System requirements
- macOS 14 or later
- Universal app for Apple silicon and Intel
Help when you need it
Questions, problems or feedback—we’re here to help.
Replies usually within
two business days.
Transcription is CPU and GPU bound and runs entirely on your device, so a long recording takes a while. A two-to-three-minute clip is enough to see speaker separation working.
A recording with no speech is reported as such. Check that the selected file contains audible voices.
Very short or overlapping speech is harder to attribute reliably. MinuteMark may omit uncertain stretches rather than assign the wrong speaker.