🔍 web-search-plus-mcp

License: MIT

Multi-provider web search MCP server with intelligent auto-routing.

A Model Context Protocol server that gives AI assistants access to 7 search providers with intelligent auto-routing. Analyzes query intent and picks the best provider automatically — no manual switching needed. Install, configure your keys, and go.

✨ Features

Intelligent auto-routing — analyzes query intent and picks the best provider automatically
7 search providers — use one or all, graceful fallback if any key is missing
Zero install option — run instantly with uvx web-search-plus-mcp
MCP-native — works with Claude Desktop, NanoBot, and any MCP-compatible host

🔎 Supported Providers

Provider	Best for	Free tier
Serper (Google)	Facts, news, shopping, local businesses	2,500 queries/month
Tavily	Deep research, analysis, explanations	1,000 queries/month
Querit	Multi-lingual AI search with rich metadata and real-time info	1,000 queries/month
Exa (Neural)	Semantic discovery, finding similar content	1,000 queries/month
Perplexity	AI-synthesized answers with citations	Via API key
You.com	Real-time RAG, LLM-ready snippets	Limited free tier
SearXNG	Privacy-first, self-hosted, $0 cost	Free (self-hosted)

🧠 Auto-Routing Examples

Query	Routed to	Why
"iPhone 16 Pro price"	Serper	Shopping intent detected
"how does TCP/IP work"	Tavily	Research/explanation intent
"latest multilingual EV market updates"	Querit	Real-time AI search
"companies like Stripe"	Exa	Discovery/semantic intent
"what is quantum computing"	Perplexity	Direct answer intent

🚀 Quick Start

# Run instantly with uvx (no install needed)
uvx web-search-plus-mcp

# Or install globally with pip
pip install web-search-plus-mcp
web-search-plus-mcp

⚙️ Claude Desktop Config

Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "web-search-plus": {
      "command": "uvx",
      "args": ["web-search-plus-mcp"],
      "env": {
        "SERPER_API_KEY": "your_serper_key",
        "TAVILY_API_KEY": "your_tavily_key",
        "QUERIT_API_KEY": "your_querit_key",
        "EXA_API_KEY": "your_exa_key",
        "PERPLEXITY_API_KEY": "your_perplexity_key",
        "YOU_API_KEY": "your_you_key",
        "SEARXNG_BASE_URL": "https://your-searxng-instance.example.com"
      }
    }
  }
}

🤖 NanoBot Config

Add to your NanoBot config.json under mcp_servers:

{
  "mcp_servers": [
    {
      "name": "web-search-plus",
      "command": "uvx",
      "args": ["web-search-plus-mcp"],
      "env": {
        "SERPER_API_KEY": "your_serper_key",
        "TAVILY_API_KEY": "your_tavily_key",
        "QUERIT_API_KEY": "your_querit_key",
        "EXA_API_KEY": "your_exa_key",
        "PERPLEXITY_API_KEY": "your_perplexity_key",
        "YOU_API_KEY": "your_you_key",
        "SEARXNG_BASE_URL": "https://your-searxng-instance.example.com"
      }
    }
  ]
}

🔑 Environment Variables

Variable	Provider	Sign up
`SERPER_API_KEY`	Serper (Google)	console.serper.dev
`TAVILY_API_KEY`	Tavily	tavily.com
`QUERIT_API_KEY`	Querit	querit.ai
`EXA_API_KEY`	Exa	exa.ai
`PERPLEXITY_API_KEY`	Perplexity	docs.perplexity.ai
`YOU_API_KEY`	You.com	you.com/api
`SEARXNG_BASE_URL`	SearXNG (self-hosted)	docs.searxng.org

At least one provider is required. More providers = better routing coverage. SearXNG needs no API key — just point SEARXNG_BASE_URL at your self-hosted instance.

You can also drop a .env file next to the server:

SERPER_API_KEY=xxx
TAVILY_API_KEY=xxx
QUERIT_API_KEY=xxx
EXA_API_KEY=xxx
PERPLEXITY_API_KEY=xxx
YOU_API_KEY=xxx
SEARXNG_BASE_URL=https://your-searxng-instance.example.com

🛠 Tool Reference

`web_search`

Parameter	Type	Default	Description
`query`	string	required	Search query
`provider`	string	`"auto"`	Force a provider: `auto`, `serper`, `tavily`, `querit`, `exa`, `perplexity`, `you`, `searxng`
`count`	integer	`5`	Number of results to return

Credits

Built on the web-search-plus routing logic — a multi-provider search skill for OpenClaw with intelligent auto-routing.

🔍 web-search-plus-mcp