PlumyCat/Multi-Provider-MCP-Server

Multi-Provider MCP Server - Universal LLM "Swiss Army Knife"

A unified Model Context Protocol (MCP) server that provides a single, consistent interface to multiple AI providers. Perfect for Microsoft Copilot Studio and other MCP-compatible applications.

🎯 Key Concept: One Server, One Provider, One Model

Unlike traditional multi-provider servers, this MCP server is designed to be configured once per provider. Each server instance focuses on a single AI provider with its optimal default settings, making it perfect for enterprise deployments.

🚀 Perfect for Microsoft Copilot Studio

This MCP server provides a unified interface for all AI providers, making it ideal for:

• Microsoft Copilot Studio custom connectors
• Enterprise AI deployments with provider flexibility
• Consistent bot behavior across different LLM providers
• Easy provider switching without code changes

📦 Installation

Quick Start with npm

npm install -g multi-provider-mcp-server

Manual Installation

git clone https://github.com/PlumyCat/Multi-Provider-MCP-Server.git
cd Multi-Provider-MCP-Server
npm install
npm run build

🔧 Configuration

Supported Providers

Choose one provider per server instance:

Environment Variables

Set the appropriate API key based on your chosen provider:

# For DeepSeek
DEEPSEEK_API_KEY=your-deepseek-api-key

# For OpenAI
OPENAI_API_KEY=your-openai-api-key

# For Azure OpenAI
AZURE_OPENAI_API_KEY=your-azure-openai-api-key
AZURE_OPENAI_BASE_URL=https://your-resource.openai.azure.com

# For Claude
CLAUDE_API_KEY=your-claude-api-key

# For Mistral
MISTRAL_API_KEY=your-mistral-api-key

# For Codestral
CODESTRAL_API_KEY=your-codestral-api-key

# For Gemini
GEMINI_API_KEY=your-gemini-api-key

Basic MCP Configuration

For Claude Desktop (claude_desktop_config.json)

{
  "mcpServers": {
    "LLM": {
      "command": "node",
      "args": ["F:/MCP-Server/Multi-Provider-MCP-Server/build/index.js"],
      "cwd": "F:/MCP-Server/Multi-Provider-MCP-Server",
      "env": {
        "PROVIDER": "azure-openai"
      }
    }
  }
}

For Cursor (mcp.json)

{
  "mcpServers": {
    "LLM": {
      "command": "node",
      "args": ["F:/MCP-Server/Multi-Provider-MCP-Server/build/index.js"],
      "cwd": "F:/MCP-Server/Multi-Provider-MCP-Server",
      "env": {
        "PROVIDER": "azure-openai"
      }
    }
  }
}

🎯 Usage

Simple Interface

The server provides two simple tools with no complex configuration:

1. chat_completion

Basic single-turn chat completion:

{
  "message": "Hello, how are you?"
}

2. multi_turn_chat

Multi-turn conversation with maintained history:

{
  "messages": "Continue our conversation about AI"
}

No Configuration Required

The beauty of this approach: The bot/client doesn't need to know:

• Which AI provider is being used
• Which model is active
• What parameters are optimal
• How to handle different APIs

Everything is handled automatically by the server configuration.

🏢 Enterprise Deployment Examples

Scenario 1: Multiple Bots, Different Providers

# Bot A: Customer Support (Azure OpenAI)
PROVIDER=azure-openai → Uses gpt-4o for consistent responses

# Bot B: Code Assistant (DeepSeek)
PROVIDER=deepseek → Uses deepseek-reasoner for technical queries

# Bot C: Creative Writing (Claude)
PROVIDER=claude → Uses claude-3-5-sonnet for creative tasks

Scenario 2: Same Bot, Easy Provider Migration

# Week 1: Testing with OpenAI
PROVIDER=openai

# Week 2: Switch to Azure OpenAI for compliance
PROVIDER=azure-openai

# Week 3: Try DeepSeek for cost optimization
PROVIDER=deepseek

Zero code changes required! 🎉

🔍 Testing and Development

Build and Test

npm run build
npm test

Test with MCP Inspector

npm run build
npx @modelcontextprotocol/inspector node ./build/index.js

Environment File Support

Create a .env file in the project root:

# Provider Selection
PROVIDER=azure-openai

# API Configuration
AZURE_OPENAI_API_KEY=your-key-here
AZURE_OPENAI_BASE_URL=https://your-resource.openai.azure.com

🚀 Microsoft Copilot Studio Integration

This MCP server is specifically designed for Microsoft Copilot Studio custom connectors:

Benefits:

• Consistent Interface: Same tools across all providers
• No Provider Lock-in: Easy to switch providers
• Enterprise Ready: Built for production deployments
• Simplified Management: One configuration per bot

Implementation:

1. Deploy MCP Server with your chosen provider
2. Configure Copilot Studio to use MCP endpoints
3. Use Simple Tools - chat_completion and multi_turn_chat
4. Switch Providers by changing environment variables only

📚 API Reference

Tools

chat_completion

• Description: Single-turn chat completion
• Input: { message: string }
• Output: Text response from configured provider

multi_turn_chat

• Description: Multi-turn conversation with history
• Input: { messages: string }
• Output: Contextual response maintaining conversation flow

Resources

• models: List available models for the configured provider
• model-config: Configuration options for the current provider

🔒 Security

• API Keys: Stored as environment variables only
• Provider Isolation: Each server instance uses one provider
• No Credential Exposure: Client never sees API keys or provider details

🆘 Support

• GitHub Issues: Report bugs or request features
• GitHub Discussions: Community support

📄 License

MIT License - see file for details.

Built for the future of AI integration 🤖✨