Jan
Offline-first, privacy-focused desktop app to run LLMs locally on any hardware
Overview
Jan is an open-source, cross-platform desktop application that lets users download and run LLMs entirely offline. It bundles llama.cpp as its inference engine and provides a clean chat UI along with an OpenAI-compatible local API server. Models are downloaded from the built-in Hub or imported manually from Hugging Face.
Where it falls short of ChatGPT
- Desktop-only; no headless server deployment mode
- Multi-user collaboration not supported
- Limited to llama.cpp-compatible model formats
We list the gaps honestly so you can decide if the trade-off is worth owning your data.
Tags
Claim this listing to keep it accurate, add a deploy template, or feature it on relevant pages.
Embed the Jan difficulty badge in your README — it links back here.
[](https://openreplace.com/jan)Similar open-source projects
Other self-hostable tools in the same space worth comparing.
Run large language models locally with a simple CLI and REST API
The most widely used web interface for running Stable Diffusion image generation locally
Feature-rich self-hosted chat UI for Ollama and OpenAI-compatible APIs
Modern AI chat framework with multi-provider support and MCP marketplace