Try the speaker filter and switch export formats below. This sample is an editorial illustration of a transcript, with no uploaded recording.
What changes when we start listening to a city instead of looking at it?
You notice the small things. The tram at six. A door closing. Someone making coffee.
Almost like the city has its own rhythm.
One transcript. Your format.
1 00:00:12,000 --> 00:00:18,800 What changes when we start listening to a city instead of looking at it?
From the first upload
to the final edit.
Automatic transcripts are a starting point. Review names, specialist vocabulary and passages with overlapping speech before publishing.
Bring your recording.
WAV, MP3, FLAC, M4A, MP4, MOV or WebM. Maximum 2 GB and 4 hours per recording. Video is transcribed from its audio track.
Make it sound like you.
Name the speakers, review low-confidence words and add terminology. Keep timestamps attached as you edit the text.
Send it downstream.
Export plain text for notes, JSON for your application, or SRT and VTT for captions. Original media retention follows your plan.