Tool 03SpeechWhisper large-v3 turbo

Audio to text. Transcribe audio and video online

Drop a recording on the reel. It comes back as text you can read along with, search, and take away as subtitles.

MP3 · WAV · M4A · MP4 · MOV · WEBM · up to 15 minutes

0:00/ 0:00
01

How it works

  1. 1

    Drop the file

    Audio or video, straight onto the reel. You see its length and what it will cost before anything runs.

  2. 2

    Read along

    Press play and the words light up as they are spoken. Click any word to jump the recording to that moment.

  3. 3

    Take it with you

    Copy the text, or download TXT for reading and SRT or VTT for subtitles, with the timings already in.

02

What it costs

Credit packs →
3 creditsper minute of audio, signed in

Without an account you get 15 free runs across every tool here, for recordings up to 15 minutes each.

Signed in, a recording can run to 4 hours and 500 MB. You see the exact figure before you press the button, and nothing is charged if a run fails.

03

Questions

Not answered here? Ask us →

Which files can I transcribe?

MP3, WAV, M4A, AAC, OGG, FLAC and OPUS audio, and MP4, MOV, WEBM and MKV video. The sound is taken out of a video automatically, so there is nothing to convert first.

Which languages does it understand?

About a hundred, including English, Hindi, Urdu, Bengali, Tamil, Telugu, Spanish, French, German, Arabic, Japanese and Chinese. It detects the language on its own; choose one from the list if you want to be sure.

Can I get subtitles for a video?

Yes. Download SRT or VTT and the timings are already in the file, ready for YouTube, Premiere, CapCut or any player that takes subtitles.

What happens to my recording?

It is sent to the speech model, turned into text and deleted from our server as soon as the transcript is ready. The copy you play on this page never leaves your browser.

How long can a recording be?

Up to 15 minutes on a free run and up to 4 hours when you are signed in. Long recordings are split behind the scenes and stitched back together with their timings intact.

How accurate is it?

It runs Whisper large-v3 turbo, which handles clear speech very well and copes with accents and background noise better than most. Names and jargon are where any transcript needs a quick read-through.