Connecting llama.cpp to Open WebUI #202

Start llama.cpp's OpenAI-compatible server, then add it as an OpenAI connection in Open WebUI.

1. Start the server (adjust the model path):

   ./llama-server --model /path/to/model.gguf --port 10000

2. In Open WebUI, go to Settings → Admin Settings → Connections → OpenAI → Add Connection.

3. Set the URL to http://127.0.0.1:10000/v1, set Auth to None, and under Advanced choose llama.cpp as the provider.

4. Save the connection. The model served by llama.cpp should then appear in the model selector.

If Open WebUI itself is running in Docker, use http://host.docker.internal:10000/v1 instead of 127.0.0.1. For the Desktop app, 127.0.0.1 is the appropriate local address.

Official guide: https://github.com/open-webui/docs/blob/main/docs/getting-started/quick-start/connect-a-provider/starting-with-llama-cpp.mdx