This comes straight from OvertChat’s own GitHub page, and it’s free. Need help installing?
About this app
OvertChat is a phone app for talking to AI models that you host yourself. Your conversations, files and projects live on your server rather than in someone else's account, and you decide which models to point it at.
The server side works with local inference engines like vLLM, llama.cpp and SGLang, and with hosted providers using your own API keys. It can also connect to coding agents running on your machine or over SSH, so you can start or resume a session, follow the plan and tool calls as they happen, and review changes without sitting in a terminal.
The Android app is a thin client, so it is only half the story: you have to install and run the OvertChat server first, which is a Docker setup on a computer you control. With no server to connect to, the app has nothing to talk to.
From the app’s home page
- See the codeHow it’s made, every line is public
- WebsiteThe app’s own home page
- Questions & commentsTalk with the people who make it
- Report a problemTell the makers something is broken
- All versionsOlder downloads and what changed
- LicenseThe rules for using and sharing it
More like this
Ollama App
A friendly phone remote for your Ollama server
Reins
A calm, simple front end for Ollama chats
RikkaHub
One chat app for many different AI assistants
Kelivo
A tidy chat app for the AI model of your choice
Cherry Studio
Chat with AI models using your own API keys
Maid
Chat with AI models running right on your phone