Skip to main content

LM Studio

Run AI agents with local models using LM Studio. LM Studio provides a GUI for downloading, managing, and serving local models with an OpenAI-compatible API.

Setup​

  1. Download LM Studio from lmstudio.ai
  2. Open LM Studio, download a model from the built-in model browser
  3. Start the local server (Server tab → Start Server)
  4. LM Studio serves on localhost:1234 by default. Configure in droid.yaml:
lm_studio:
  base_url: "http://localhost:1234/v1"

No API key required.

Agent Config​

name: "local-agent"
model:
  provider: "lm_studio"
  name: "local-model"       # Use the model identifier shown in LM Studio
  max_tokens: 2048
  temperature: 0.7

The model.name should match the model loaded in LM Studio. You can check the current model in LM Studio's Server tab.

Notes​

  • LM Studio provides a friendly GUI for model management — useful if you prefer a visual interface over Ollama's CLI.
  • The OpenAI-compatible API means it works seamlessly with Sinaptic® DROID+.
  • Tool use (function calling) support depends on the model you load.
  • LM Studio supports GGUF model format. Most popular open models are available in GGUF.
  • If running Sinaptic® DROID+ in Docker, use host.docker.internal:1234 to connect to LM Studio on the host machine.