What is Mcp Search Hub

MCP Search Hub is an intelligent multi-provider search aggregation server built on FastMCP 2.0, designed to unify search results from multiple leading providers into a single interface.

Use cases

Use cases include integrating multiple search providers for enhanced information retrieval, reducing operational costs by 30-45% compared to single-provider solutions, and leveraging unique strengths of each provider for optimal search results.

How to use

Users can deploy MCP Search Hub using Docker or as a standalone Python service. It supports both HTTP and STDIO transport methods, allowing for easy integration with existing systems.

Key features

Key features include embedding official MCP servers from five leading providers, automatic query routing based on characteristics, intelligent result ranking and deduplication, cost control mechanisms, performance caching, graceful error handling, and zero maintenance for provider updates.

Where to use

MCP Search Hub can be utilized in various fields such as academic research, content aggregation, real-time information retrieval, and any application requiring efficient and accurate search capabilities.

MCP Search Hub

🔍 Intelligent multi-provider search aggregation server built on FastMCP 2.0

MCP Search Hub embeds official MCP servers from five leading search providers within a unified interface, intelligently routes queries to the most appropriate provider(s), and combines/ranks results for optimal relevance.

✨ Key Features

• 🏗️ Unified Interface: Single MCP server for all search providers
• 🧠 Intelligent Routing: Automatically selects optimal providers based on query characteristics
• ⚡ Official Provider Integration: Embeds official MCP servers (Linkup, Exa, Perplexity, Tavily, Firecrawl)
• 💰 Cost Optimization: 30-45% cost reduction through intelligent provider selection
• 🚀 Zero Maintenance: Provider updates flow through automatically
• 📊 Smart Result Merging: Deduplication and multi-factor ranking
• ⚡ Performance Optimized: Multi-tier caching and parallel execution
• 🔧 Production Ready: Error handling, rate limiting, monitoring

🚀 Quick Start

Get up and running in 5 minutes:

# 1. Clone and install
git clone https://github.com/BjornMelin/mcp-search-hub
cd mcp-search-hub
uv pip install -r requirements.txt

# 2. Interactive setup (recommended)
python scripts/setup_claude_desktop.py

# 3. Start searching!
# The script will guide you through API key setup and Claude Desktop integration

After setup, restart Claude Desktop and ask: “What MCP tools do you have available?”

📖 Complete Setup Guide | ⚡ Quick Setup Script

🎯 Why MCP Search Hub?

Provider Strengths

Each provider excels in different areas. MCP Search Hub automatically routes queries to leverage these strengths:

Accuracy scores from SimpleQA benchmark

Architecture Benefits

graph TB
    User[Claude Desktop/Code] --> Hub[MCP Search Hub]
    Hub --> Router[Intelligent Router]
    
    Router -->|Academic Query| Exa[Exa MCP Server]
    Router -->|News Query| Perplexity[Perplexity MCP Server]
    Router -->|Factual Query| Linkup[Linkup MCP Server]
    Router -->|Scraping Task| Firecrawl[Firecrawl MCP Server]
    
    Hub --> Cache[Smart Caching]
    Hub --> Merger[Result Merger]
    
    style Hub fill:#e1f5fe
    style Router fill:#f3e5f5
    style Cache fill:#e8f5e8
    style Merger fill:#fff3e0

Benefits:

• Single Integration: One MCP server instead of five
• Zero Configuration: No need to manage multiple servers
• Automatic Updates: Provider improvements flow through automatically
• Cost Efficient: Intelligent routing reduces unnecessary API calls
• Quality Results: Multi-provider consensus improves accuracy

📚 Documentation

For Users

• Getting Started - Complete setup guide and first steps
• Configuration - Detailed configuration reference
• API Reference - Complete tool and endpoint documentation
• Troubleshooting - Solutions to common problems

For Developers

• Contributing - How to contribute to the project
• Development - Development workflows and practices
• Architecture - System design and components

Advanced Topics

• Provider Integration - How providers are embedded
• Caching Strategy - Multi-tier caching system
• Performance Tuning - Optimization strategies

📖 Complete Documentation Index

🔑 Quick Configuration

At minimum, you need API keys for the providers you want to use:

# Copy environment template
cp .env.template .env

# Add your API keys
LINKUP_API_KEY=your_linkup_key
EXA_API_KEY=your_exa_key
PERPLEXITY_API_KEY=your_perplexity_key
TAVILY_API_KEY=your_tavily_key
FIRECRAWL_API_KEY=your_firecrawl_key

Getting API Keys

⚙️ Complete Configuration Guide

🚀 Available Tools

MCP Search Hub provides all tools from embedded provider MCP servers:

Core Search Tools

• search - Intelligent multi-provider search with automatic routing
• get_provider_info - Information about available providers

Provider-Specific Tools (25+ tools available)

• Firecrawl: firecrawl_scrape, firecrawl_search, firecrawl_crawl, firecrawl_extract
• Exa: web_search_exa, research_paper_search, company_research, linkedin_search
• Perplexity: perplexity_ask, perplexity_research
• Linkup: linkup_search_web
• Tavily: tavily_search, tavily_extract

🛠️ Full Tool Documentation

🏗️ Architecture

MCP Search Hub uses a modular architecture with embedded MCP servers:

graph TB
    subgraph "MCP Search Hub"
        Server[FastMCP Server]
        Router[Query Router]
        Cache[Tiered Cache]
        Merger[Result Merger]
    end
    
    subgraph "Embedded MCP Servers"
        LinkupMCP[Linkup MCP]
        ExaMCP[Exa MCP]
        PerplexityMCP[Perplexity MCP]
        TavilyMCP[Tavily MCP]
        FirecrawlMCP[Firecrawl MCP]
    end
    
    Client[Claude Desktop] --> Server
    Server --> Router
    Router --> LinkupMCP
    Router --> ExaMCP
    Router --> PerplexityMCP
    Router --> TavilyMCP
    Router --> FirecrawlMCP
    
    style Server fill:#e1f5fe
    style Router fill:#f3e5f5
    style Cache fill:#e8f5e8
    style Merger fill:#fff3e0

🏗️ Architecture Documentation

🤝 Contributing

We welcome contributions! Here’s how to get started:

1. Fork the repository
2. Create a feature branch
3. Make your changes with tests
4. Submit a pull request

🤝 Contributing Guide | 🛠️ Development Setup

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgements

• FastMCP - The framework powering this server
• Provider Teams - For their excellent APIs and official MCP servers

🔗 Links

• GitHub Repository
• Documentation
• Issue Tracker
• Releases

Ready to get started? Follow our Getting Started Guide or run the interactive setup:

python scripts/setup_claude_desktop.py