Docs
How to use lmgw from your own code, and the reference of its HTTP API.
Build on lmgw
lmgw speaks protocols your code already knows. Point an OpenAI or Anthropic client at it, read what each model can do from /v1/models, and add voice and tools when you need them.
Read →API reference
The complete HTTP API of lmgw: the OpenAI-, Anthropic- and llama.cpp-compatible routes, lmgw's own extensions and MCP, the device Chat API, the dashboard API with Knowledge, labs and ops, and the agent run API, generated from the gateway's OpenAPI document.
Read →API compatibility
What lmgw's OpenAI and Anthropic routes do against the official SDKs, model by model: a matrix from a live check of the running gateway, with every failure and its reason.
Read →