macOS · Apple Silicon
aarch64 · .dmg
Get .dmgAPK · DMG · EXE · KunoEngine
Local AI chatbot powered by KunoEngine (WebLLM + WASM). Default model profile targets FP8 · ~6GB GPU memory. macOS and Windows ship via Tauri; Android via Capacitor APK.
Detecting your system…
cd kuni && npm install && npm run buildcd desktop && npm install && npm run build → .dmg / .exenpx cap add android && npm run android:sync then open Android Studio → APKRequires WebGPU on the device for KunoEngine inference. First model load downloads ~6GB weights.