Live API

Web Scraper

Give your AI agent access to the live web. One API call returns clean markdown, meta information, or full HTML — with JavaScript fully rendered. No headless browser setup required.

Pay-per-call via HTTP 402JS rendering includedMarkdown, HTML, or metaTry it free

Try it now

Integration

API Reference

Two ways to connect: a standard REST endpoint or an MCP server for agent frameworks.

REST Endpoint

POST

Endpoint

POST https://web-scraper.fourotwo.xyz/scrape

Request Body (JSON)

Example

{
  "url": "https://example.com",   // required
  "markdown": true,               // optional, default: true
  "html": false                   // optional, default: false
}

Parameters

url

required

The URL to scrape. Must be a publicly accessible web page.

markdown

optional · default: true

Return the page content as clean markdown.

html

optional · default: false

Return the full rendered HTML after JavaScript execution.

cURL Example

curl -X POST https://web-scraper.fourotwo.xyz/scrape \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com"}'

MCP Server

Streamable HTTP

Connect any MCP-compatible AI agent or framework directly to the Web Scraper via the Model Context Protocol. Uses streamable HTTP transport — no WebSocket setup required.

MCP Endpoint

https://web-scraper.fourotwo.xyz/mcp

MCP Config Example

{
  "mcpServers": {
    "fourotwo-web-scraper": {
      "url": "https://web-scraper.fourotwo.xyz/mcp"
    }
  }
}

Learn more about the Model Context Protocol at modelcontextprotocol.io