LiteLLM is an open-source Python SDK and proxy server providing a single OpenAI-compatible interface for over 100 language models β including Claude, GPT-4, Llama, Mistral, and Gemini β with unified cost tracking, rate limiting, and provider fallback logic. It is the leading model-routing layer for AI applications that need provider portability without SDK rewrites.
Call 100+ LLMs with a unified OpenAI-compatible API. Switch providers, manage rate limits, and track costs across every model from one library.
LiteLLM is an open-source Python library and proxy server that provides a unified OpenAI-compatible API for 100+ language models β including Claude, Llama, Gemini, Mistral, and Groq. It handles model routing, cost tracking, rate limiting, and fallback logic so your application does not need to change when you switch providers. Used by teams building provider-agnostic LLM applications that need to hedge against pricing changes or model deprecations.
Also see: Anthropic API Β· OpenRouter
Also see: Anthropic API Β· Claude
LiteLLM supports 100+ providers including OpenAI, Anthropic (Claude), Google Gemini, Cohere, Azure OpenAI, Mistral, Groq, Ollama, and any OpenAI-compatible endpoint. You switch providers by changing the model string β no SDK rewrites required. It also supports local models via Ollama for fully offline inference.
Also see: Haystack Β· Instructor
0β100 viral momentum index combining social buzz, search trends & growth velocity
A.R.C. ratings are calculated for developer infrastructure and API-first tools. This tool hasn't been evaluated yet or falls outside the A.R.C. scope.
Lower = more portable. 0 = fully open, 100 = maximum lock-in.
GitHub health score, founder track record, full A.R.C. breakdown, category peer comparison, and 14-day score forecast β in one printable report.