Open Websearch

What is Open Websearch?

Open Websearch is a Model Context Protocol (MCP) server that provides web search capabilities using multiple search engines without requiring API keys or authentication. It is aimed at developers and MCP clients who need free, multi-engine search results with optional article content fetching.

How to use Open Websearch?

Install locally by cloning the repository, running npm install and npm run build, then adding the server to your MCP client configuration using the streamableHttp or SSE transport at http://localhost:3000/mcp or http://localhost:3000/sse. Docker deployment is also available via docker-compose up -d or a direct Docker run command. The server exposes three tools: search, fetchCsdnArticle, and fetchLinuxDoArticle. Configure default search engine, proxy, and CORS via environment variables.

Key features of Open Websearch

• Multi-engine web search (Bing, Baidu, CSDN, DuckDuckGo, Exa, Brave)
• No API keys or authentication required
• Configurable number of results and default search engine
• HTTP proxy support for accessing restricted resources
• Fetch complete article content from CSDN and Linux.do

Use cases of Open Websearch

• Integrate free web search into AI assistants and chatbots
• Search multiple sources simultaneously without individual API keys
• Fetch full article content from technical blogs (CSDN, Linux.do)
• Enable web research in environments that require HTTP proxies

FAQ from Open Websearch

What runtime or dependencies does Open Websearch require?

The server requires Node.js (npm) for local installation. Docker deployment is also supported for containerized environments.

Does Open Websearch need API keys or authentication?

No. The server scrapes search engine results directly, so no API keys or authentication are needed.

Where does search data come from?

Data is obtained by scraping multiple search engines (Bing, Baidu, CSDN, DuckDuckGo, Exa, Brave). Article content is fetched from CSDN and Linux.do.

What transports does Open Websearch support?

The server supports both streamable HTTP and SSE (Server‑Sent Events) transports.

Are there any usage limitations?

Yes. Rate limiting may cause temporary blocks from search engines. Result accuracy depends on the HTML structure of the engines. The tool is intended for personal use only, and users must comply with each engine's terms of service.