up to 30 minutes on the free plan, no minute limit on Pro
For UX research
Transcribe a user interview
Transcribe a user interview for UX research: timecodes on every line, speakers named on request, quotes you can point at in the recording.
3 files a day up to 30 minutes, free99+ languagesSpeaker separation and timestamps
In short
Upload the recording of the interview and the transcript comes back with the time of every line and the voices named on request, so every quote in the report can point at the moment it came from. One file is up to 30 minutes; the free plan is $0.
The case
Why a quote needs the moment behind it
The sentence in the report and the sentence in the recording are not always the same, and that difference is where a research finding goes wrong. A quote without a timecode cannot be checked by anyone but the person who was in the room. A quote with one can be played again, argued with, and thrown out.
What a research interview gives you
0.84 s to transcribe a short file
one row per segment, with the time and the speaker
How to transcribe a research interview
- 1
Record the session the way the team agreed
Keep the file of the session as it is: the recording of one device is enough, and it does not have to be converted before the upload.
- 2
Upload it and separate moderator and participant
Choose the recording in the widget above, leave the language on automatic detection, and switch on the split by speaker, so the questions of the moderator are marked apart from the answers.
- 3
Pull the quotes into a sheet
Export CSV: one row per segment, with the time and the speaker. Tag the rows in the spreadsheet and every quote keeps the place it came from.
- 4
Check the moments that carry the finding
Open the recording at the timecode of the three quotes that carry the finding and listen to them. That is where the meaning of a study is decided.
Tips
Read the transcript before the report is written, not after.
Give the moderator and the participant two different microphones.
Keep the quotes that contradict the hypothesis.
Export CSV, and tag the rows while the session is fresh.
Questions this page answers
Can the moderator and the participant be told apart?
Yes. Switch on the split by speaker and the transcript names the voices separately on request, so the question and the answer are marked apart. Without it the questions of the moderator are not marked, and the coding gets slower.
How do I keep the quotes with their time?
Export CSV. Every row carries the start and the end time of the segment and the speaker, so a tagged row in the spreadsheet keeps the place of the quote in the recording.
How accurate is the transcript?
We have not measured the accuracy of the transcript, so we publish no percentage. Check the first minute against the recording yourself: that is where a wrong word costs the most.
How fast is a short recording?
On a short file, our own measurement of 21 October 2026 gives 0.84 s to transcribe and 0.39 s to detect the language. A full interview takes longer than that.
Reading: 6 minutesUpdated Transcribsi editorial team
Drop a recording — the text arrives in a minute
Speakers by name, a time on every line, export to TXT, SRT, VTT, CSV and Markdown.
3 files a day up to 30 minutes, free99+ languagesSpeaker separation and timestamps