Transcribe an interview
Upload the recording of an interview and get the text with a start and an end time on every segment and the two voices separated, so every quote can be found by the moment it was said. Free is $0 with 3 files a day; Pro is $10 a month when you pay yearly ($120 a year).
| What goes in | What comes out | What the price covers |
|---|---|---|
| What goes in | One recording of an interview: the audio of the conversation, ideally with a separate microphone for each voice | One file: up to 64 MB and 120 minutes (two hours) on every plan, including Team |
| What comes out | The text, and a start and an end time on every segment; speaker labels when you ask for them | On every plan, including the free one |
| Export | TXT, Markdown, SRT, VTT, CSV and DOCX | DOCX is a Pro feature; the other five are on every plan |
| Price and limits | Free: $0, 3 files a day, up to 30 minutes per file, no card and no trial clock | Pro: $10 a month paid yearly ($120 a year) or $20 a month, with fair use of 1,500 minutes of audio per calendar month |
| Where it runs | Our own transcription engine, a model profile tuned for Russian and for the Russian-English mix | A Russian or mixed file is processed on our own infrastructure outside the United States; a file that is entirely in English may go to a cloud speech provider in the United States when that fallback is enabled on the server |
How to do it in four steps
-
Agree with the other side before you record
An interview is a conversation with a person, so the consent of the other side is the normal rule, not a formality. We do not give legal advice and the rules differ by country.
-
Record so that the voices can be told apart
One microphone for the interviewer and one for the interviewee is the simple case: the engine then has one voice per channel. With a single room microphone the voices are separated by the diarization pass alone, which is the harder case. Keep the microphones at the same distance as usual — we are not writing a recording manual, only saying what helps.
-
Upload, split by speaker, read
Drop the file into the widget on this page, switch on Split by speaker, and read the first lines. The labels are placeholders: you rename them in the editor and the names stay in the export.
-
Find the quote and export it
Every segment carries its start and end time, so you can find the moment a quote was said. Export to DOCX or Markdown for the write-up, or to CSV if the quotes go into a table.
What goes into the transcript and what comes out
| Stage | What it is | What it costs |
|---|---|---|
| The recording | Usually 30 to 60 minutes of audio | Minutes count against the allowance: 1,500 minutes a month on Pro |
| The transcript | Text with timecodes on every line | Included on every plan, including the free one |
| The speakers | Two voices separated and labelled | Included on every plan; the labels are renamed in the editor |
| The quotes | Found by the moment they were said | Included: the start and end time of every segment |
| The write-up | DOCX or Markdown for the text, CSV for a table | DOCX is a Pro feature |
What it costs against the 1,500-minute allowance
| Your workload | Minutes a month | Where it lands |
|---|---|---|
| 4 interviews of 30 minutes | 120 | 8% of the allowance (calculated) |
| 10 interviews of 45 minutes | 450 | 30% of the allowance (calculated) |
| 20 interviews of 60 minutes | 1,200 | 80% of the allowance (calculated) |
| 30 interviews of 60 minutes | 1,800 | over the allowance by 300 minutes: the API answers with a clear error and the account stays open |
Put a file on the page and read the first lines.
The widget below is the same one as on the home page: file → transcript with timecodes and speakers → export. The transcript is saved to your account.
Drop an audio or video file
or — mp3, wav, m4a, mp4, up to 64 MB and two hours
Transcription runs under your account: sign in or create a free account.
No account yet? The free plan needs no card. Create a free account or sign in.
Questions this page answers.
How accurate is the transcript?
We publish no single accuracy number. The method of our benchmark is public — corpus, split, normalisation, metric, engine build and date — and the numbers come after the run. Ask us for the evaluation report and the sample set.
Will the interviewer and the interviewee be separated?
Yes, if you switch on Split by speaker: the two voices are told apart and labelled, and you rename the labels in the editor. One microphone per voice is the easier case; a single room microphone is the harder one.
Can I find a quote by the moment it was said?
Yes. Every segment carries its start and end time, so you can find the line in the text and the moment in the recording.
How long can the interview be?
Up to 120 minutes and 64 MB per file. A longer interview has to be split into parts — there is a step-by-step guide on the page how to transcribe a long recording.
How much does it cost?
Free is $0 with 3 files a day and up to 30 minutes per file, and no card is needed. Pro is $10 a month paid yearly ($120 a year) or $20 a month, with fair use of 1,500 minutes of audio per calendar month.
What happens to the recording of a person?
The file is processed to produce the transcript you asked for. It is not sold and not used to train models for other customers. Deleted files leave the live system immediately and the backups within 30 days.
Related pages.
The general checklist for calls, with speakers and timecodes.
Transcribe a phone call recording
A phone recording is narrowband — what that means for the text.
When the interview does not fit 120 minutes.
Every plan, every limit, in one table.
Upload, edit, export and keep the history of every transcript.
The live answer base: 99 questions about the price, the limits, the languages, the exports, the privacy and the API, each with the page of this site it comes from.
Mark Vesyoly, Founder.
Every number on this page comes from the service contract, the pricing page or our own measurements. The author of the page and the publisher of the site are the same person. Questions and corrections: support@transcrib.si.