Automatic Speaker Recognition in MacWhisper

MacWhisper 12 now supports Automatic Speaker Recognition! If you're using a WhisperKit Model it will automatically group your transcript by speaker.

Starting with MacWhisper 12.0, you can now automatically recognize speakers in your transcript. This currently only works for local models that run on the WhisperKit engine, or on cloud transcription providers ElevenLabs and Deepgram.

Select a compatible model

In order to use automatic speaker recognition you will need to use a supported model. The local models that currently support all use the WhisperKit engine, and can be recognised in the list with their green Speaker Recognition badge.

Renaming a speaker

After the transcription is completed MacWhisper will automatically assign default speaker names to each of the detected speakers (Speaker 1, Speaker 2 etc). You can easily rename a speaker from the sidebar. Just click on the speaker name and change it. When you unfocus the textfield or you hit return, the new name will be applied.

Merging Two Speakers together

Sometimes the app may assign different parts of the transcript said by the same person, to two different speakers. You can easily merge two speakers together by right clicking on the speaker name in the transcript or in the sidebar, and then click Merge Speaker To..

Autofill speaker names from existing speakers

If you have a list of common speaker names that you use across different transcripts, you can easily auto-fill the speaker name to one of the existing speakers. While you type the name of the speaker, autosuggestions based on the speakers you've created in settings will appear. Hit ⌘ + Return to select it.

Changing the color of a speaker.

You can easily change the color of a speaker within the transcript by clicking on the little color palette icon in the sidebar.

Quickly adjust segments with the keyboard shortcuts

You can use the keyboard shortcuts that belong to each speaker (1, 2, 3 etc) to quickly adjust segments to another speaker. Just hover over the segments that you want to change to a different speaker and hit the keyboard shortcut.