Peter Steinberger · 2026-04-08 · major
OpenClaw — self-hosted personal AI assistant
MIT-licensed personal AI assistant that runs on your own devices and bridges any LLM to 20+ messaging channels — now past 355k GitHub stars.
An MIT-licensed personal AI assistant that runs on your own devices and already has 355k stars — tagline: 'the AI that actually does things.'
Key specs
| License | MIT |
|---|---|
| GitHub stars | 355k |
| Channels | 24+ |
What is it?
OpenClaw is an open-source personal AI assistant built by Peter Steinberger. You run it on your own Mac, Windows or Linux machine. It plugs local or remote LLMs (Claude, GPT, Gemini, DeepSeek, Ollama, and more) into a single gateway that exposes them through every chat app you already use: WhatsApp, Telegram, Slack, Discord, Signal, iMessage, Teams, Matrix, and over a dozen others.
How does it work?
A local-first gateway daemon manages sessions, channels, tools and memory. You point it at one or more model providers and connect the channel adapters you want; the same assistant is reachable from every channel and shares context across them. It can drive a browser, run shell commands, read and write files, and extend itself through custom skills.
Why does it matter?
It is the first self-hosted personal AI assistant polished enough that non-developers are installing it. The repo jumped from 9k to 60k stars in days and is now over 355k — one of the fastest star-growth curves in GitHub history. If you wanted a personal assistant without shipping every message to a vendor, you finally have one.
Who is it for?
Privacy-conscious users, indie developers, self-hosters.
Try it
npm install -g openclaw@latest && openclaw onboard --install-daemon