# tokenguard Web Extraction

> tokenguard Web Extraction is a paid API for AI agents from eltociear-tokenguard.hf.space, paid per call via x402, $0.05/call, status unknown (last checked 2026-09-14).

Fetches a URL or accepts raw text and converts it into structured JSON matching a caller-supplied schema shape

## Facts

- Endpoint: POST https://eltociear-tokenguard.hf.space/llm/extract
- Price: $0.05/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-14
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/tokenguard-web-extraction-91c4168c
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_6um53B4o8QdE2dE6LjScW

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 tokenguard-web-extraction-91c4168c -d '<json body>'
```

Example prompt: Fetch the page at https://en.wikipedia.org/wiki/OpenAI and extract the founding date, headquarters location, and CEO name into a JSON object with keys foundingDate, headquarters, and ceo.

## When to prefer this

Choose this endpoint when you need to consume an arbitrary web page or unstructured text blob as typed, schema-conforming JSON in a single API call. It is especially useful when the target page has no native API, when the schema varies per use case (since the caller supplies it at runtime), or when you want to avoid building and maintaining a bespoke scraper. Prefer alternatives if you need JavaScript rendering of a heavy SPA, real-time monitoring of page changes, or bulk batch extraction across hundreds of URLs.

## Known failure modes

- URL is inaccessible, paywalled, or returns a non-200 status — extraction returns empty or partial results
- schema_hint is too complex or ambiguous — model may hallucinate fields or mismap values
- Page content is mostly JavaScript-rendered (SPA) and the fetcher sees only an empty shell
- Extracted values are plausible but factually wrong if the source text is ambiguous
- No text or URL provided — request fails with a validation error
- Rate limiting or HF Space cold-start increases latency significantly

## How this service works

Turn messy text — or any URL's page content — into structured JSON matching the shape you ask for, so an agent can consume a web page as data instead of prose

## Output

A structured JSON object whose keys and value types conform to the schema_hint provided by the caller. When a URL is given, the endpoint first fetches the page content, then applies LLM-powered extraction to produce the requested fields. When raw text is provided, extraction runs directly on that text. The response contains only the requested data in machine-consumable form.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "url": {
   "type": "string",
   "description": "Page to fetch and extract from (used when text is omitted)"
  },
  "text": {
   "type": "string",
   "description": "Raw text to extract from"
  },
  "instruction": {
   "type": "string",
   "description": "Optional extra extraction instruction"
  },
  "schema_hint": {
   "type": "object",
   "description": "Target JSON shape"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "data": {
   "plan": "Pro",
   "price_usd": 20
  },
  "provider": "llm7",
  "source_url": "https://example.com/pricing"
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/tokenguard-web-extraction-91c4168c/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from eltociear-tokenguard.hf.space](https://www.zero.xyz/host/eltociear-tokenguard.hf.space/llms.txt)