Hold the hotkey and talk. Your words are typed at your cursor, in any app. Nothing leaves your Mac, and if you weren't in a text field, nothing you said is lost.
Download for macOSBoth permission prompts now advance on their own the moment a grant lands, so nobody has to hunt for a Continue button.
the onboarding screen feels like it's doing too much
It asks for three permissions, explains the hotkey, and runs the try-it in one view. Splitting it into steps would let each one hold a single idea. Want me to draft that?
At the smallest window width the function blurb wraps under its action button and the rows lose their alignment. Everything reads fine at the default size.
Your voice is transcribed on your Mac and never sent anywhere. Yorick reads whether a text field is focused, nothing else on screen. MIT licensed.
Text usually lands as you let go of the key. It uses Apple's on-device speech model, with Whisper available as an alternative engine.
Ums and false starts come out before the text is typed. Costs about half a second and needs an Apple Intelligence Mac. A miss types your words as spoken.
Talk with no text field focused and your words are saved instead of dropped. Copy them out to wherever you meant them to go.
Built by someone who talks to his Mac all day. Free, open source, and yours to audit. Tell me what breaks and I'll fix it: open an issue or email yorick@rebman.dev.
Requires macOS 14 or later.