Skip to main content
All documentation

Offline voice dictation

Download a Whisper model, record from a shortcut, tune local text cleanup, and use Dictionary and Snippets.

How dictation works

Side Whisper records microphone audio locally, converts it for the selected Whisper model, transcribes on the device, then copies or pastes the final text according to your settings.

  • Choose Tiny English, Tiny multilingual, Small, Large Turbo, or Large.
  • Use the default Alt+Q shortcut or record a different shortcut.
  • Model downloads continue even if Settings closes.
  • A native Side Whisper notification confirms completion.

Dictionary and Snippets

  • Dictionary entries correct names, acronyms, and recurring recognition mistakes.
  • Snippets expand a short dictated trigger into a longer local text block.
  • Both collections can be imported and exported independently.

Privacy boundary

Voice recordings and transcriptions are not uploaded to the Side Whisper account backend. Model downloads require the internet, but transcription itself is local after the model is installed.