# Parallel Extract

> Parallel Extract is a paid API for AI agents from Parallel at parallelmpp.dev, paid per call via MPP or x402, $0.01/call, status unknown (last checked 2026-10-11).

Scrapes and extracts structured content (title, excerpts, full text) from one or more URLs using the Parallel Web Systems API.

## Facts

- Endpoint: POST https://parallelmpp.dev/api/extract?utm_source=zero.xyz
- Price: $0.01/call
- Payment: MPP, x402
- Status: unknown
- Last checked: 2026-10-11
- Activations on Zero: 0
- Provider: Parallel
- Docs: https://parallelmpp.dev/#agents
- Website: https://parallel.ai
- Tags: answers, search, web, extraction, research, ai, x402
- Canonical page: https://www.zero.xyz/c/parallel-extract-ed5ce778
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_yAZErE4KSovkcL7XAxNwQ

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 parallel-extract-ed5ce778 -d '<json body>'
```

Example prompt: Can you scrape the content from https://techcrunch.com/2024/05/01/openai-news/ and extract the main article text and headings?

## When to prefer this

Use this endpoint when you need to retrieve and read the actual textual content of known URLs — ideal for AI pipelines that need to ingest web page text, for research workflows pulling from specific sources, or when you already have URLs from a prior search step and want to extract their contents. Prefer this over a search endpoint when you know the exact URLs; prefer it over a generic HTTP fetch when you need clean, parsed, readable text rather than raw HTML.

## Known failure modes

- URL is unreachable or returns a non-200 HTTP status — error reported per URL in the errors array
- URL points to a JavaScript-heavy SPA that cannot be scraped server-side — excerpts may be empty
- Rate limiting or upstream scraping block — partial results with errors
- Invalid URI format in the urls array — validation error returned
- Network timeout on slow or large pages — may return null full_content

## How this service works

Extract content from URLs

## Output

Returns a structured result per URL including the page title, an array of content excerpts, optionally the full page content, the publish date if available, and a unique extract_id. Also includes usage counts, any errors per URL, and global warnings.

## Example request

```json
{
 "urls": [
  "https://example.com"
 ],
 "objective": "Extract main content and headings"
}
```

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "required": [
  "urls"
 ],
 "properties": {
  "urls": {
   "type": "array",
   "items": {
    "type": "string",
    "format": "uri"
   },
   "description": "URLs to extract content from"
  },
  "objective": {
   "type": "string",
   "description": "Optional extraction objective to guide content selection"
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/parallel-extract-ed5ce778/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from parallelmpp.dev](https://www.zero.xyz/host/parallelmpp.dev/llms.txt)