§ Transcripts
Live + final transcripts
A streaming live transcript appears while recording, and a stable final transcript replaces it after stop. Raw and polished display modes. Timestamped rows. Speaker labels backfill from diarization. Export as Markdown, plain text, JSON, SRT, or VTT.
Who this is for
Anyone working in the Meetings workspace — taking notes against a transcript, exporting for distribution, or comparing the live and final view.
Live vs final
The live transcript streams in segments while you record. It is approximate — segments may be re-emitted as the model decides where sentence boundaries fall, and speaker labels are not yet attached.
After you stop, Dettivo runs a final transcription against the retained audio. The final transcript is canonical: stable timestamps, speaker labels filled in, polished output applied. The live transcript stays accessible as a fallback if final never produces (e.g., a corrupted audio buffer).
View modes in the workspace
Raw transcript
Verbatim STT output. Filler stays in. Useful for evidence and quotes.
Polished transcript
Deterministic cleanup, no LLM. The default reading view.
Auto-scroll
Toggle in the transcript header. Off pins the view; on follows the latest row.
Export formats
| Format | Use for |
|---|---|
| Markdown | Notes, vault sync, blog posts, downstream summarization. |
| Plain text | Email, distribution to people who do not want markdown rendering. |
| JSON | Scripts, downstream tooling, custom analysis. |
| SRT | Subtitles, captioning workflows. Speaker labels optional. |
| VTT | Web video captions. Speaker labels optional. |
Launching April 2026
Be first on the launch list.
Dettivo opens for new purchases in April 2026. One-time license, three Macs, lifetime updates.