# Vevang AI Web Extractor (Structured JSON)

> Vevang AI Web Extractor (Structured JSON) is a paid API for AI agents from extract.vevang.com, paid per call via x402, $0.05/call, status unknown (last checked 2026-10-10).

Scrapes any public URL and returns structured JSON data shaped by a natural-language instruction you provide.

## Facts

- Endpoint: POST https://extract.vevang.com/extract?utm_source=zero.xyz
- Price: $0.05/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-10
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/vevang-ai-web-extractor-structured-json-cb45573c
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_sbA81keXSCGmQDs2pBtjR

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 vevang-ai-web-extractor-structured-json-cb45573c -d '<json body>'
```

Example prompt: Go to https://www.example.com/products and extract all product names, prices, and availability into structured JSON for me.

## When to prefer this

Choose this endpoint when you need web page content returned as structured, queryable JSON rather than raw markdown or HTML. It is ideal when you have a specific schema in mind (prices, contacts, listings, specs) and want an AI model to do the field mapping for you. Prefer this over the sibling markdown endpoint when downstream processing requires structured data rather than free-form text. Best for one-off extractions from public pages where you can describe the desired output in plain language.

## Known failure modes

- URL is behind a login or paywall — extraction fails or returns incomplete data
- Page uses heavy JavaScript rendering that prevents content from being read
- Instruction is too vague or ambiguous — returned JSON may be incomplete or misaligned
- URL is malformed or unreachable — HTTP error or timeout returned
- Page structure changes break expected extraction patterns
- Rate limiting or bot protection on the target site blocks the fetch
- x402 payment fails — request not processed

## How this service works

AI Web Extractor (structured) — scrape a url into the exact JSON you ask for. POST {url, instruction}; a fast model reads the page and returns structured data per your instruction (prices, contacts, listings, specs, tables). Web scraping to structured JSON for agents. x402-native.

## Output

A structured JSON object shaped according to the natural-language instruction provided. The response contains the extracted fields (e.g. prices, names, contacts, specs, table rows) parsed directly from the live page content at the given URL. The exact schema of the returned JSON mirrors what was requested in the instruction.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "url": {
   "type": "string",
   "description": "the public http(s) URL to read"
  },
  "instruction": {
   "type": "string",
   "description": "what structured data to extract from the page"
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/vevang-ai-web-extractor-structured-json-cb45573c/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from extract.vevang.com](https://www.zero.xyz/host/extract.vevang.com/llms.txt)