### Feature proposal

**Context:** LiteLLM abstracts 100+ LLM providers behind a unified interface. Users routing between models need accurate, up-to-date pricing to make cost-optimal decisions.

**Proposal:** [WhichModel](https://whichmodel.dev) is a free MCP server that tracks pricing and capabilities across 100+ models, refreshed every 4 hours. It could complement LiteLLM's routing by providing real-time pricing data.

**Potential integration points:** 1. Use WhichModel pricing data to power LiteLLM's cost-based routing 2. Add WhichModel as an MCP tool example in LiteLLM docs 3. Use `check_price_changes` to keep LiteLLM's internal pricing tables current

**MCP Config:** ```json {"mcpServers":{"whichmodel":{"url":"https://whichmodel.dev/mcp"}}} ```

**4 tools:** - `recommend_model` — cost-optimized suggestions by task type, complexity, budget - `compare_models` — head-to-head comparison with volume cost projections - `get_pricing` — raw pricing data with filters (model, provider, capabilities, price ceiling) - `check_price_changes` — pricing change tracker since a given date

No API key needed for free tier (1,000 req/mo). MIT licensed.

- GitHub: https://github.com/Which-Model/w…