Description
LiteLLM is an open-source AI gateway that puts many LLM providers behind one OpenAI-compatible API. It normalizes requests to the OpenAI format across more than 100 providers and gives platform teams virtual keys, spend tracking, budgets, rate limits, fallbacks and routing to the right model for each request. You can see every request, cap spending per key, team or model, and send traffic to a cheaper model when appropriate. It is MIT licensed and can be self-hosted in your own cloud or fully air-gapped. An Enterprise tier adds SSO, role-based access, audit logs, SCIM and dedicated support on the same open-source core. The site also offers a Python SDK for calling many models in the OpenAI format, and a liteagents SDK for routing agent turns across providers. The project is on GitHub, with a Discord community and a quickstart script for macOS, Linux and Windows.…
Preview
Highlights
- One OpenAI-compatible API for 100+ LLMs
- Spend tracking and budgets
- Virtual keys and rate limits
- Fallbacks and model routing
- Self-hosted, including air-gapped
Ratings & Reviews
No ratings yet
Be the first to rate LiteLLM.
People also visit

Inference Gateway
An open-source, cloud-native, high-performance gateway unifying multiple LLM…
LiteLLM FAQ
What is LiteLLM?
An open-source AI gateway and LLM proxy that gives one OpenAI-compatible API for many model providers.
Is it free?
The gateway is MIT licensed and free to self-host. An Enterprise tier adds SSO, RBAC, audit logs and support.
Can I track costs?
Yes. The site says you can track and cap spend per key, team and model.
Can I self-host it?
Yes. It can run in your own cloud or in a fully air-gapped environment.
How do I start?
Use the quickstart script on macOS and Linux, or install the Python SDK.
What are the best LiteLLM alternatives?
Popular alternatives to LiteLLM include LLM Gateway, YoloRouter and Inference Gateway. See all LiteLLM alternatives



