Jan

Run LLMs offline on your computer

41.8k|stable|v0.7.9

Download for Windows

Download .zip — extract and run

How to install

1Download the file above

2Extract the .zip file

3Run jan.exe

Local LLM execution — Download and run models like Llama, Gemma, and Qwen entirely offline.

Cloud model integration — Connect to OpenAI, Anthropic, Mistral, and other providers when needed.

OpenAI-compatible API — Local server at localhost:1337 for integration with other applications.

Custom assistants — Create specialized AI assistants for specific tasks.

Model Context Protocol support — MCP integration enables agentic capabilities.

Jan is a desktop application that lets you run large language models locally without sending data to the cloud. It supports both local models from HuggingFace and cloud providers, with an OpenAI-compatible API for third-party integrations. Full privacy control and offline operation make it ideal for users who want to keep their AI interactions local.

macOS 13.6+ (8GB-32GB RAM depending on model size), Windows 10+, or Linux with optional GPU support

Source on GitHub

pavilion