______ __ _ __ __
/ ____/____/ /_ ____ / | / /___ ____/ /__
/ __/ / ___/ __ \/ __ \ / |/ / __ \/ __ / _ \
/ /___/ /__/ / / / /_/ / / /| / /_/ / /_/ / __/
/_____/\___/_/ /_/\____/ /_/ |_/\____/\__,_/\___/
Hands-free local voice assistant with wake-word activation, local STT/TTS, and flexible backend routing. Optimized for Surface Laptop Studio 2 / RTX 4050-class hardware.
- Wake word: OpenWakeWord
hey_rhasspy(ONNX) - STT: Parakeet TDT v2 through
onnx-asr - VAD: Silero VAD for silence detection and barge-in
- TTS: Kokoro ONNX, with espeak-ng fallback
- Backend: Ollama, OpenAI-compatible servers, Hermes, or Odysseus
cd v2
./wizardOr manually:
cd v2
./setup.sh
./test.sh
./run.sh| Platform | Command |
|---|---|
| Fedora 43 / GNOME / PipeWire | cd v2 && ./install-fedora |
| WSL2 Ubuntu with WSLg audio | cd v2 && ./install-wsl2 |
| Native Windows 11 | cd v2 && .\install-windows.ps1 |
| macOS ARM | cd v2 && ./install-macos-arm |
| Platform | Command | Default Binding |
|---|---|---|
| GNOME/Fedora | ./install-hotkey-fedora |
Ctrl+Alt+V |
| Windows | .\install-hotkey-windows.ps1 |
Ctrl+Alt+V |
See v2/docs/configuration.md for backend URL, model name, wake-word mode, TTS voice, VAD settings, barge-in, and hotkeys.
See v2/docs/avatar.md for the optional animated avatar (PyQt6 sidecar with Rhubarb lip-sync).
v2/ # Current best build (active development)
echo_node/ # Core package (agent profiles, pipeline stubs)
avatar/ # Animated avatar subsystem
tools/ # Voice audition, wake-word recording, deploy wizard
docs/ # Configuration and avatar docs
assistant_v2.py # Main entry point
specs/ # Original design specs
docs/ # Setup guides, provider docs, troubleshooting
sprites/ # Avatar sprite sheets (raw source images)
archive/ # Historical v1 code, transcripts, design docs
See DEPLOYMENT.md.
MIT. See LICENSE.