MCP Catalogs
Homegoogle-surf-mcp screenshot

google-surf-mcp

by HarimxChoi·169·Score 50

A robust Google search MCP server combining search, content extraction, and academic paper analysis without API keys.

searchweb-scrapingai-llm
23
Forks
2
Open issues
this month
Last commit
2d ago
Indexed

Overview

Google-surf-mcp is a comprehensive MCP server that provides Google search capabilities along with URL content extraction and academic paper analysis. It uses Playwright with a persistent Chrome profile to bypass CAPTCHAs through human intervention when needed, and drops sponsored ads and knowledge panels to deliver clean search results. The server supports both abstract and full content extraction modes, with special support for academic sources like arXiv, Nature, and PubMed. It's designed for both local use and headless/serverless environments with proper configuration.

Try asking AI

After installing, here are 7 things you can ask your AI assistant:

you:AI agents needing real-time web information retrieval without API key limitations
you:Academic researchers requiring efficient access to papers across multiple repositories
you:Content creators who need to analyze search results before following links
you:Automation workflows that combine web search with content analysis
you:Does this require Google API keys?
you:How does it handle CAPTCHA challenges?
you:Can it extract content from PDFs?

When to choose this

Choose this when you need a complete search, fetching, and extraction solution without API keys and want CAPTCHA handling built-in.

When NOT to choose this

Don't choose for high-volume applications or serverless environments without manual CAPTCHA solving capability.

Tools this server exposes

5 tools extracted from the README
  • searchsearch(query, limit?)

    Perform a single Google search and return results with ads and knowledge panels removed.

  • search_parallelsearch_parallel(queries[], limit?)

    Execute multiple Google searches in parallel (up to 10 queries per call).

  • extractextract(url, max_chars?, mode?)

    Fetch and extract content from a URL with options for full text, abstract, or metadata.

  • search_extractsearch_extract(query, limit?, max_chars?, mode?)

    Combine Google search with content extraction in one call, with abstract or full mode options.

  • healthhealth()

    Check server status including cascade mode, rate-limiter usage, cache size, and configuration.

Comparable tools

bing-search-mcpweb-scraper-mcpduckduckgo-search-mcp

Installation

npm install -g google-surf-mcp

Configure your MCP client:

{
  "mcpServers": {
    "google-surf": {
      "command": "npx",
      "args": ["-y", "google-surf-mcp"]
    }
  }
}

First tool call auto-bootstraps the warm profile (Chrome may open briefly).

FAQ

Does this require Google API keys?
No, this MCP server works without any API keys by using browser automation.
How does it handle CAPTCHA challenges?
It has four CAPTCHA recovery modes: OS notification (default), visible Chrome, remote debugging, or cloud mode with fail-fast.
Can it extract content from PDFs?
Yes, it supports PDF extraction from academic sources like arXiv, Nature, PubMed, and others.

Compare google-surf-mcp with

GitHub →

Last updated · Auto-generated from public README + GitHub signals.