Install on Linux and macOS
Before you start
Section titled “Before you start”- Linux x64/arm64 or macOS Apple Silicon/Intel.
- About 200 MB of free space after unzipping.
- An AI provider: Ollama locally or an OpenAI-compatible endpoint.
You do not need Node.js, an installer or an account.
- Download the Linux package and unzip it somewhere permanent.
- Open a terminal in the folder and make the launcher executable:
chmod +x start.sh./start.sh- The browser should open at
http://localhost:3001.
If the browser does not open, visit the URL manually. Keep the terminal open while the app is running.
- Download the macOS package for Apple Silicon or Intel.
- Unzip it and move the folder to a place you will remember.
- Open Terminal in the folder and run:
./start.sh- The browser should open at
http://localhost:3001.
After the first launch
Section titled “After the first launch”Open System → Providers, add Ollama or an OpenAI-compatible endpoint, test the connection and choose a model.
Next: connect a provider or first steps.
