Voice-to-text in the Mac app you’re already using.
Choose your keys, speak, and release. Dictaty transcribes locally, then inserts the finished text, leaves a clipboard copy, or keeps it in History for recovery.
Dictaty is built for live dictation into the focused Mac app. It does not transcribe meetings or imported audio and video files.
Every completed dictation follows a visible delivery path. It is inserted, copied, or retained for recovery instead of silently disappearing.
Toggle or hold to talk.
Set one shortcut to start and stop recording and another to record only while held. Either shortcut can be disabled. A recording control remains visible while the microphone is active.
Know where the finished text went.
Accessibility permission is optional. It enables global shortcuts and automatic insertion into most editable fields. Clipboard delivery and local History remain available without it.
- Editable field with Accessibility allowed
- Dictaty inserts the text and keeps a clipboard copy.
- Accessibility not allowed
- Dictaty copies the text so you can paste it yourself.
- Target changed or insertion cannot be verified
- The clipboard or local History remains the fallback.
- Secure field
- Dictaty does not insert or copy dictated text into the system pasteboard.
- Transcription failed
- The recording stays in local History for another attempt.
Teach Dictaty the spelling that matters to you.
Add names, product terms, and specialist phrases under Custom words. Dictaty preserves exact matches and carefully corrects similar recognized variants. It avoids replacing a valid known word just because it resembles a custom entry.
Custom words stay on the Mac with the rest of the dictation settings. They are a conservative spelling aid, not a promise that every spoken variant will be rewritten.
A failed transcription is not a lost recording.
Dictaty saves audio locally while you speak. If transcription fails, open History and try again. History is removed automatically after seven days; successful audio can be removed when the next recording starts, when local audio storage reaches its limit, or once it is more than 24 hours old.