Artificial Intelligence
LiteLLM: A Proxy to Unify Model Providers
The LiteLLM proxy sits between your applications and your model providers, exposing a single endpoint in the OpenAI format that routes each call to OpenAI, Anthropic, Cohere, or a self-hosted model. It earns its place at the second provider, where clean code breaks down: every SDK brings its own client, message format, streaming semantics, errors, and function calling.