Official SDKs & MCP Server

Skip the raw HTTP requests. Use the official Node.js and Python SDKs, or plug Scrapingdog straight into Claude, Cursor, and any other MCP-compatible AI agent.

Pick your integration

Every package is open source, free to use, and covers every Scrapingdog API.

Zero dependencies

Node.js SDK

77 methods, one for every Scrapingdog API, plus a generic get() for anything new.

$ npm install scrapingdog-sdk

const { ScrapingDog } = require('scrapingdog-sdk');

const client = new ScrapingDog('APIKEY');
const serp = await client.google({ query: 'coffee' });
console.log(serp.organic_results);
View on npm
Only requests

Python SDK

75 methods covering search engines, e-commerce, social profiles, and more.

$ pip install scrapingdog

import scrapingdog

client = scrapingdog.Client(api_key="APIKEY")
results = client.google(query="coffee")
print(results)
View on PyPI
For AI agents

MCP Server

Exposes 35 tools so Claude Desktop, Claude Code, Cursor and other MCP clients can scrape directly.

$ npx -y scrapingdog-mcp

{
  "mcpServers": {
    "scrapingdog": {
      "command": "npx",
      "args": ["-y", "scrapingdog-mcp"],
      "env": { "SCRAPINGDOG_API_KEY": "APIKEY" }
    }
  }
}
View on npm

Built for how you already ship code

  • Typed, Structured Responses

    Every method returns parsed JSON with bundled TypeScript definitions — no manual response parsing.

  • Zero Extra Dependencies

    The Node.js SDK uses the built-in fetch; the Python SDK needs only requests. No bloated dependency tree.

  • Every Endpoint Covered

    Dedicated methods for every Scrapingdog API, plus a generic get() for anything released after you install.

  • Works With Any AI Agent

    The MCP server plugs Scrapingdog straight into Claude Desktop, Claude Code, Cursor, and other MCP-compatible clients.

  • Built-In Reliability Options

    Configure per-call timeouts, custom fetch implementations, and AbortSignal cancellation out of the box.

  • Secure By Default

    Your API key stays in environment variables — never hardcoded into agent context, transcripts, or logs.

What you can build with the SDKs

  • AI Agents & RAG Pipelines

    Give Claude, Cursor, or your own agent live web access through the MCP server, no custom tool code required.

  • Backend Data Pipelines

    Drop the Node.js or Python SDK into an existing ETL job and get structured JSON on every run.

  • Competitor Monitoring Scripts

    Schedule a script that calls a few SDK methods to track competitor pricing, ranking, or ads.

  • Internal Dashboards & Tools

    Wire SDK calls straight into your app backend to power internal scraping dashboards.

  • Rapid Prototyping

    Go from idea to working scraper in minutes with typed methods and sane defaults.

  • Research & Data Collection

    Use the Python SDK in a notebook to collect data for analysis without writing raw HTTP calls.

Get started in minutes

  1. 1

    Create a free account

    Sign up and get 200 free credits to test any SDK or the MCP server.

  2. 2

    Get your API key

    Copy your key from the dashboard — it goes into an environment variable, never into your code.

  3. 3

    Install your package of choice

    npm install scrapingdog-sdk, pip install scrapingdog, or wire the MCP server into your AI agent.

  4. 4

    Call any method

    Every method returns parsed, structured data — ready to use in your app or agent.

Transparent & Simple Pricing

Start your web scraping journey with 200 free credits. Test our service and upgrade to one of the plans below. Cancel anytime.

2 months Extra
  • FREE

    $0 forever
    • 200 Requests Credits
    • 1 Concurrency
    • Access To All APIs
    • No Credit Card Required
    • Community Support
    • Team Support
    Start for Free
  • LITE

    $40 /month
    • 200,000 Requests Credits
    • 5 Concurrency
    • Access To All APIs
    • Geotargeting
    • Email Support
    • No Team Support
    Start Trial
  • STANDARD

    $90 /month
    • 1,000,000 Requests Credits
    • 50 Concurrency
    • Access To All APIs
    • Geotargeting
    • Priority Email Support
    • No Team Support
    Start Trial
  • Popular

    PRO

    $200 /month
    • 3,000,000 Requests Credits
    • 100 Concurrency
    • Access To All APIs
    • Geotargeting
    • Priority Email Support
    • Team Support
    Start Trial
  • PREMIUM

    $350 /month
    • 6,000,000 Requests Credits
    • 150 Concurrency
    • Access To All APIs
    • Geotargeting
    • Priority Email Support
    • Team Support
    Start Trial

Frequently Asked Questions

  • Node.js and Python. Both cover every Scrapingdog API with dedicated, typed methods, plus a generic method for calling any endpoint directly.
  • It exposes Scrapingdog as tools that Claude Desktop, Claude Code, Cursor, and other MCP-compatible clients can call directly, so an AI agent can scrape the web without any custom integration code.
  • No. All three packages are free and open source (MIT licensed). You only pay for the API credits your requests use, same as calling the REST API directly.
  • Yes. Both the Node.js and Python SDKs work in any standalone script, serverless function, or backend service — no framework required.
  • No. The MCP server reads your key from the SCRAPINGDOG_API_KEY environment variable, so it never appears in the model's context, transcripts, or logs.

Try Scrapingdog for Free!

Get 200 free credits to spin the API. No credit card required!