Long audio in.
Editable, speaker-labeled transcript out.

Drop a meeting recording, an interview, a podcast — up to five hours. Whisper does the words, pyannote splits the speakers, you do the editing.