lm

Connect Claude to any OpenAI-compatible LLM endpoint and offload routine work to a local model.

Install

Add this to your `claude_desktop_config.json` file:

{
  "mcpServers": {
    "lm": {
      "command": "npx",
      "args": [
        "@houtini/lm"
      ]
    }
  }
}

Runnable packages

npm@houtini/lm3.3.3stdio

What it can do

No publishable tool enumeration has been recorded.

Community

Evidence

This server has Registry metadata but no public verification run yet.