# TikTok Video Transcript Extractor

> TikTok Video Transcript Extractor is a paid API for AI agents from api.scrapeforagents.tech, paid per call via x402, $0.025/call, status unknown (last checked 2026-10-02).

Extracts transcripts and captions from public TikTok videos given their URLs

## Facts

- Endpoint: POST https://api.scrapeforagents.tech/v1/tiktok-transcript?utm_source=zero.xyz
- Price: $0.025/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-02
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/tiktok-video-transcript-extractor-63c13507
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_vwe-neqnZL3Fbr_QTG6zc

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability tiktok-video-transcript-extractor-63c13507 -d '<json body>'
```

Example prompt: Can you pull the full transcript of this TikTok video for me: https://www.tiktok.com/@username/video/1234567890? I want to see exactly what's being said with the subtitle language noted.

## When to prefer this

Use this endpoint when you need structured transcript data from one or more public TikTok videos without building your own scraping pipeline. It handles subtitle extraction and language detection in one call, and does not charge for failed or empty runs, making it cost-safe for uncertain inputs.

## Known failure modes

- Video is private or deleted — item returned with null transcript fields
- Video has no subtitles or captions available — hasTranscript returns false
- Invalid or malformed TikTok URL — item may return with nulls or be skipped
- TikTok rate limiting or anti-scrape measures — may result in empty result set (not charged)
- maxItems set too low for the number of submitted URLs — excess videos silently dropped

## How this service works

Pay-per-call structured web data. Failed or empty runs are not charged.

## Output

Returns a JSON object with a count of results and an array of items, each containing the video ID, URL, caption, transcript text, whether a transcript was available, the subtitle source, and the subtitle language detected.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "videos": {
   "type": "array",
   "description": "Public TikTok video URLs or official short links. One dataset item is returned per available video."
  },
  "maxItems": {
   "type": "integer",
   "default": 0,
   "minimum": 0,
   "description": "Maximum number of unique videos to return. Set to 0 for no cap."
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "count": 1,
  "items": [
   {
    "id": null,
    "url": null,
    "caption": null,
    "transcript": null,
    "hasTranscript": null,
    "subtitleSource": null,
    "transcriptText": null,
    "subtitleLanguage": null
   }
  ],
  "product": "tiktok-transcript"
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/tiktok-video-transcript-extractor-63c13507/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from api.scrapeforagents.tech](https://www.zero.xyz/host/api.scrapeforagents.tech/llms.txt)