TikTok Transcript Scraper — TikTok Video to Text

Get the TikTok transcript of public videos as clean JSON, plain text, SRT, VTT, or RAG chunks with timestamps — in bulk, with no API key. If you need to turn a TikTok video to text, paste the URLs and run.

It reads TikTok's own native caption tracks (no speech-to-text guessing), and you only pay for transcripts actually delivered. TikTok throttles and captcha-walls datacenter IPs, so this actor runs a hosted residential-proxy layer with automatic retries on fresh IPs.

Run in 30 seconds

Minimal input — one field:

{
  "videoUrls": ["https://www.tiktok.com/@khaby.lame/video/7106658991907802411"]
}

Back comes one dataset item per video with a transcript:

{
  "videoId": "7106658991907802411",
  "url": "https://www.tiktok.com/@khaby.lame/video/7106658991907802411",
  "author": "khaby.lame",
  "language": "en",
  "captionSource": "auto-generated",
  "text": "the whole transcript as one clean string ...",
  "segments": [{ "start": 0.12, "duration": 2.4, "text": "first caption line" }],
  "segmentCount": 14,
  "durationSeconds": 21
}

What it does

Give it a list of TikTok video URLs (or short vm.tiktok.com links, or raw IDs) and it returns the transcript for each:

• Clean JSON with per-line timestamps
• Plain text, SRT, VTT, and RAG chunks — richer output than bare-text scrapers
• Video metadata: author, description, duration, like/view counts, available caption languages

Output formats

Option	What you get
includePlainText	Whole transcript as one string
includeSegments	[{ start, duration, text }]
includeSrt / includeVtt	Subtitle file strings
ragChunking	Overlapping timestamped chunks for embeddings

What you'll pay

$0.003 per video with a transcript. The arithmetic is plain:

Transcripts delivered	Cost
1	$0.003
100	$0.30
1,000	$3.00

Videos with no captions (no speech, captions off, photo posts) go to the skipped list and are never charged. No subscription, no API key.

RAG / LLM use cases

Built for retrieval-augmented generation and AI agents. Turn TikTok content into clean, chunked, timestamped context for vector stores (LangChain, LlamaIndex) — or feed plain text straight to an LLM for summaries and analysis.

Use as an MCP tool / with AI agents

This actor is exposed as an MCP tool (streamable HTTP) at:

https://mcp.apify.com/?tools=jamhimself/tiktok-transcript-scraper

Example MCP client configuration:

{
  "mcpServers": {
    "apify": {
      "type": "streamable-http",
      "url": "https://mcp.apify.com/?tools=jamhimself/tiktok-transcript-scraper",
      "headers": { "Authorization": "Bearer <YOUR_APIFY_TOKEN>" }
    }
  }
}

Any MCP client — Claude, Cursor, or agent frameworks — can call this actor directly and get the transcript back as structured output.

Why scrapers return empty transcripts (and why this one doesn't)

TikTok serves a verification/captcha interstitial instead of the real video page to IPs it doesn't trust — which includes most datacenter ranges that free libraries run on. A scraper that doesn't detect the wall "succeeds" with an empty page and returns no transcript. This actor detects walled responses, discards that IP, and retries on fresh residential sessions (several rotations per video) before reading TikTok's native caption tracks directly. Genuinely unavailable videos (deleted, private, region-locked) are recognized as terminal and skipped without burning retries. Actively maintained.

FAQ

• Do I need a TikTok API key? No.
• What if a video has no captions? Reported in the skipped list, not charged. (Photo posts and silent videos have no transcript.)
• Auto-captions and translations? Yes — original auto-captions are preferred; machine translations are used as a fallback for your requested language.
• Is this legal? It accesses publicly available caption data. You are responsible for complying with TikTok's Terms of Service and applicable law in your use of the output.

---

If this actor saved you time, an honest review helps others find it.

TikTok is a trademark of ByteDance Ltd. This actor is not affiliated with or endorsed by them. Built and maintained by Jamhimself LLC.