MCP Connector

Add web scraping to
any AI assistant

The ScrapingBot MCP server connects your AI tools to the web. Scrape any site, extract structured data, get search results — directly inside your AI workflow. Works with Claude, Cursor, Windsurf, Continue, Zed and more.

Get free API key View on GitHub

Open standard

Works with your favorite AI tools

MCP is an open protocol. One config, any compatible client.

Claude Desktop Anthropic's desktop app — Windows & macOS

Cursor AI code editor, most popular for developers

Windsurf AI-powered IDE by Codeium

Continue Open-source AI coding assistant — VS Code & JetBrains

Zed Fast open-source editor with built-in AI

LibreChat Open-source ChatGPT-style interface

Quick setup

Ready in 3 steps

Same config works in every MCP-compatible client.

Get your API key

Sign up at scraping-bot.io — 500 free credits/month, no credit card required. You'll find your username and API key in your dashboard.

Create free account →

Install the MCP server

The config is the same for all clients. Add this JSON block to your client's MCP config file:

{
  "mcpServers": {
    "scraping-bot": {
      "command": "npx",
      "args": ["-y", "github:Sene/scraping-bot-mcp"],
      "env": {
        "SCRAPING_BOT_USERNAME": "your-username",
        "SCRAPING_BOT_API_KEY": "your-api-key"
      }
    }
  }
}

Claude Desktop

~/Library/Application Support/Claude/claude_desktop_config.json %APPDATA%\Claude\claude_desktop_config.json

macOS path on the left, Windows on the right

Cursor

Settings → MCP → Add server — paste the config above

Windsurf

~/.codeium/windsurf/mcp_config.json

Continue

~/.continue/config.json

Add inside the "mcpServers": { ... } key

Zed

~/.config/zed/settings.json

Use the "context_servers" key

Ask your AI to scrape

Restart your AI client. The ScrapingBot tools are now available. Just ask naturally — your AI will call the right tool automatically.

Requirements: Node.js 18+ must be installed on your machine. Download Node.js →

Prompts

Real prompts, real results

Ask naturally in any supported client — your AI picks the right tool.

Scrape the Amazon product page at [URL] and tell me the title, price, and rating.

scrape_retail Returns structured JSON with price, title, brand, reviews

Search Google for 'best web scraping API 2026' and summarize the top 10 results

scrape_search_engine Returns organic results, titles, descriptions, URLs

Get the property details for this Rightmove listing: [URL]

scrape_real_estate Returns price, bedrooms, address, description

Scrape this Booking.com hotel and list all available room types and prices

scrape_booking Returns room data, pricing, amenities

Get the raw HTML of this website and extract all the product names from it

scrape_url Returns full HTML for custom parsing

MCP tools

5 scraping tools

Each tool maps to a ScrapingBot API endpoint and returns clean, structured data.

scrape_url

Scrape any URL

Returns raw HTML from any webpage. Handles proxies, CAPTCHAs, and JS rendering automatically.

scrape_retail

E-commerce data

Extract structured product data from Amazon, eBay, Walmart, and 100+ retailers.

scrape_real_estate

Real estate listings

Get structured listing data from Zillow, Rightmove, SeLoger, and other property sites.

scrape_search_engine

Search results

Extract Google or Bing SERP results as structured JSON with titles, URLs, and descriptions.

scrape_booking

Travel & booking

Scrape hotel prices, availability and reviews from Booking.com, Airbnb, and more.

△

Open source — The MCP server is MIT licensed. View the source, contribute, or fork it on GitHub. github.com/Sene/scraping-bot-mcp

Configuration

Advanced options

Pass options to any scraping tool to fine-tune behavior for difficult targets.

Option	Description
useChrome	JavaScript rendering with real Chrome
premiumProxy	Residential proxies for tough sites
proxyCountry	Target a specific country (US, GB, FR…)
waitForNetworkRequests	Wait for async content to load before returning HTML

Add web scraping toany AI assistant