# agdata YouTube Transcript Extractor

> agdata YouTube Transcript Extractor is a paid API for AI agents from agdata.shveik.dev, paid per call via x402, $0.032/call, status unknown (last checked 2026-10-15).

Fetches the full transcript of a YouTube video in multiple formats (JSON, plain text, SRT, VTT) for a specified language.

## Facts

- Endpoint: POST https://agdata.shveik.dev/v1/mpp/youtube-transcript?utm_source=zero.xyz
- Price: $0.032/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-15
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/agdata-youtube-transcript-extractor-945b0587
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_d9uARvhMYfm-hL3roDCAV

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 agdata-youtube-transcript-extractor-945b0587 -d '<json body>'
```

Example prompt: Can you grab the English transcript of this YouTube video — https://www.youtube.com/watch?v=dQw4w9WgXcQ — and give it back to me as plain text?

## When to prefer this

Use this endpoint when you need to extract the spoken content from a YouTube video as structured or plain-text data, especially when you want timing metadata (SRT/VTT), multi-language support, or rich video metadata alongside the transcript. Prefer this over generic web unblockers when the target is specifically a YouTube transcript, as it handles YouTube's caption API natively and returns structured segment data.

## Known failure modes

- Video has no captions or subtitles available — transcript array will be empty
- Requested language not available for the video — may fall back or return error
- Invalid or inaccessible video URL — returns error response
- Video is private or age-restricted — may be blocked
- Truncation of very long transcripts indicated by truncated:true flag

## How this service works

Unblock any URL as Markdown, or get search, social, maps, jobs and news as data. Paid per call in stablecoins over x402 or MPP, with no account.

## Output

Returns a JSON object containing the video ID, title, channel name, channel URL, publish date, view count, duration in seconds, transcript segments (with timing), the language used, subtitle type (user-generated or auto), number of segments, available languages, output format, and a truncation flag.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "lang": {
   "enum": [
    "any",
    "en",
    "de",
    "es",
    "fr",
    "it",
    "ja",
    "ko",
    "nl",
    "pt",
    "ru"
   ],
   "type": "string"
  },
  "video": {
   "type": "string",
   "maxLength": 2048
  },
  "format": {
   "enum": [
    "json",
    "text",
    "srt",
    "vtt"
   ],
   "type": "string"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "url": "https://…",
  "title": "Rick Ast…",
  "format": "json",
  "videoId": "dQw4w9Wg…",
  "language": "en",
  "truncated": false,
  "viewCount": 1822672170,
  "channelUrl": "https://…",
  "transcript": [],
  "channelName": "Rick Ast…",
  "publishedAt": "2009-10-…",
  "segmentCount": 61,
  "subtitleType": "user_gen…",
  "durationSeconds": 213,
  "availableLanguages": []
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/agdata-youtube-transcript-extractor-945b0587/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from agdata.shveik.dev](https://www.zero.xyz/host/agdata.shveik.dev/llms.txt)