§ Speakers
Speaker labels + diarization
After a meeting stops, on-device diarization assigns speaker labels to transcript rows. Rename a speaker once and Dettivo saves the suggestion for future meetings. A talk-time timeline shows percentages and a strip view of who spoke when.
Who this is for
Anyone with multi-participant meetings — interviews, team syncs, panel discussions. Especially useful when distributing transcripts where attribution matters.
How diarization works
Diarization runs after the final transcript is produced. The model identifies distinct voices in the audio and tags rows with anonymous speaker IDs ("Speaker 1", "Speaker 2"). Diarization is on-device — no audio leaves your Mac. The diarization model is downloaded on first use from Settings › Models.
Working with speakers
Rename a speaker
Click a speaker name in the transcript. Type the real name. Apply.
Saved suggestions
Dettivo remembers names you have applied before and offers them as suggestions next time the same voice appears.
Talk-time timeline
Strip view at the top of the transcript shows who spoke when. Percentages on hover.
Name detection
Dettivo scans transcript context for likely names ("thanks Maria, what do you think") and suggests them.
Re-run diarization
Meetings workspace → More menu → Re-run Diarization. Useful when you have added new saved voices.
Launching April 2026
Be first on the launch list.
Dettivo opens for new purchases in April 2026. One-time license, three Macs, lifetime updates.