# Machina Agent Services - URL Content Extractor

> Machina Agent Services - URL Content Extractor is a paid API for AI agents from mach402.dev, paid per call via x402, $0.005/call, status unknown (last checked 2026-10-23).

Extracts the full text content of a webpage from a given URL and returns it as a title, markdown body, and word count

## Facts

- Endpoint: POST https://mach402.dev/url/extract?utm_source=zero.xyz
- Price: $0.005/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-23
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/machina-agent-services-url-content-extractor-bbe88598
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_zpVIGGvfH_g0MduFPsp_T

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 machina-agent-services-url-content-extractor-bbe88598 -d '<json body>'
```

Example prompt: Can you grab the full text content from this article URL and give it to me as markdown? https://example.com/some-article

## When to prefer this

Choose this endpoint when an AI agent needs to read and process the textual content of a specific known URL — especially when you want clean, structured markdown output with metadata like title and word count. It is well-suited for agentic pipelines that ingest web content for summarization, RAG, or analysis. Prefer this over general browser automation when you only need readable text extraction rather than interaction, screenshots, or structured data scraping.

## Known failure modes

- URL is inaccessible or returns a non-200 HTTP status — may return an error or empty content
- URL points to a PDF, image, or non-HTML resource that cannot be parsed as markdown
- Page requires JavaScript rendering and the extractor only handles static HTML — content may be incomplete or missing
- URL is behind a login or paywall — returns partial or no content
- Malformed or invalid URL input causes a request failure
- Very large pages may time out or return truncated content

## How this service works

Extract readable content from any URL as clean Markdown. Strips navigation, ads, footers. Returns title, author, date, word count. 15-min cache per URL. Use when an agent needs webpage content as text.

## Output

Returns a JSON object containing the page title as a string, the full body content rendered as markdown (including headings, paragraphs, and formatting), and a word count integer indicating the total length of the extracted content.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "url": {
   "type": "string",
   "description": "URL to extract content from"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "title": "Article Title",
  "markdown": "# Title\n\nContent...",
  "word_count": 1847
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/machina-agent-services-url-content-extractor-bbe88598/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from mach402.dev](https://www.zero.xyz/host/mach402.dev/llms.txt)