LiteLLM
used in
0
stacks
Open-source LLM gateway exposing 100+ providers through a single OpenAI-compatible API, as a Python SDK or proxy server.
Translates requests to OpenAI, Anthropic, Gemini, Bedrock, Azure, and roughly 100 other providers into one OpenAI-compatible format. The proxy server adds retries and fallbacks across deployments, plus spend tracking and budget enforcement per project or user. Core is open source with additional features under a commercial enterprise license.
Strengths
- Single OpenAI-compatible interface across 100+ LLM providers
- Built-in retry and fallback logic across model deployments
- Per-key, per-user, and per-project spend tracking and budgets
- Usable as a lightweight Python SDK or a deployable proxy/gateway
Tradeoffs
- Fast-moving release cadence; practitioners report regressions and breaking changes
- Running the proxy adds an extra operational service in the request path
- Some features (e.g. SSO) are gated behind the enterprise license