Mobile
Android and iOS. The backend runs in-process — nothing extra to install. Run a bundled on-device model, or point it at a provider.
macOS
Self-contained .app — embeds the backend and a Python runtime. Native builds for Apple Silicon and Intel, plus a universal build for both.
Windows
An installer, or a portable zip you unzip and run. Either way the backend and Python runtime live next to the app — keep the folder intact.
Linux
A .deb for Debian/Ubuntu, or a portable AppImage for any modern distro. Self-contained — app, backend, Python and the WPE WebKit webview runtime.
FreeBSD
Terminal UI only — a single self-contained binary that runs the whole workspace in your shell: files, agents, databases, charts and automations. No GUI build yet.
After downloading
See the installation guide for per-OS first-run steps, and the quick start for your first ten minutes. To enable AI features, set up a model in LLM settings — a bundled on-device model, a local Ollama install, or a provider you configure.
All releases and checksums are on GitHub Releases.