Transcribing files

Back to docs
The short version: Open Transcribe audio/video file… from the menu, drop a file in, get text. Same Whisper model, same on-device processing, same lack of cloud involvement.

Opening the dialog

Two ways:

  • Click the menu bar icon and choose Transcribe audio/video file…
  • Set up a global keyboard shortcut in Settings → General and trigger it from anywhere. Useful when you transcribe files often and don’t want to fish through the menu every time.

Adding a file

Drop a file onto the dialog, or click the dashed box to browse. TongueType handles audio and video, including most things you’ll throw at it:

  • Audio: WAV, MP3, M4A, AIFF, CAF, FLAC.
  • Video: MOV, MP4, M4V, 3GP. (TongueType ignores the video and transcribes the audio track.)

While it’s working

You’ll see a spinner and a rotating status line (“Decoding syllables…”, “Untangling sounds…”, “Squinting at waveforms…”), just to keep things interesting. Closing the dialog cancels the transcription, so feel free to bail if you picked the wrong file.

Getting your text out

Once the transcript appears, you have three buttons:

  • Copy & Close: copies the text and closes the dialog (also bound to Return).
  • Copy: copies without closing, in case you want to re-read or run another file.
  • Clear: resets the transcript area so you can drop in another file.

Language

File transcription uses the same language setting as voice dictation. Pick it from the menu bar under Language, or leave it on Auto-detect if you don’t know what’s in the file.

Postprocessing

Your replacement rules and Whisper-annotation filtering apply to file transcripts too, so a recording with “[music]” or “new line” in it gets the same treatment as live dictation. See Postprocessing for the details.

The 10-second limit

On the free version, files are trimmed to the first 10 seconds. The dialog tells you up front, and the transcript ends with “[Audio trimmed to 10 seconds]” as a reminder.

TongueType Pro removes the limit entirely. Transcribe a 90-second voice memo, a 20-minute interview, or that two-hour meeting nobody should have sat through.

Where to next