99+ languages — set by you or detected automatically
For translators
Transcribe an audio for a translator
Transcribe an audio for a translator: the source text with timecodes, in any of the any of the 99+ languages, ready to translate. Free $0, 3 files a day.
3 files a day up to 30 minutes, free99+ languagesSpeaker separation and timestamps
In short
Upload the recording and the source text comes back in any of the any of the 99+ languages, with a timecode on every line and the speakers named on request. One file is 30 minutes; the free plan is $0 with 3 files a day.
The case
Why the source text needs the times
A translator works on a file that is spoken in one language and has to be read in another. The first half of that job is the transcript: the exact words of the interview, with the time they were said and the name of the person who said them. Only after that does the translation begin.
What a translator gets out of the recording
up to 30 minutes on the free plan, no minute limit on Pro
$0: 3 files a day, up to 30 minutes per file, no card
How to transcribe for a translation
- 1
Upload the file and name the language
Send the file to the widget above. Choose 99+ languages, or leave the detection on: our own measurement of 21 October 2026 gave 0.39 s to detect the language of a file.
- 2
Split the speakers
Switch on the split by speaker, so the interviewer and the guest each get a name on their lines. The source lines are what you translate against.
- 3
Read the source with its times
Every line carries the start and the end time of the segment, so the source sentence can be found in the recording and checked against the translation.
- 4
Hand the result to the editor
Export TXT for the source, CSV when the segments have to be worked through in a sheet, and DOCX on Pro for the file that goes into the job.
Tips
Check the names of the speakers in the first minute.
Keep the language on auto for a mixed recording.
Export CSV when the job is worked through segment by segment.
Keep the source text next to the recording while you translate.
Questions this page answers
Which languages can the recording be in?
9the recording language is detected automatically. We publish the 99+ languages the engine really has, not a long list of engines we do not have.
Does the service translate the recording?
No. It transcribes and gives you the source text with the timecodes; the translation is yours to write or to run through your own CAT tool.
Does the timecode survive the export?
Yes, in SRT, VTT, CSV and in the table of the Markdown file. The plain TXT export carries the words only, without the times.
Is the Word file free?
No. DOCX is a Pro feature: $10 a month when you pay yearly, or $20 a month when you pay monthly. The free plan is $0 with 3 files a day and up to 30 minutes per file.
Can the speakers be told apart?
Yes, with the split by speaker the voices are named separately on request: the interviewer and the guest stay apart. Without it no line says who spoke it, and a translated line cannot be checked against the question it answers.
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