LiteLLM Overview

LiteLLM Overview

Summary

LiteLLM is an open-source tool and proxy server that enables developers to call over 100 large language models (LLMs) from different providers using a unified OpenAI-compatible API. It supports both a Python SDK and an AI Gateway (proxy server) for flexible integration and deployment.

Content

What is LiteLLM?

LiteLLM allows users to interact with a wide range of LLMs (OpenAI, Bedrock, Azure, VertexAI, Anthropic, Groq, and more) through a single interface. It provides endpoints for chat, completions, embeddings, images, audio, and more, and supports agent invocation and advanced routing.

Key Features

  • Unified API for 100+ LLMs
  • Python SDK and AI Gateway (proxy server)
  • Support for multiple endpoints and providers
  • Agent invocation and A2A protocol support
  • Easy deployment via Docker, Render, or Railway

Use Cases

  • Developers building multi-provider LLM applications
  • Organizations seeking to simplify LLM integration and management
  • Teams needing a flexible, open-source AI gateway

References