Open-source, offline-first AI assistant that runs language models locally on your own hardware.
Jan is an open-source desktop application that enables users to run large language models entirely locally on their own computer, with no internet connection or cloud API required. It provides a clean ChatGPT-like interface for local models from Hugging Face and other sources, supports GPU acceleration on NVIDIA and AMD hardware, and includes a built-in model hub for discovering and downloading optimized GGUF models. Jan also supports connecting to remote APIs like OpenAI and Anthropic as a unified frontend. Privacy-conscious users, developers, and organizations with sensitive data use Jan to enjoy AI capabilities without sending data to third-party servers. The application runs on macOS, Windows, and Linux, and its modular extension system allows community-contributed features.