LM Studio
Free tierRun AI models locally and privately on your own hardware
Free tier available·All audiences·Powered by Local open models (llama.cpp, MLX) and US-hosted open source models (Kimi K3, GLM 5.3, DeepSeek V4 Flash/Pro, GLM 5.2)·API available
Key strengths
Natively local LLM inference with llama.cpp and MLXZero Data Retention (ZDR) cloud servicesReal-time local voice transcriptionAgentic capabilities for work, coding, and computer controlSupport for frontier open models (DeepSeek, Kimi, GLM)
Free tier + paid plans · from $20 USD/mo
United States
Self-hostable
Facts verified against lmstudio.ai on
Recent changes
- Plans updated: Free plan ($0) includes local LLMs, Bionic Agent, offline voice transcription, LM Link for up to 5 devices, and limited web search. Bionic+ is $20/month with US-hosted open source models and web search. Pro is $100/month with 5× usage limits and early access to new features. Team/organization billing available.
- Models updated: Local open models (llama.cpp, MLX) and US-hosted open source models (Kimi K3, GLM 5.3, DeepSeek V4 Flash/Pro, GLM 5.2)
Get alerts when LM Studio changes
Free. We'll email you when pricing, the free tier, API access or open-source status changes.
No ratings yet
Developer Use Cases
LM Studio is well-suited for developers and technical teams in the following scenarios:
- Local LLM inference for applications: Use the local API to power AI features in your own apps without sending data to external cloud providers
- Privacy-sensitive or air-gapped environments: Run llama.cpp or MLX models fully offline with no external network dependency
- Apple Silicon optimization: Leverage the MLX backend for high-performance inference on M-series Macs
- Agentic coding workflows: Use the Bionic Agent for code generation, debugging, and computer control tasks within a local environment
- Hybrid cloud/local architectures: Route to US-hosted frontier models (DeepSeek V4, Kimi K3, GLM 5.3) with Zero Data Retention when local compute is insufficient
- Multi-device development setups: Use LM Link to share a local model server across up to 5 devices
- Voice-driven developer tooling: Integrate real-time offline voice transcription into local AI pipelines
- Open source evaluation and contribution: Inspect, fork, or contribute to the codebase (5,359 GitHub stars)
Alternatives
See all alternatives