Google Custom Search
Free通过 Google 自定义搜索 API 提供 Google 搜索结果。
FreeFree tier
Inputs: text, urlOutputs: text
About Google Custom Search
A Model Context Protocol (MCP) server that provides web search capabilities using the Google Custom Search API and webpage content extraction. It exposes two key tools: a search tool for performing web searches with configurable result counts (1-10) and a webpage reader tool that fetches, parses, and cleans webpage content by removing scripts and styles, returning structured data with title, text, and URL. Designed for integration with MCP-compatible clients like Claude Desktop, it requires a Google API Key and Search Engine ID for operation.
Key Features
Search tool with configurable result count (1-10)
Webpage reader tool that extracts title, main text, and URL
Removes scripts and styles from webpage content
Returns structured JSON results (title, link, snippet) for searches
Supports general web search or site-specific search via Google Custom Search API
Installable via Smithery or npm for easy integration with MCP clients
Open source under MIT license
Pros & Cons
Pros
- Provides both search and webpage reading in a single MCP server
- Leverages the reliable Google Custom Search API for search results
- Open source with permissive MIT license, allowing free use and modification
- Easy setup via Smithery or manual configuration with environment variables
- Configurable number of search results (up to 10) to control output size
- Webpage reader cleans content by removing scripts and styles for cleaner output
Cons
- Requires a Google API Key and Search Engine ID, adding setup overhead
- Limited to a maximum of 10 search results per query
- Depends on Google Custom Search API usage limits and potential costs
- Only works with MCP-compatible clients, not standalone
Best For
Integrating web search capabilities into MCP-compatible AI assistants like Claude DesktopExtracting and cleaning webpage content for further processing or summarizationBuilding custom research tools that combine search results with page readingAutomating data gathering from the web for analysis or reporting
FAQ
How do I get a Google API Key and Search Engine ID?
Create a Google Cloud project, enable the Custom Search API, and generate an API key from the Credentials page. Then create a Custom Search Engine at the Programmable Search Engine site, enable 'Search the entire web', and copy the Search Engine ID (cx).
What is the maximum number of results I can get from a search?
The search tool supports a 'num' parameter with a maximum value of 10. The default is 5.
Can this tool read any webpage?
Yes, the webpage reader tool can fetch and extract content from any publicly accessible URL. It returns the page title, cleaned main text, and the original URL.
Is this tool free to use?
The MCP server itself is free and open source under the MIT license. However, using the Google Custom Search API may incur costs depending on your usage tier.