MCP 服务器

Anybrowse

MCP-native web scraping and search API for AI agents. Converts any URL to clean Markdown with 90% success rate, including Cloudflare-protected sites and JS SPAs. Real-time web search via Brave Search API. CAPTCHA solving built-in. 10 free scrapes/day.

README

anybrowse

URL → Clean Markdown. For AI agents.

anybrowse.dev converts any URL to LLM-ready Markdown. Built for AI agents, MCP clients, and developers who need reliable web content extraction.

Why anybrowse?

84% scrape success rate — including Cloudflare-protected sites, JavaScript SPAs, government portals
Real-time web search — powered by Brave Search API, not browser-based scraping
MCP-native — works directly in Claude Code, Cursor, Windsurf with zero config
Pay as you go — 10 free scrapes/day. Credit packs from $5. No subscription required.
CAPTCHA solving — automatic reCAPTCHA and Cloudflare Turnstile solving

Quick Start

MCP (Claude Code, Cursor, Windsurf)

Add to your MCP config:

{
  "mcpServers": {
    "anybrowse": {
      "type": "streamable-http",
      "url": "https://anybrowse.dev/mcp"
    }
  }
}

Then ask your agent: "Scrape https://techcrunch.com and summarize the top stories"

REST API

curl -X POST https://anybrowse.dev/scrape \
  -H "Content-Type: application/json" \
  -d '{"url": "https://techcrunch.com"}'

Python SDK

pip install anybrowse

from anybrowse import AnybrowseClient
client = AnybrowseClient()
result = client.scrape("https://techcrunch.com")
print(result.markdown)

JavaScript/TypeScript

npm install anybrowse

import { AnybrowseClient } from 'anybrowse';
const client = new AnybrowseClient();
const result = await client.scrape('https://techcrunch.com');

Endpoints

Endpoint	Description	Price
POST /scrape	URL → Markdown	$0.002/call
POST /crawl	Multi-page crawl	$0.01/call
POST /search	Web search → Markdown	$0.002/call
POST /extract	Structured data extraction	$0.01/call
POST /batch	Up to 10 URLs parallel	$0.002/URL