TikTok Transcript Scraper

Extract transcripts (subtitles/captions) from TikTok videos in every available language. Returns both full plain-text and timestamped segments per language — ideal for content analysis, translation, SEO research, and accessibility.

Features

• All languages at once — one result row per video per language, including creator-provided subtitles and TikTok's machine-translated versions
• Timestamped segments — precise [start, end, text] cues for each caption
• Full plain-text transcripts — ready-to-use joined text
• Video metadata included — author, caption, publish date, duration
• No cookies required — works against public videos only
• Anti-detection — Firefox stealth browser, randomized fingerprint

Input

Field	Type	Description
postUrls	Array of strings (required, min 1)	TikTok video URLs to extract transcripts from

Example Input

{
  "postUrls": [
    "https://www.tiktok.com/@khaby.lame/video/7595604254929210646",
    "https://www.tiktok.com/@charlidamelio/video/7594925330234218794"
  ]
}

Output

One dataset item per (video × language). A video with 6 available languages produces 6 output rows.

Field	Type	Description
url	String	Original video URL
success	Boolean	Whether a transcript was extracted
message	String	Status message (empty on success)
postId	String	Unique TikTok video ID
username	String	Author's username
nickname	String	Author's display name
caption	String	Video caption/description
publishDate	String	Human-readable UTC publish date
duration	Integer	Video duration (seconds)
language	String	Language code (e.g. eng-US, spa, por)
source	String	Creator (author-provided) or MT (machine-translated)
transcript	String	Full transcript text for this language
segments	Array	Timestamped cues: [{start, end, text}, ...]
segmentCount	Integer	Number of segments

Example Output (one row)

{
  "url": "https://www.tiktok.com/@khaby.lame/video/7595604254929210646",
  "success": true,
  "message": "",
  "postId": "7595604254929210646",
  "username": "khaby.lame",
  "nickname": "Khabane lame",
  "caption": "Try something different next time #learnfromkhaby",
  "publishDate": "2026-01-15 15:03:52",
  "duration": 16,
  "language": "eng-US",
  "source": "Creator",
  "transcript": "Hey guys welcome back to another video ...",
  "segments": [
    { "start": 0.0, "end": 2.5, "text": "Hey guys welcome back" },
    { "start": 2.5, "end": 4.2, "text": "to another video" }
  ],
  "segmentCount": 2
}

FAQ

Q: Do I need to provide cookies or a login? No. The scraper works entirely with public TikTok videos.

Q: What if a video has no subtitles? The scraper emits one row per video with success=false and message="No transcript available". All metadata fields remain populated.

Q: How many rows will I get per video? One row per available language. TikTok often provides 5–10 languages per popular video (one creator-uploaded plus several machine translations).

Q: Are machine-translated captions included? Yes. The source field tells you whether a transcript was uploaded by the creator (Creator) or auto-translated by TikTok (MT).

Q: Which video URL formats are supported? Standard TikTok video URLs such as https://www.tiktok.com/@username/video/1234567890123456789.

Q: Can private or deleted videos be scraped? No. Private, deleted, or region-restricted videos will return a failure row with an explanatory message.

Limitations

• Only public videos are supported
• Some videos may be region-restricted
• Videos without any TikTok-generated captions cannot produce transcripts