How it works

1. Paste a link

Any public video or audio page. We check it first and show you the title, length and thumbnail. Nothing is deducted from your allowance until you press the button.

2. Pick what you need

Transcript gives you every word with timestamps. Turn on "Translate to English" if the speech is in another language and you want English out.

Download gives you the file: 360p, 480p, 720p, 1080p, 1440p or 4K, or audio on its own as MP3.

3. We do the work

Jobs go into a queue and run on our servers. Roughly one minute of processing per four minutes of audio. Downloads up to 720p are usually instant because the site already serves them in one piece; 1080p and above have to be stitched together from separate video and audio streams, which takes a few minutes.

You can close the tab. The job keeps running and the result waits in your history.

4. Take it away

Copy the text straight out, or export it:

• SRT — the standard subtitle file. Drop it into Premiere, Resolve, or a YouTube upload.
• VTT — subtitles for the web and HTML5 players.
• Timestamped text — readable notes with a clock reading on each line.
• Plain text — just the words.
• JSON — segment data, if you are feeding it into something else.

Getting better results

Accuracy follows audio quality, so a clean recording beats a good model every time. If you can pick the source, pick the one with less background music and fewer people talking over each other. If you know the language, choose it instead of leaving it on auto — detection is good but not free of mistakes on short clips.

What will not work

Private and unlisted-with-sign-in videos, paywalled courses, DRM-protected content, and live streams that have not finished. Some sites also block requests coming from servers rather than home connections; when that happens you will see a plain message saying so rather than a silent failure.