AI REFERENCE BREAKDOWN FOR UGC CREATORS

Understand a TikTok.
Start with one link.

We collect the available text. AI breaks down how it is written when it completes.

No questionnaire or upload · text only, no video or audio review

Your TikTok reference

TikTok video URLUse a /video/ link or a vm.tiktok.com / vt.tiktok.com share link.

02 · Choose a brief optional

What are you making your next video for?

Who is it for?

What should it work toward?

03 · What can you actually film? optional A result screenshot from my own product An original product / service demo My own operating process or workflow An actual use example I have

Check only material you can film. Unchecked material stays unknown; only checked choices are treated as your filming confirmations.

Advanced: add custom details optional

Text is optional. A custom description replaces the preset description for that field; clear it to use the preset again. Checked material choices remain your confirmations.

Custom product / offer Custom audience Custom material you can film

Advanced: add transcript or on-screen text

Use only my text · skip public metadata lookup

A public video may already have published subtitles. We do not create a new audio transcription or read text from the reference video automatically. Paste the words yourself when you need to add evidence; supplied text is labelled as yours.

Spoken words (optional) On-screen text (optional)

One public scan per day. No account needed.

No link? Try practice text

What can this read?

Public metadata
Caption, creator and counts when the connector returns them.
Published subtitles
TikTok-published text and timecodes when available. The report shows the language and generation label without claiming the audio was checked.
What AI adds
A breakdown of the available text when AI completes. An optional, experimental adaptation asks at most one question per attempt, only if needed. You can skip it.

Text evidence only
Video viewing, keyframe review, OCR of the reference video, new audio transcription, comment collection and retention analysis are not enabled. Published subtitles are a text source, not proof that the audio or visuals were reviewed.

Common questions & limits

Can it review any TikTok video?

Only public video links supported by the connector. Removed, private, region-blocked and unsupported videos may return an error or a partial report. Empty metadata stays labelled as unavailable.

Are published subtitles a new audio transcription?

No. These are text and timecodes already published with the TikTok video, read from its WebVTT subtitle track when available. They may be marked auto-generated. We do not independently check them against the audio or watch the video.

Does it extract speech, on-screen text or comments automatically?

New audio transcription, automatic reading of the reference video and comment collection are not enabled in this release. The link is enough to start. We use available descriptions and published subtitles and clearly show what could not be retrieved. No upload is required.

Does it retrieve new metadata on every report?

Metadata may be reused from a cache. The report labels reused data and shows when those fields were originally fetched. The report creation time does not mean the counts were fetched at that moment.

What if AI cannot finish?

Your source report stays available, with missing data and the incomplete AI step labelled. You can still copy or download the report.

Are the video drafts already validated?

No. Optional adaptations are experimental, and drafts appear only when AI completes. The hook, proof to film, call to action and fit are proposed plans. They do not mean the reference video showed that proof or that you adopted or filmed the plan. Verify the source and your available material before filming.

Will it predict virality?

No. The output proposes small tests. Verify the source, change one part of the next post and measure the audience response.

Your TikTok reference

Your source report.