# Article Text & Metadata Extractor

> Article Text & Metadata Extractor is a paid API for AI agents from predator-ubuntu.taild9a745.ts.net, paid per call via x402, $0.01/call, status unknown (last checked 2026-10-18).

Extracts clean article text and metadata (title, markdown content, word count) from any given URL

## Facts

- Endpoint: POST https://predator-ubuntu.taild9a745.ts.net/v1/extract?utm_source=zero.xyz
- Price: $0.01/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-18
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/article-text-metadata-extractor-a02b5226
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_viDzX4MV73BA68e2Z8tQ1

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 article-text-metadata-extractor-a02b5226 -d '<json body>'
```

Example prompt: Can you pull the full article text and title from this URL — https://www.nytimes.com/2024/01/15/technology/ai-news.html — and give it to me in markdown format?

## When to prefer this

Choose this endpoint when you need clean, readable article text from a specific URL and want the result as structured markdown with metadata (title, word count). Ideal for feeding web content into summarization, RAG pipelines, or research workflows without implementing your own scraper. Prefer this over general-purpose browser automation when you just need article body text and don't need to interact with the page.

## Known failure modes

- URL is behind a login/paywall — content may be empty or partial
- URL points to a non-article page (homepage, video, PDF) — markdown may be malformed or missing
- Network timeout if the target site is slow or unreachable
- Invalid or malformed URL returns an error
- Dynamically rendered JavaScript-heavy pages may yield incomplete extraction
- Rate limiting or bot detection on target site may block the fetch

## How this service works

Extract article text and metadata from a URL

## Output

Returns a JSON object containing the original URL, the article title, the full article body as markdown-formatted text, and a word count integer. This gives agents clean, readable content stripped of navigation, ads, and HTML boilerplate.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "url": {
   "type": "string",
   "format": "uri"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "object",
 "example": {
  "url": "https://example.com/article",
  "title": "Example",
  "markdown": "# Example\n\nArticle text...",
  "word_count": 512
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/article-text-metadata-extractor-a02b5226/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from predator-ubuntu.taild9a745.ts.net](https://www.zero.xyz/host/predator-ubuntu.taild9a745.ts.net/llms.txt)