Web summarizer API

Summarize and extract structured data from any web page

Web summarizer API

The web summarizer service lets you extract summaries and structured data from any publicly accessible URL. It returns page titles, descriptions, headings, paragraphs, links, images, and Open Graph metadata.

No authentication is required. The service runs as a standalone deployment separate from the main Agentbot API.

Base URL

https://<your-web-summarizer-host>

The web summarizer runs as an independent service on port 3100 by default. Replace the base URL with your deployment’s address.

Service info

GET /

Returns service metadata and available endpoints.

Response

{
  "service": "Agentbot Web Summarizer",
  "description": "Summarize and extract data from any URL",
  "endpoints": {
    "POST /api/summarize": "Extract title, description, headings, key content — { url }",
    "POST /api/extract": "Extract all links, images, meta tags — { url }"
  },
  "payment": {
    "network": "base",
    "currency": "USDC",
    "payTo": "0x9Fc073659284575850614f6286158803F0526Bc2",
    "note": "x402 gating coming soon"
  }
}

Health check

GET /health

No authentication required.

Response

{
  "status": "ok",
  "service": "agentbot-web-summarizer",
  "version": "1.0.0"
}

Summarize a URL

POST /api/summarize

Fetches the given URL and extracts a content summary including the page title, meta description, headings, key paragraphs, and word count.

Request body

Field	Type	Required	Description
url	string	yes	The URL to summarize

curl -X POST https://<your-web-summarizer-host>/api/summarize \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com"}'

Response

{
  "url": "https://example.com",
  "title": "Example Domain",
  "description": "This domain is for use in illustrative examples.",
  "headings": ["Example Domain"],
  "paragraphs": [
    "This domain is for use in illustrative examples in documents."
  ],
  "wordCount": 28,
  "fetchedAt": "2026-03-22T19:00:00.000Z"
}

Response fields

Field	Type	Description
url	string	The URL that was summarized
title	string	Page title from the <title> tag, falling back to the first <h1>
description	string	Meta description from <meta name="description"> or <meta property="og:description">
headings	string[]	Deduplicated list of h1, h2, and h3 headings (max 20)
paragraphs	string[]	Deduplicated key content paragraphs extracted from <p>, <article>, and common content selectors (max 10)
wordCount	number	Approximate word count of the full page body text
fetchedAt	string	ISO 8601 timestamp of when the URL was fetched

Errors

Code	Description
400	url field is missing from the request body
500	The target URL could not be fetched or parsed

Error response

{
  "error": "url required"
}

Extract links, images, and metadata

POST /api/extract

Fetches the given URL and extracts all links, images, and Open Graph / Twitter Card metadata.

Request body

Field	Type	Required	Description
url	string	yes	The URL to extract data from

curl -X POST https://<your-web-summarizer-host>/api/extract \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com"}'

Response

{
  "url": "https://example.com",
  "links": [
    {
      "href": "https://www.iana.org/domains/example",
      "text": "More information..."
    }
  ],
  "images": [
    {
      "src": "https://example.com/logo.png",
      "alt": "Example logo"
    }
  ],
  "meta": {
    "og:title": "Example Domain",
    "og:description": "This domain is for use in illustrative examples.",
    "description": "This domain is for use in illustrative examples."
  },
  "fetchedAt": "2026-03-22T19:00:00.000Z"
}

Response fields

Field	Type	Description
url	string	The URL that was extracted from
links	object[]	Deduplicated list of links found on the page (max 50)
links[].href	string	Resolved absolute URL of the link
links[].text	string	Link text, truncated to 100 characters
images	object[]	Deduplicated list of images found on the page (max 20)
images[].src	string	Resolved absolute image URL
images[].alt	string	Image alt text, truncated to 100 characters
meta	object	Open Graph (og:*), Twitter Card (twitter:*), and description meta tags
fetchedAt	string	ISO 8601 timestamp of when the URL was fetched

Errors

Code	Description
400	url field is missing from the request body
500	The target URL could not be fetched or parsed

Error response

{
  "error": "url required"
}

Fetch behavior

All URL fetching uses the following defaults:

Setting	Value
User-Agent	Agentbot-WebSummarizer/1.0
Timeout	10 seconds
Redirects	Followed automatically
Accepted content	text/html, application/xhtml+xml

If the target URL returns a non-2xx HTTP status, the service responds with a 500 error containing the upstream status code in the error message.