core

@van1s1mys/ai-router

Semantic search routing for SPAs — find the best route by meaning, not keywords.

Documentation | Live Demo | npm | GitHub

Runs a HuggingFace embedding model inside a Web Worker and uses Orama hybrid (text + vector) search to match user queries by semantic similarity.

Install

npm install @van1s1mys/ai-router

Usage

import { SmartRouter } from '@van1s1mys/ai-router';

const router = new SmartRouter({
  routes: [
    { path: '/pricing', title: 'Pricing', description: 'cost, plans, subscription' },
    { path: '/contact', title: 'Contact', description: 'support, phone, address' },
    { path: '/docs',    title: 'Docs',    description: 'documentation, API, guides' },
  ],
  threshold: 0.5,
});

await router.ready; // model loads (~22 MB, cached after first run)

const result = await router.search('how to reach support?');
// { path: '/contact', score: 0.91 }

router.destroy(); // cleanup when done

Progressive model loading

Start with a fast model, upgrade to a better one in the background:

const router = new SmartRouter({
  routes,
  model: ['Xenova/all-MiniLM-L6-v2', 'Xenova/multilingual-e5-small'],
  onModelUpgrade: (modelId) => console.log(`Upgraded to ${modelId}`),
});

await router.ready; // first model ready — search works immediately

Instance caching & preloading

// Pre-warm at page load
SmartRouter.preload({ routes, model: ['Xenova/all-MiniLM-L6-v2', 'Xenova/multilingual-e5-small'] });

// Later — returns cached instance, no re-download
const router = SmartRouter.create({ routes, model: ['Xenova/all-MiniLM-L6-v2', 'Xenova/multilingual-e5-small'] });
await router.ready; // instant if preload finished

Multilingual support

The default model (Xenova/all-MiniLM-L6-v2) works best for English. For other languages, use the multilingual model:

const router = new SmartRouter({
  routes,
  model: 'Xenova/multilingual-e5-small',
});

API

new SmartRouter(options)

Option	Type	Default	Description
routes	RouteConfig[]	required	Routes to index
model	string | string[]	"Xenova/all-MiniLM-L6-v2"	Model ID or ordered array for progressive loading
threshold	number	0.5	Minimum similarity score (0-1)
onModelUpgrade	(modelId: string) => void	—	Called when the router switches to the next model

SmartRouter.create(options): SmartRouter

Returns a cached instance for the given model config. Safe to call on every component mount.

SmartRouter.preload(options): SmartRouter

Same as create(), but intended to be called at page load to pre-warm the model.

router.ready: Promise<void>

Resolves when the first model is loaded and routes are indexed. Resolves immediately during SSR.

router.search(query): Promise<SearchResult | null>

Returns { path, score } or null if no route meets the threshold. Returns null during SSR.

router.destroy()

Terminates the worker, cleans up resources, and removes from cache. Safe to call multiple times.

SSR

SSR-safe out of the box. On the server ready resolves immediately and search() returns null. The worker is only spawned in the browser.

Framework plugins

Auto-scan your pages directory instead of listing routes manually:

• @van1s1mys/ai-router-plugin-vite
• @van1s1mys/ai-router-plugin-next
• @van1s1mys/ai-router-plugin-webpack

License

MIT © IvanMalkS