricing registry.
#24628 - [Bug]: S3 filename too long
- Type: Bug
- Created: 2026-03-26T12:53:04Z
- Author: rodriciru
- Labels: bug, proxy
- Link: [Bug]: S3 filename too long BerriAI/litellm#24628
- Description: S3 logging fails with 500 error due to filename length exceeding 255 characters.
#24627 - [Bug]: Pass-through multipart audio transcription endpoint returns UnicodeDecodeError
- Type: Bug
- Created: 2026-03-26T12:29:20Z
- Author: nhyy244
- Labels: bug, proxy
- Link: [Bug]: Pass-through multipart audio transcription endpoint returns UnicodeDecodeError BerriAI/litellm#24627
- Description: Pass-through endpoint for Scaleway audio transcription returns UnicodeDecodeError.
#24626 - [Bug]: Gemini file retrieval fails: Error parsing file retrieve response
- Type: Bug
- Created: 2026-03-26T11:16:40Z
- Author: MyroslavaTarcha
- Labels: bug, proxy, llm translation
- Link: [Bug]: Gemini file retrieval fails: Error parsing file retrieve response BerriAI/litellm#24626
- Description: GET /v1/files/{file_id} fails with 500 error for Gemini provider.
#24625 - fix(mcp): block arbitrary command execution via stdio transport
- Type: Pull Request
- Created: 2026-03-26T10:27:37Z
- Author: Sameerlite
- Link: fix(mcp): block arbitrary command execution via stdio transport BerriAI/litellm#24625
- Description: Security fix - adds command allowlist for MCP stdio test endpoints.
#24624 - fix(proxy): sanitize user_id input and block dangerous env var keys
- Type: Pull Request
- Created: 2026-03-26T10:04:30Z
- Author: Sameerlite
- Link: fix(proxy): sanitize user_id input and block dangerous env var keys BerriAI/litellm#24624
- Description: Security hardening for input validation gaps in proxy server.
#24623 - fix docker documentation MASTER_KEY -> LITELLM_MASTER_KEY
- Type: Pull Request
- Created: 2026-03-26T09:35:14Z
- Author: D0wn3r
- Link: fix docker documentation MASTER_KEY -> LITELLM_MASTER_KEY BerriAI/litellm#24623
- Description: Documentation fix for correct environment variable name.
#24622 - fix(vertex_ai): Gemini tool-use prompt tokens ignored, causing wrong usage counts
- Type: Pull Request
- Created: 2026-03-26T09:18:28Z
- Author: SyedShahmeerAli12
- Link: fix(vertex_ai): Gemini tool-use prompt tokens ignored, causing wrong usage counts BerriAI/litellm#24622
- Description: Fixes toolUsePromptTokenCount being ignored in Gemini usage metadata.
#24621 - [Bug]: Cannot generate 2K images with Gemini 3.1 Flash Image Preview - extra_body is stripped
- Type: Bug
- Created: 2026-03-26T09:09:15Z
- Author: ieQ-strecker
- Labels: bug, proxy, llm translation
- Link: [Bug]: Cannot generate 2K images with Gemini 3.1 Flash Image Preview (stuck at 1K) - extra_body is stripped BerriAI/litellm#24621
- Description: extra_body parameter for imageConfig is stripped, forcing 1K images instead of 2K.
#24620 - fix: add openrouter/minimax/minimax-m2.7 model pricing
- Type: Pull Request
- Created: 2026-03-26T08:31:28Z
- Author: WhoisMonesh
- Link: fix: add openrouter/minimax/minimax-m2.7 model pricing BerriAI/litellm#24620
- Description: Adds pricing for minimax-m2.7 model from OpenRouter.
#24618 - fix(ollama): preserve image_url blocks in ollama_chat multimodal requests
- Type: Pull Request
- Created: 2026-03-26T08:05:56Z
- Author: WhoisMonesh
- Link: fix(ollama): preserve image_url blocks in ollama_chat multimodal requests BerriAI/litellm#24618
- Description: Fixes image_url content blocks being dropped in ollama_chat multimodal requests.
