Transcribe an interview to text

In an account — free or Pro, with the same features — the transcript comes with:

Who spoke

In American English, British English and Indian English, the transcript separates the voices into Speaker 1, Speaker 2 and so on; you rename each label once, and the name holds for the whole transcript. In Australian English and Philippine English, it separates them too, on recordings of up to 2 hours. If the separation fails, the transcript comes without speaker labels, and says so.

A time on every word

Click any line to hear that passage. The transcript follows the audio while it plays, so checking a quote does not mean hunting through the recording.

Editing in the text itself

Fix names, acronyms and misheard words right in the transcript, with the passage playing beside it.

Export

Word (DOCX), TXT, SRT and VTT, on every plan.

A summary that quotes instead of paraphrasing

For an interview, the summary gathers the themes, keeps notable quotes in the speaker's own words — quoted exactly, grammar untidied — and lists the factual claims a reader might want to check.

Questions about the interview

Ask, for example, what the interviewee said about a subject. The answer points to the passages and the moment in the recording they are in.

Every automatic transcript has errors, and they cluster on proper names, numbers and technical terms. Before you quote a passage, click the line and listen: with a time on every word, that check takes seconds.

If your department asks for a literal transcript, with hesitations, repetitions and fillers such as “like” and “you know”, check that point against the audio: an automatic transcript does not promise to record each of them.

An interview you quote from may also be an interview you promised to anonymise. Rename each speaker label once, then edit names and places out of the text before you export it. Your recordings and transcripts are not used to train models.

Without an account, the trial transcribes up to the first 10 minutes of a recording. On Pro, each file can be up to 8 hours, within the month's included hours; what each plan includes, the pricing page shows. See plans and prices.

On long recordings, speaker labels can wander from one section to the next — a consequence of how long audio is processed, not a setting.

Name the main language rather than leaving it on automatic. Automatic detection has to decide what language each stretch of speech is in, and where a recording switches language, what was said in the other one may come back translated instead of transcribed — or not come back at all. Either way nothing is marked as missing: the transcript will look complete.

Philippine English is marked “preview” in the language list: offered, but without a quality commitment yet, so check its transcript more closely before you use it.