Settings guide: language, models and recording preferences

Open Settings from VoicePad’s gear icon or footer. The panel groups the choices that affect recognition, appearance and how you use the result. Controls can differ by platform and build.


Language

Choose the language you will speak before recording. The Help Center documents English and German. VoicePad uses the selected language for recognition; changing the language of a website does not change your app’s speech setting.

Speech model: Whisper

Whisper turns audio into text. Start with Base, then compare another model offered by your device if needed. Smaller models generally use less storage and processing time; larger models need more resources and are not a guarantee of error-free text.

  • Mobile: documented model choices include Tiny, Base and Small.
  • Desktop: larger models can be available with Pro. Do not assume every model listed in an older guide is offered by your current build.

The Help Center’s approximate model download sizes are 75 MB for Tiny, 142 MB for Base and 466 MB for Small. Use the actual download screen for the storage needed by your version. Download and prepare a model before you need offline dictation.

AI model: Ollama


Desktop AI Tools use a separate local language model through Ollama. This is different from the Whisper speech model. Install Ollama and its selected model once while connected; ordinary dictation does not require it.

The documented Qwen options progress from qwen2.5:0.5b to 1.5b and 3b. Try the smaller option first if your computer processes text slowly.


Copying and insertion

  • Auto-Copy: puts the completed transcript on your clipboard.
  • Auto-Inject or Direct: sends text into a supported editable field.

Turn automatic insertion off if you prefer to review the text before placing it in another app.


Microphone and feedback

If your app offers a sensitivity control, change it gradually and test the same short sentence. A threshold control can work differently from a microphone-volume slider.

Sound effects provide recording feedback. Android can also offer vibration through Haptic Feedback. These controls do not improve recognition accuracy.


Other settings

Appearance controls change the theme or supported fonts. Sync connects supported phone–computer pairs. License controls show your current entitlement. Smart Grammar and dictionary options can adjust recognised text; review their output rather than assuming every correction is appropriate.