Transcribe TikTok Video

1. Copy the link. In the app: tap the share arrow on the video, then Copy link. On the web: copy the address bar. Both the long form (tiktok.com/@user/video/123…) and the short vt.tiktok.com share links work — short links are resolved automatically.
2. Paste it into the tool. One link for a single transcript, or several at once — 2 per request on the free plan, up to 10 on Pro.
3. Click transcribe. Most clips finish in well under a minute.
4. Take what you need. Copy the text from the page, or download it: TXT to write with, SRT or WebVTT to caption with.
5. Check the risky words. Names, brands, and numbers are where speech recognition slips. Use the timestamps to jump back and confirm anything you plan to quote or publish.

There is no account step anywhere in that list, and no card. The free tier is 15 transcriptions per week, up to 2 videos per request.

Accuracy is real but conditional, and it is more honest to tell you where it breaks than to claim a single number. On clear speech — a creator talking to camera, voiceover narration, a piece-to-camera explainer — expect around 95%. That is usually good enough to publish after a skim.

Where it degrades:

• Loud music under the voice. A voiceover over a quiet backing track transcribes fine. A creator rapping over a beat, or a track mixed louder than the speech, will cost you accuracy — sometimes a lot.
• People talking over each other. Duets, stitches, and group videos produce a single merged text stream; the transcript will not tell you who said what.
• Very fast delivery and heavy accents land lower than clear standard speech, though usually still readable.
• No speech at all. A TikTok that is a trending sound plus on-screen text has no spoken audio to convert, and will come back empty. That is not a bug — transcription reads audio, and the words in that video were never spoken.

The practical rule: transcribe first, then verify the specific words you intend to rely on. The 5% that is wrong is disproportionately the names and numbers you most want to be right.

One video is a novelty. The reason people keep a transcription tool open is the batch: paste up to 10 links in a single run on Pro and get every transcript back together, plus one combined AI summary across the whole set. That is what turns transcription into research.

The workflows this unlocks:

• Competitor sweeps — transcribe a creator's last ten videos and read how they actually talk about the problem. It is voice-of-customer data that is already public.
• Hook libraries — the first line of each transcript is the hook. Ten at a time, the patterns become obvious.
• Your own back catalogue — turn a year of videos into a searchable text archive of everything you have said on camera.

A 60-second TikTok yields roughly 100–150 words, so ten of them is about the length of a blog post — read in a few minutes instead of watched in ten.