About Jan
Jan is a free, open-source desktop app for chatting with AI models. It is built by Menlo Research for people who want a ChatGPT-style assistant they control: you can run open models on your own machine or connect to online models from providers you already use. The project is developed in public on GitHub, where it has over 45,000 stars.
Key benefits of Jan include:
- Local models: Run open models such as Llama, Mistral, Qwen, DeepSeek, Gemma and Kimi on your own computer.
- Cloud model support: Plug in online models including OpenAI's ChatGPT, Anthropic's Claude and Google's Gemini from the same app.
- Open source: The code is public on GitHub, so you can inspect it, contribute or build on it.
- Free to download: The desktop app is free, with more than 7 million downloads reported.
- Active community: Discord, GitHub and a HuggingFace presence with 123 published models.
If you are leaving ChatGPT, Gemini or Copilot, Jan gives you one interface that can run models offline on your own hardware, which keeps conversations on your machine, while still letting you reach hosted models when you need them. The trade-offs: local models depend on your computer's resources and may be slower or less capable than the largest hosted models, and using cloud models still means sending data to, and paying, those providers. Some features, like persistent memory across chats, are listed as coming soon rather than available today.

