CPU: any modern x86-64 processor (recognition runs on CPU by default).
GPU (optional): NVIDIA GPU for one-click CUDA acceleration, with automatic CPU fallback.
Disk: space for the app plus your chosen Whisper model(s).
Installing
1. Run the installer
Download the 64-bit installer above and run it. QuillVoice installs like any Windows app — no account or sign-in required.
2. Pick a model
On first launch, choose a Whisper model (tiny through large-v3-turbo). Smaller models are faster; larger models are more accurate. Everything runs on your device.
3. Press your hotkey and speak
Press your hotkey in any text field and start talking — QuillVoice types real keystrokes into the focused window. See the full walkthrough on the how it works section.
Activating Pro
Bought Pro? Paste your license key into Settings → License. It is verified locally — no sign-in, no phone-home — consistent with the no-account promise.
Updates
QuillVoice can check for new versions on launch. That update check is the only outbound request it makes — it sends no audio, no transcripts, and no personal data, and it never touches the dictation path. It can be turned off.