Get a YouTube transcript

Paste any YouTube URL above and the transcript comes back timestamped and ready to export. If the video carries YouTube's own caption track, that text is what you get, cleaned up into whole sentences instead of the short, speech-model-timed fragments the player shows. If it doesn't — an older upload, a livestream replay, a video where the uploader turned captions off — the same request transcribes the audio directly and returns the identical formats.

That last part is the reason most people end up here. YouTube's transcript panel disappears whenever the uploader disables it, and the third-party scrapers built on the caption endpoint fail with it, because there is no caption track left to scrape. Transcribing the audio sidesteps the question entirely: the spoken words exist whether or not anyone published a text track for them.

  • Works on the full YouTube range — a 30-second Short, a two-hour lecture, a livestream VOD.
  • Captions disabled, auto-captions missing, or non-English audio: all handled by transcribing the audio itself.
  • Timestamped segments, so a quote you pull can be cited back to the second it was said.
  • Playlists and channels can go through the REST API or CLI rather than one paste at a time.

Captions on YouTube, as they actually are

Most videos carry YouTube's automatic captions, which cover dozens of spoken languages, plus a smaller set of creator-uploaded caption tracks. Coverage drops for older uploads, livestream VOD replays, and heavily accented or noisy audio, where auto-captions technically exist but are unreliable enough that transcribing the audio directly is often the better source anyway.

Typical length
Widest range of any platform here — under a minute for Shorts, up to multi-hour lectures, podcasts, and livestream replays.
URL shape
youtube.com/watch?v=...

What you're doing with the transcript

The export format and the workflow both depend on what the text is for. These go into detail for YouTube specifically:

FAQ

The free tier includes 15 minutes of video a month, transcripts in the language spoken, and TXT/SRT/VTT/Markdown/JSON export — enough to run a YouTube video through and judge the output before committing to anything.