Definition
Jan is an open-source desktop application that runs large language models locally on a user's own computer, positioned as a privacy-focused alternative to hosted chat assistants. Released under the Apache 2.0 license, it runs on Windows, macOS, and Linux and is built so that, by default, data stays on the device and internet access is optional. It targets users who want a polished chat experience without sending their conversations to a third party, which aligns closely with a self-sovereign approach to computing.
How it works
Jan lets a user download and run open-weight models such as Llama, Gemma, and Qwen from model hubs, then chat with them through a clean desktop interface. Under the hood it manages the local model runtime so the user does not have to assemble a serving stack manually. It supports creating custom assistants and integrates the Model Context Protocol (MCP) for connecting tools, while keeping everything resident on the local machine.
Local API and optional cloud
Jan also runs a local OpenAI-compatible API server, so other applications on the same machine can use the locally hosted model as if it were a remote service. While its core promise is offline operation, it can optionally connect to external providers when a user chooses, leaving the privacy trade-off in the user's hands rather than enforcing a default of remote processing.
Jan is a packaged consumer front end rather than a low-level engine; for the underlying techniques that make local models practical, see model quantization and the GGUF format.
Compare local-AI apps in the sovereign self-hosting catalog.
In Simple Terms
Jan is an open-source desktop application that runs large language models locally on a user’s own computer, positioned as a privacy-focused alternative to hosted chat…