LiteLLM-RS
A high-performance AI gateway written in Rust that provides OpenAI-compatible APIs with intelligent routing across 100+ providers.
Features
- OpenAI-compatible HTTP APIs (chat, completions, embeddings, images)
- Multi-provider routing, load balancing, and failover
- Streaming, tool/function calling, and response formatting
- Caching, authentication (API keys/JWT), rate limiting, and observability
- Extensible provider system and protocol gateways (MCP, A2A)
Getting Started
- Documentation: see
docs/README.md - Configuration: copy templates from
config/*.exampleorconfig/*.template - Run locally:
make devorcargo run
Library Usage
The crate exposes Rust helpers for OpenAI-style calls. For end-to-end usage, run the gateway binary and configure providers via the config file.
License
MIT. See LICENSE.