1. Pick a Model
On first launch, Voicetypr asks you to download a speech model.
Recommended for beginners:
- Whisper Base (~140 MB) — Fast, accurate for clear speech, 99+ languages
- Whisper Large v3 (~2.9 GB) — Best accuracy, handles noise and accents better
The model downloads once and stays on your machine. No internet needed after that.
2. Set Your Hotkey
Default hotkey is ⌘ + Shift + Space (macOS) or Ctrl + Shift + Space (Windows).
To change it:
- Open Voicetypr settings
- Click Hotkey
- Press your preferred key combination
- Click Save
3. Your First Dictation
- Open any app — Notes, ChatGPT, Gmail, VS Code
- Place your cursor where you want text
- Press and hold your hotkey
- Speak naturally
- Release the hotkey — text appears where your cursor was
Tips
- Pause punctuation: Say "comma", "period", "new line" — Voicetypr inserts them
- Numbers: Say "one twenty three" or "one hundred and fifty"
- Corrections: Backspace and re-dictate. There's no cloud delay.
What's Next?