What is a TikTok transcript API?
A TikTok transcript API takes a public video URL and returns spoken or captioned text as structured data. Social Fetch: GET /v1/tiktok/videos/transcript — WebVTT in data.transcript.content when a caption track exists, optional useAiFallback for videos without one.
What the response actually contains
Pass public video URL (/@user/video/… or vm.tiktok.com). Returns { data, meta } with lookupStatus plus transcript when found.
Check lookupStatus first. found → WebVTT in data.transcript.content. not_found → no caption and no fallback unless opted in. Photo carousels → transcript_target_not_video.
The AI fallback, and why it's TikTok-specific
Most TikTok clips lack native captions. Optional useAiFallback=true runs speech-to-text when no caption track exists.
YouTube transcript endpoint has no equivalent — captions only. Leave fallback off for compliance scans that care about creator-added captions; turn on when you need spoken words regardless.
What it costs
1 credit base for existing caption track; useAiFallback adds speech-to-text credits — check operation page and meta.creditsCharged.
found and not_found bill; lookup_failed and 503 don't. Retry only the latter.
When teams reach for this
RAG pipelines, brand-safety scans, repurposing tools needing WebVTT cue timing.
Try /tools/tiktok-transcript (plain text, no WebVTT or useAiFallback) then API for production.
Common mistakes
Retrying not_found without useAiFallback — same billed answer. Use useAiFallback=true if you need text anyway.
Photo carousel URLs → transcript_target_not_video, not lookupStatus.
data.transcript.content is WebVTT, not plain text — parse or strip timestamps deliberately.
How this compares to other ways of getting a TikTok transcript
Unofficial scrapers: free, break often, no speech-to-text fallback. TikTok publishes no public transcript API.
YouTube endpoint: same lookupStatus pattern, plainText + segments, no AI fallback.
FAQ
How do I get a TikTok transcript via API?
GET /v1/tiktok/videos/transcript with x-api-key and url. Check lookupStatus, read WebVTT in data.transcript.content.
Does the TikTok transcript API support videos without captions?
Yes with useAiFallback=true — speech-to-text when no caption track. Without it → not_found.
Does it always return captions?
No. Most clips have no native track. useAiFallback generates from audio at added credit cost.
What URL formats work?
Public video URLs including vm.tiktok.com. Carousels → transcript_target_not_video.
How much does the TikTok transcript endpoint cost?
1 credit for caption read; useAiFallback adds cost. found/not_found bill; lookup_failed/503 don't.
Is this the same shape as YouTube transcripts?
Same lookupStatus envelope. TikTok: WebVTT + useAiFallback. YouTube: plainText + segments, no fallback.
Is there a free way to try the TikTok transcript API?
/tools/tiktok-transcript — Turnstile-gated, plain text only, no useAiFallback.
Can I get transcript text with timestamps for clipping?
Yes via API — WebVTT carries per-cue timestamps. Free tool is plain text only.