Speaker diarization

Speaker diarization is identifying who spoke when in a recording, so a transcript shows each person's lines separately.

In practice: The tool splits the audio by voice and labels each segment, for example Speaker 1 and Speaker 2, or real names if it can match them to meeting attendees. It is what turns a wall of text into a readable meeting transcript.

Why it matters when picking a tool

If you transcribe interviews, calls or meetings, diarization quality matters as much as word accuracy. Test with people who have similar voices or talk over each other, since that is where tools mix speakers up.