llmgateway

Why we included this project

Companies running production apps on more than one LLM provider quickly accumulate a mess of credentials, SDKs, and billing spreadsheets. LLM Gateway sits in front of those providers and exposes a single endpoint that speaks the OpenAI API format, so an app already written against OpenAI can be pointed at the gateway and then send traffic to Anthropic, Google Vertex AI, or anything else without touching its code. It also centralizes API key management and rate limiting, and it logs token usage and per-request costs, which makes it much easier to see what you are actually spending. The project can be self-hosted with Docker for teams that want the data and configuration on their own infrastructure, and there is a hosted service for those who would rather not run it. For anyone tired of wiring up each provider SDK separately, this is the kind of infrastructure that quietly saves a lot of engineering time.