Skip to main content
Glama
Serper MCP Server

Serper MCP Server

by garylab

MCP 服务器

铁匠徽章

通过 Serper 提供 Google 搜索的模型上下文协议 (MCP) 服务器。该服务器使 LLM 能够从 Google 获取搜索结果信息。

可用工具

Related MCP server: Scrapeless MCP Server

用法

通过 Smithery 安装

要通过Smithery自动为 Claude Desktop 安装 Serper MCP 服务器:

npx -y @smithery/cli install @garylab/serper-mcp-server --client claude

使用uv (推荐)

  1. 确保您已在操作系统上安装了uv 。

  2. 在您的 MCP 客户端代码配置或Claude设置(文件claude_desktop_config.json )中添加serper mcp 服务器:

    {
        "mcpServers": {
            "serper": {
                "command": "uvx",
                "args": ["serper-mcp-server"],
                "env": {
                    "SERPER_API_KEY": "<Your Serper API key>"
                }
            }
        }
    }

    uv将使用uvx从pypi.org自动下载 mcp 服务器并应用到您的 MCP 客户端。

使用pip进行项目

  1. 将serper-mcp-server添加到您的 MCP 客户端代码requirements.txt文件中。

    serper-mcp-server
  2. 安装依赖项。

    pip install -r requirements.txt
  3. 为您的客户端添加配置:

    {
        "mcpServers": {
            "serper": {
                "command": "python3",
                "args": ["-m", "serper_mcp_server"],
                "env": {
                    "SERPER_API_KEY": "<Your Serper API key>"
                }
            }
        }
    }

使用pip进行全局使用

  1. 确保pip或pip3在您的操作系统中。

    pip install serper-mcp-server
    # or
    pip3 install serper-mcp-server
  2. MCP 客户端代码配置或者Claude设置,添加serper mcp 服务器:

    {
        "mcpServers": {
            "serper": {
                "command": "python3",
                "args": ["serper-mcp-server"],
                "env": {
                    "SERPER_API_KEY": "<Your Serper API key>"
                }
            }
        }
    }

调试

您可以使用 MCP 检查器来调试服务器。对于uvx安装:

npx @modelcontextprotocol/inspector uvx serper-mcp-server

或者,如果您已将软件包安装在特定目录中或正在其上进行开发:

cd path/to/servers/src/serper
npx @modelcontextprotocol/inspector uv run serper-mcp-server -e SERPER_API_KEY=<the key>

执照

serper-mcp-server 采用 MIT 许可证。这意味着您可以自由使用、修改和分发该软件,但须遵守 MIT 许可证的条款和条件。更多详细信息,请参阅项目仓库中的 LICENSE 文件。

Available Tools

13 tools
google_search_autocompleteC

Search Google for results

Inputs: q (required), gl, location, hl, +2 more.

google_search_imagesD

Search Google for results

Inputs: q (required), gl, location, hl, +3 more.

google_search_lensC

Search Google for results

Inputs: url (required), gl, hl.

google_search_mapsC

Search Google for results

Inputs: q (required), ll, placeId, cid, +3 more.

google_search_newsD

Search Google for results

Inputs: q (required), gl, location, hl, +3 more.

google_search_patentsC

Search Google for results

Inputs: q (required), num, page.

google_search_placesD

Search Google for results

Inputs: q (required), gl, location, hl, +2 more.

google_search_reviewsD

Search Google for results

Inputs: fid (required), cid, placeId, sortBy, +4 more.

google_search_scholarC

Search Google for results

Inputs: q (required), gl, location, hl, +2 more.

google_search_shoppingD

Search Google for results

Inputs: q (required), gl, location, hl, +3 more.

google_search_videosC

Search Google for results

Inputs: q (required), gl, location, hl, +3 more.

webpage_scrapeC

Scrape webpage by url

Inputs: url (required), includeMarkdown.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    C2.6/5.0

    Scored across 13 tools

    Disambiguation5/5

    Each tool targets a distinct search vertical (web, images, news, etc.) as indicated by the tool name, making them clearly distinguishable. The identical descriptions are a minor issue but do not cause ambiguity because the names are self-explanatory.

    Naming Consistency4/5

    All search tools follow a consistent 'google_search_<type>' pattern, but 'webpage_scrape' breaks this pattern with a different verb and structure, causing a slight inconsistency.

    Tool Count5/5

    13 tools is well within the recommended 3-15 range and covers the major Google search verticals plus scraping, which fits the server's purpose without being excessive.

    Completeness4/5

    The tool set covers most common search types (web, images, news, maps, etc.) and includes scraping. Minor omissions like books or finance searches exist but do not significantly hinder typical usage.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      B
      quality
      D
      maintenance
      A Model Context Protocol server that enables LLMs to perform web searches using Google's Custom Search API through a standardized interface.
      1
      47
      MIT
    • A
      license
      Not graded
      quality
      D
      maintenance
      A Model Context Protocol server that enables AI assistants to perform web searches using Google Search API, returning up to 20 search results in JSON format.
      2
      Apache 2.0
    • F
      license
      C
      quality
      D
      maintenance
      A Model Context Protocol server that enables web search capabilities using the Tavily API, allowing AI models to retrieve current information from the internet through natural language commands.
      3
      -