v1.0.11.0 is here: six tools, one workspace.See what's new →

Module 05 / 06

Speech Studio

Drop in meeting recordings, interviews or video and get a timestamped transcript on-device, with speakers told apart; you can also see captions as you record, and one more click turns it into a report you can actually read.

Speech Studio: the three-pane transcription workstation, with player and synced transcript

FEATURE 01

Speech to text

Upload audio or video and the on-device engine produces a timestamped transcript; the text lights up as playback moves.

  • Follows alongWhichever line the playhead is on is highlighted; click a line to jump to that moment.
  • Subtitle exportExport SRT and VTT subtitles, or plain text and Markdown.
  • Three modelsLight on resources, standard, or highest quality — switch based on your machine and what you need.
  • Speaker separationTranscription can work out who is speaking automatically, and you can rename or merge speakers; first use needs a separate component download.
product-meeting-0730.m4aSimulated demo · clickable

Click any line to jump to that moment

FEATURE 01+BETA

Live transcription (Beta)

No waiting for the recording to end: see captions as you record, write notes along the way, and get a full transcript when you stop. Before every session you are asked to confirm that everyone taking part has agreed.

Live captions favor speed, so they may contain mistakes or stop now and then; the transcript made after you stop is more accurate. For anything important, keep a separate recording.

Live Transcription: live captions and notes during a recording

FEATURE 02

Insight reports

The transcript goes to a local language model for post-processing, which organizes it into a summary, chapter outline, action items, quotes and a mind map.

  • Grows as it goesThe report doesn't wait for everything to finish; each section appears as it lands, with a skeleton held for the rest.
  • Organized by contentIt first works out whether this is a meeting, an interview or a lecture, then organizes it into sections that fit; for recordings with speakers identified, action items show who owns them.
  • Ask follow-upsThe drawer on the right takes questions about this transcript, and answers trace back to the passages they cite; "Ask about this" on a section asks about just that part.
  • Export itThe whole report exports as HTML or Markdown.
Speech Studio — InsightsGenerating
Key summaryDoneThe next version puts the offline experience first; cloud-related work gets an assessment first, with numbers brought back next week for discussion.
Chapter outlineDone00:04 Scope check
00:18 Assessment and timeline
Action itemsGenerating□ Deliver a feasibility assessment — Speaker B
Quotes
Mind map

Next

Got the machine? Four steps to start

Install, agree, activate, download models — most of the time is just the download.