AI WikiAI Wiki
LM Studio logo

LM Studio

Free tier

Run AI models locally and privately on your own hardware

Free tier available·All audiences·Powered by Multiple (Qwen, Gemma, DeepSeek, GPT-OSS, Apple MLX, etc.)·API available
Visit

Key strengths

Fully local & private — no data leaves your machineSupports a wide range of open-source LLMs (Qwen, Gemma, DeepSeek, MLX, etc.)OpenAI-compatible API for seamless developer integrationHeadless/server deployment via llmstер CLI toolOfficial JavaScript and Python SDKs with MCP client support
Free tier + paid plans
United States
Self-hostable
No ratings yet

Developer Setup & API Usage

SDKs

JavaScript/TypeScript:

npm install @lmstudio/sdk

Full docs: lmstudio-js

Python:

pip install lmstudio

Full docs: lmstudio-python

Headless / Server Deployment (llmster)

Install without a GUI for Linux, cloud VMs, or CI pipelines:

# macOS / Linux
curl -fsSL https://lmstudio.ai/install.sh | bash

# Windows
irm https://lmstudio.ai/install.ps1 | iex

OpenAI-Compatible API

LM Studio serves a local OpenAI-compatible REST API, allowing you to point any OpenAI SDK or HTTP client at http://localhost:1234/v1 and use your local models as drop-in replacements.

Key Integration Points

  • MCP Client — LM Studio acts as an MCP (Model Context Protocol) client for tool-use workflows.
  • Apple MLX — Native support for Apple MLX model format, leveraging Apple Silicon GPU acceleration.
  • CLI (lms) — Script model loading, inference, and server management from the terminal.
  • LM Studio Hub — A model registry and sharing layer for teams (currently in beta).