janhq/jan — reverse-engineered prompt

Reverse engineered prompt

GitHub

Build me a desktop app like ChatGPT that I can run on my own computer, with a clean chat interface and strong privacy. I want people to be able to download local AI models like Llama, Gemma, Qwen, or GPT oss from Hugging Face, run chats offline when they want, and see basic model status like download progress, loaded model, and whether it is using the computer locally.

Also let users connect cloud providers like OpenAI, Anthropic, Mistral, Groq, and others if they add their own keys. Include a simple way to create custom assistants for different tasks, save chat history, manage models, and switch between local and cloud models without confusion.

Please include a local API server on localhost 1337 that other apps can call in an OpenAI compatible way, plus basic MCP support for agent tools. Make it work as a polished desktop app for macOS, Windows, and Linux. Look up current docs online if needed.

Want more depth? Deep Reverse