LMStudio

Use LMStudio as a local LLM inference server.

Setup

import { LMStudioLLMProvider } from '@llmrtc/llmrtc-provider-lmstudio';

const llm = new LMStudioLLMProvider({
  baseUrl: 'http://localhost:1234/v1',
  model: 'llama-3.2-3b'
});

Env vars

• LMSTUDIO_BASE_URL
• Optional: LMSTUDIO_MODEL

Notes

• Great for prototyping and edge use cases; ensure the model is loaded in LMStudio before starting the backend.