#24617 - fix: Bedrock internalServerException mapping, AuthError no-retry, xai drop_params, SSE error handling
- Type: Pull Request
- Created: 2026-03-26T08:01:14Z
- Author: naarob
- Link: fix: Bedrock internalServerException mapping, AuthError no-retry, xai drop_params, SSE error handling BerriAI/litellm#24617
- Description: Four independent bug fixes for error handling and routing correctness.
#24616 - fix(router): 429 routing — cooldown bypass, providers.json mapping, Anthropic credit balance fallback
- Type: Pull Request
- Created: 2026-03-26T07:34:36Z
- Author: naarob
- Link: fix(router): 429 routing — cooldown bypass, providers.json mapping, Anthropic credit balance fallback BerriAI/litellm#24616
- Description: Three related fixes for 429/rate-limit routing failures.
#24615 - fix(ollama): preserve image_url blocks in ollama_chat multimodal requests
- Type: Pull Request
- Created: 2026-03-26T07:26:13Z
- Author: NIK-TIGER-BILL
- Link: fix(ollama): preserve image_url blocks in ollama_chat multimodal requests BerriAI/litellm#24615
- Description: Fixes image_url blocks being dropped in Ollama multimodal requests.
#24613 - Feature/add hpc ai provider
- Type: Pull Request
- Created: 2026-03-26T07:08:36Z
- Author: lioZ129
- Link: Feature/add hpc ai provider BerriAI/litellm#24613
- Description: Adds HPC-AI as an OpenAI-compatible provider.
#24612 - fix(model): add supports_reasoning for gemini-3.1-flash-image-preview
- Type: Pull Request
- Created: 2026-03-26T06:57:12Z
- Author: Ryze0323
- Link: fix(model): add supports_reasoning for gemini-3.1-flash-image-preview BerriAI/litellm#24612
- Description: Adds missing supports_reasoning flag for Gemini image preview model.
#24611 - feat(router): order-based fallback across deployment priority levels
- Type: Pull Request
- Created: 2026-03-26T06:25:16Z
- Author: Sameerlite
- Link: feat(router): order-based fallback across deployment priority levels BerriAI/litellm#24611
- Description: Implements automatic fallback across order levels when deployments fail.
#24610 - feat(gemini): Lyria 3 preview models in cost map and docs
- Type: Pull Request
- Created: 2026-03-26T06:11:26Z
- Author: Sameerlite
- Link: feat(gemini): Lyria 3 preview models in cost map and docs BerriAI/litellm#24610
- Description: Adds Lyria 3 preview models to cost map and documentation.
#24609 - [Bug]: No Error Handling in /v1/messages Path
- Type: Bug
- Created: 2026-03-26T05:52:56Z
- Author: urainshah
- Labels: bug, proxy, llm translation
- Link: [Bug]: No Error Handling in /v1/messages Path BerriAI/litellm#24609
- Description: async_sse_wrapper has no try/except block, causing raw Bedrock errors to propagate.
#24608 - [Bug]: [Bedrock] internalServerException mid-stream error incorrectly mapped to BadRequestError
- Type: Bug
- Created: 2026-03-26T05:41:15Z
- Author: urainshah
- Labels: bug, proxy, llm translation
- Link: [Bug]: [Bedrock] internalServerException mid-stream error incorrectly mapped to BadRequestError (400) instead of internalServerException (500) BerriAI/litellm#24608
- Description: Bedrock internalServerException mapped to BadRequestError (400) instead of 500.
#24606 - Fix Ollama model info URL normalization
- Type: Pull Request
- Created: 2026-03-26T05:05:03Z
- Author: LittleChenLiya
- Link: Fix Ollama model info URL normalization BerriAI/litellm#24606
- Description: Fixes Ollama model info URL construction when api_base already contains endpoint path.
Note: This summary includes both issues and pull requests that are currently open in the repository. For the most up-to-date information, please visit the litellm repository.