Transcribe Audio to Text

MP3, WAV, M4A, FLAC, OGG, AAC cover most of what you'll upload, and most other common formats work too.

No hard ceiling, and here's the mechanism. Past about thirteen minutes, the file gets cut into roughly eight-minute pieces, each piece transcribed separately and rejoined, with the timestamps kept on your original clock. A two-hour recording comes back as one transcript, not four disconnected ones.

No, not from the sound of the voices. The transcript comes back as words and lines with their timings, nobody identified by voice. Ask for names anyway and you get a guess inferred from what was said, not from how the voices sound, which you then have to check yourself; on a two-person interview, one asking and one answering, that check is easy. For anything more reliable, record each person to their own file, or split a stereo recording into its channels before you transcribe.

One clear voice comes back readable, wrong mostly on names it has no way of knowing. Two people talking over each other come back as one line, not two. The transcript runs on a single clock, so when voices overlap, one wins and the other gets dropped or blended in.

Yes. Ask for bullet-point summaries, key takeaways, action items, or any other format. Chat Octopus works from the full transcript to produce exactly what you need.

Yes. Ask for a DOCX for documents or an SRT for subtitles; both download from your account.

Yes. Drop in the video and Chat Octopus transcribes the spoken audio directly. There is no need to extract the audio yourself first.

Only you can open the files you upload; nobody else using the product can reach them. Temporary copies made while a job runs are cleared inside a week. Recordings submitted on or after August 2, 2026 are used to train our systems unless you opt out, which starts with an email to [email protected].

Yes, that's the normal way to use it. Record the memo on your phone, then upload the finished file: nothing has to be listening while you talk. Works the same for a dictated note, a voicemail, or a lecture recording as it does for an interview.

No. Phone dictation writes words down live, while you're talking into it, and stops the moment you stop. This works the other way round: you hand over a recording that already exists and get text back from it. If you only have dictation, you can still record yourself with a voice memo app and upload that instead.