Dictation tab

The Dictation tab is where you tune how Ctrl+D behaves. It shows the live hotkey, the source language, the speech-to-text tier, and — if you've subscribed to Keyboard Unlimited — the option to run dictation entirely offline using Whisper. A list of recent dictations sits at the bottom.

1. 1
   Hotkey tile

   Shows your current dictate hotkey (default Ctrl+D) and a live "Ready" / "Listening" indicator that flips red while the microphone is recording.

2. 2
   Source language

   The language Vavus expects to hear when you speak. The list is filtered by the chosen tier — Medical, for example, only supports English, Spanish, German, and French.

3. 3
   Dictation mode — Online / Offline

   Online uses Vavus cloud STT. Offline runs Whisper plus m2m100 fully on-device — your audio never leaves the computer. Offline is gated behind the Keyboard Unlimited plan.

4. 4
   Online tier

   Standard — balanced accuracy and speed for everyday dictation. Pro — higher accuracy on noisy audio and technical vocabulary. Medical — clinical terms, drug names, and dosages.

5. 5
   Speech model (offline)

   Whisper model size to use offline. Choices range from Lightweight (~39 MB, fastest) to Best Quality (~1.55 GB, slowest). Bigger models need more RAM at transcription time.

6. 6
   Translation model (offline)

   The on-device translator (m2m100 family) used when you also turn on offline translation. Sizes and language coverage vary by model.

7. 7
   Download offline models

   Fetches the selected speech model and (if enabled) translation model and saves them to your app data folder. You need to do this once before the offline tier will work.

8. 8
   Recent dictations

   The last ten dictation results, with duration, time, and a one-click copy button.