How can we help?

Download the installer for your platform from the Download page. On Linux, download the single portable binary, make it executable and run. On Windows, run the .exe installer. On first launch, the app will download your selected model tier (1–3 GB one-time download).

Click into any text field (editor, chat, terminal, etc.), press Ctrl+D (configurable), and speak. When you stop, the transcribed text is typed into the active window. For longer dictation, hold the hotkey — the app streams text as you speak.

Yes. After the initial model download and license activation, Voice Pro works fully offline. No internet connection is needed for transcription.

Try these steps in order:

1. Switch to a higher model tier (Settings → Model). Fast Balanced or Max Balanced offer the best quality-to-size ratio.
2. Enable noise reduction if you are in a noisy environment.
3. Add hotwords for domain-specific terms (Settings → Hotwords).
4. Lock the language instead of using auto-detect (Settings → Language).

Personal voice training (fine-tuning) is on the roadmap for Q3 2026 — see the roadmap post. Once it ships, training on your own voice will become the top fix for recognition issues.

Any microphone works. For best results, use a headset or USB condenser microphone. Built-in laptop microphones work but may pick up more background noise. The built-in noise reduction helps compensate.

Any modern GPU with Vulkan or CUDA support. NVIDIA: GTX 10-series or newer. AMD: RX 500-series or newer. Intel Arc GPUs and recent integrated graphics also work. You can also run in CPU-only mode — transcription will be slower but fully functional.

After purchase, log into the app with your account email and password. Your license is automatically activated. You can activate on up to 2 devices (Personal) or 1 device per seat (Business).

Yes. Deactivate the license on your old device from Settings → License, then log in on your new device. Alternatively, log into your account to manage devices.

After 14 days, the app stops transcribing. You can purchase a license at any time to continue. Your settings and any trained models are preserved.

Go to Settings → Audio and select your microphone from the dropdown. On Linux, make sure your audio server (PipeWire or PulseAudio) is running. On Wayland, you may need to grant the app microphone permissions through your desktop environment's settings.

On Linux with Wayland, the app needs text injection permissions. Check that your compositor supports the virtual keyboard protocol (most modern compositors do). On X11, the app uses XTest which works universally. On Windows, some applications with elevated privileges may block input — try running Voice Pro as administrator.

Check your internet connection and firewall settings. The download server is on Cloudflare R2. If downloads fail behind a corporate proxy, try downloading from home or ask your IT team to whitelist the download URL. You can also manually download models from your account page.