Jan

Getting Started

1. Download Jan from jan.ai for macOS, Windows, or Linux and install it.
2. Open the model hub within Jan and download a recommended model suited to your hardware capabilities.
3. Start a new conversation and begin chatting with fully private, offline AI — no data leaves your device.
4. Explore the extension system to add features like RAG, API connectivity, and remote model support.

Key Features

• 100% offline operation runs entirely on your local machine with no internet connection required after model download.
• Privacy-first design ensures no conversation data is ever sent to external servers or telemetry services.
• Built-in model hub provides one-click downloads of popular models optimized for various hardware configurations.
• Extension system adds functionality through plugins for RAG, remote models, TensorRT acceleration, and more.
• OpenAI-compatible API serves your local models through a familiar API interface for application integration.
• Cross-platform and open-source available on all major desktop platforms with fully open-source code under AGPLv3.