Expressive voice
Kokoro delivers a more human read than system text-to-speech, with local inference on Apple silicon.
Local text-to-speech for macOS
Sotto turns text from any app into expressive speech on your Mac. No account, no subscription, and no text sent to a cloud voice API.
Keep Sotto in the menu bar. Highlight a passage in a browser, editor, or chat app, then use your own global shortcut to hear it.
Kokoro delivers a more human read than system text-to-speech, with local inference on Apple silicon.
Sotto warms the model after startup and begins long passages with a smaller first section for faster audio.
Selected text and generated audio stay on your Mac. The network is only used for the first model download.
Change speed during playback, skip between sections, stop instantly, add pronunciation rules for names and acronyms, and let Sotto pause for voice apps.
Select anything readable in the app you are using.
Sotto reads the active selection without interrupting your workflow.
Use the menu bar for pause, speed, navigation, and Sotto Studio.
Preview release
Apple silicon required. This initial Homebrew build is unsigned, so macOS may require a one-time Gatekeeper override on first launch.