searxng-mcp
Privacy-focused web search MCP server using SearXNG with Streamable HTTP transport, supporting authentication and advanced search parameters.
README
SearXNG MCP Server
A Model Context Protocol (MCP) server that provides privacy-focused web search capabilities through SearXNG, using Streamable HTTP transport.
Features
- Privacy-Focused Search: Leverages SearXNG for anonymous, tracker-free web searches
- Streamable HTTP Transport: Modern HTTP-based MCP protocol for remote access
- Environment Variables: All configuration via environment variables
- HTTP Basic Auth: Support for authentication-protected SearXNG instances
- Advanced Search Parameters: Support for categories, language filters, and time ranges
- Async Architecture: Built with async/await for non-blocking operations
Installation
# Clone the repository
git clone <repo-url>
cd searxng-mcp
# Install dependencies
pip install -e .
# Or using uv
uv pip install -e .
Configuration
All configuration is done via environment variables:
| Variable | Description | Default |
|---|---|---|
SEARXNG_BASE_URL |
Base URL of your SearXNG instance | http://127.0.0.1:8888 |
SEARXNG_USERNAME |
Username for HTTP Basic Auth (optional) | - |
SEARXNG_PASSWORD |
Password for HTTP Basic Auth (optional) | - |
SEARXNG_TIMEOUT |
Request timeout in seconds | 10 |
MAX_RESULTS_LIMIT |
Maximum results allowed per query | 50 |
HOST |
Server host address | 0.0.0.0 |
PORT |
Server port | 3000 |
BASE_URL |
Base URL path for the MCP server | /searxng-mcp |
LOG_LEVEL |
Logging verbosity | INFO |
Using Environment File
Create a searxng.env file from the example:
cp searxng.env.example searxng.env
Edit searxng.env with your settings:
# SearXNG Configuration
SEARXNG_BASE_URL=https://example.com/searxng
SEARXNG_USERNAME=your_username
SEARXNG_PASSWORD=your_password
SEARXNG_TIMEOUT=10
MAX_RESULTS_LIMIT=50
# Server Configuration
HOST=0.0.0.0
PORT=3000
BASE_URL=/searxng-mcp
# Logging
LOG_LEVEL=INFO
Running
Using Python (with env file)
# Load from searxng.env file and run
export $(grep -v '^#' searxng.env | xargs) && python -m searxng_mcp
Using CLI
# With environment variables inline
SEARXNG_BASE_URL=https://example.com/searxng \
SEARXNG_USERNAME=your_username \
SEARXNG_PASSWORD=your_password \
searxng-mcp
Using Docker
Build the image:
docker build -t searxng-mcp .
Run with environment file:
docker run -d \
--name searxng-mcp \
--env-file searxng.env \
-p 127.0.0.1:3000:3000 \
searxng-mcp
Run with inline environment variables:
docker run -d \
--name searxng-mcp \
-e SEARXNG_BASE_URL=https://example.com/searxng \
-e SEARXNG_USERNAME=your_username \
-e SEARXNG_PASSWORD=your_password \
-e HOST=0.0.0.0 \
-e PORT=3000 \
-p 127.0.0.1:3000:3000 \
searxng-mcp
SearXNG Instance
You need a running SearXNG instance. Options:
- Local:
docker run -d --name searxng -p 8888:8080 searxng/searxng:latest - Remote: Use any public or private SearXNG instance
If your SearXNG instance is protected with HTTP Basic Auth, set SEARXNG_USERNAME and SEARXNG_PASSWORD.
MCP Client Configuration
Configure your MCP client to connect to this server:
{
"mcpServers": {
"searxng": {
"type": "streamable-http",
"url": "http://localhost:3000"
}
}
}
If using a reverse proxy:
{
"mcpServers": {
"searxng": {
"type": "streamable-http",
"url": "http://your-server/searxng-mcp"
}
}
}
Usage
Once connected, use the web_search tool:
Basic Search
Search for "Python async programming tutorials"
Search with Maximum Results
Search for "machine learning news" with max_results=20
Category-Specific Search
Available categories:
general- General web searchnews- News articlesimages- Image searchvideos- Video contentfiles- File downloadsscience- Scientific papers and resourcesit- IT and tech resources
Language-Specific Search
Use ISO 639-1 language codes: en, de, fr, es, zh, ja, etc.
Time-Range Filtered Search
Available time ranges: day, week, month, year
Using with Reverse Proxy
For production deployment with a custom base URL, use a reverse proxy:
Nginx Example
location /searxng-mcp/ {
proxy_pass http://127.0.0.1:3000/;
proxy_http_version 1.1;
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection "upgrade";
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
}
License
MIT License
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。