ModelPilot MCP Server

Model Context Protocol server for ModelPilot - enables AI coding assistants to integrate ModelPilot directly into developer workflows.

Features

• 🔄 Code Migration - Convert OpenAI/Anthropic code to ModelPilot
• 🎯 Router Management - Create and configure AI model routers
• 🧪 Live Testing - Test routers with real requests and see cost breakdowns
• 📊 Usage Analytics - Track spending and get optimization tips
• 💰 Cost Optimization - Get recommendations to reduce AI costs
• 🔍 Model Comparison - Compare models on cost, quality, and speed
• 💳 Balance Checking - View credits and subscription status
• 💻 Code Generation - Generate production-ready integration code

Installation

npm install -g @modelpilot/mcp-server

Usage

1. Get your ModelPilot API Key

Sign up at modelpilot.app and create an API key in the dashboard.

2. Configure in your AI Assistant

For Cascade / Windsurf

Add to your MCP settings:

{
  "mcpServers": {
    "modelpilot": {
      "command": "modelpilot-mcp",
      "env": {
        "MODELPILOT_API_KEY": "mp-your-api-key-here"
      }
    }
  }
}

For Cline

Add to ~/.cline/mcp.json:

{
  "mcpServers": {
    "modelpilot": {
      "command": "npx",
      "args": ["-y", "@modelpilot/mcp-server"],
      "env": {
        "MODELPILOT_API_KEY": "mp-your-api-key-here"
      }
    }
  }
}

3. Start Using!

Talk to your AI assistant:

"Migrate this OpenAI code to ModelPilot"
"Create a cost-optimized router for my chatbot"
"Test my router with this prompt and show me the cost"
"How can I reduce my AI costs?"

Available Tools

migrate_code

Convert OpenAI or Anthropic SDK code to ModelPilot. Shows converted code and estimated cost savings.

Example:

User: "Migrate this code to ModelPilot: [paste OpenAI code]"
AI: [Shows converted code] "This will save you ~45% on costs"

create_router

Create a new AI model router with guided configuration.

Example:

User: "Create a cost-optimized router for my chatbot"
AI: "What models would you like to include?"
User: "GPT-4o-mini and Claude 3.5 Sonnet"
AI: [Creates router] "Router created! Here's your sample code..."

test_request

Test a router with a real request. Returns response, cost, latency, and model used.

Example:

User: "Test my router 'chatbot-123' with: Hello, how are you?"
AI: [Shows response] "Cost: $0.0008 | Model: gpt-4o-mini | Latency: 450ms"

get_usage_summary

View usage statistics and costs for the last 24h, 7d, or 30d.

Example:

User: "Show my usage for the last 7 days"
AI: "Total: $12.45 | 1,234 requests | Top model: GPT-4o-mini (65%)"

optimize_router

Get cost-saving recommendations for a router.

Example:

User: "How can I optimize my chatbot router?"
AI: "Switch to GPT-4-Turbo → Save $45/month | Enable cost-optimized routing"

compare_models

Compare models side-by-side on cost, quality, and speed.

Example:

User: "Compare GPT-4o vs Claude 3.5 Sonnet vs Gemini Pro"
AI: [Shows comparison table] "Gemini Pro is cheapest, Claude 3.5 Sonnet has best quality"

get_balance

Check current credit balance and subscription status.

Example:

User: "What's my balance?"
AI: "Balance: $45.67 | Subscription: Pro | This month: $12.34 spent"

generate_integration_code

Generate production-ready code in Python, JavaScript, or TypeScript.

Example:

User: "Generate Python code for my router with error handling and retries"
AI: [Generates complete code with retries, error handling, and examples]

Documentation

The MCP server includes 10 comprehensive documentation resources:

• Quickstart Guide - Get started in 5 minutes
• OpenAI Migration - Complete migration guide from OpenAI
• Anthropic Migration - Complete migration guide from Anthropic
• Router Configuration - Best practices for router setup
• Python SDK - Full Python SDK reference
• JavaScript SDK - Full JavaScript SDK reference
• Routing Strategies - Understanding cost/quality/speed optimization
• Cost Optimization - Top strategies for reducing costs
• Chatbot Example - Production chatbot implementation
• Pricing Guide - Understanding billing and pricing

Access via:

User: "Show me the OpenAI migration guide"
AI: [Displays modelpilot://migration/openai]

Development

Setup

cd modelpilot-mcp
npm install

Development Mode

npm run dev

Build

npm run build

Test Locally

MODELPILOT_API_KEY=mp-your-key npm start

Requirements

• Node.js 18+
• ModelPilot API key

Support

• Documentation: docs.modelpilot.app
• Dashboard: modelpilot.app/dashboard
• Support: [email protected]

License

MIT