Azure Pricing MCP Server 💰

A Model Context Protocol (MCP) server that provides AI assistants with real-time access to Azure retail pricing information. Query VM prices, compare costs across regions, estimate monthly bills, and discover available SKUs—all through natural language.

🚀 Quick Start

# 1. Clone the repository
git clone https://github.com/msftnadavbh/AzurePricingMCP.git
cd AzurePricingMCP

# 2. Set up virtual environment
python -m venv .venv
source .venv/bin/activate  # Linux/Mac
# .venv\Scripts\activate   # Windows

# 3. Install dependencies
pip install -r requirements.txt

# 4. Test the server
python -m azure_pricing_mcp

Then configure your AI assistant (VS Code, Claude Desktop, etc.) to use the MCP server.

✨ Features

Feature	Description
🔍 Price Search	Search Azure prices with filters (service, region, SKU, price type)
⚖️ Price Comparison	Compare costs across regions or between different SKUs
💡 Cost Estimation	Calculate monthly/yearly costs based on usage hours
🎫 Reserved Instances	Compare 1-year and 3-year RI pricing with break-even analysis
💰 Savings Plans	View 1-year and 3-year savings plan pricing
🎯 Smart SKU Discovery	Fuzzy matching for service names ("vm" → "Virtual Machines")
🌍 Region Recommendations	Find the cheapest Azure regions for any SKU with savings analysis
💱 Multi-Currency	Support for USD, EUR, GBP, and more
📊 Real-time Data	Live data from Azure Retail Prices API
🏷️ Customer Discounts	Apply discount percentages to all pricing queries
⚠️ Retirement Warnings	Automatic alerts for retiring/retired VM SKUs with migration guidance
🐳 Docker Support	Run in containers for easy deployment and isolation

🛠️ Available Tools

Tool	Description
`azure_price_search`	Search Azure retail prices with flexible filtering
`azure_price_compare`	Compare prices across regions or SKUs
`azure_ri_pricing`	Get Reserved Instance pricing and savings analysis
`azure_cost_estimate`	Estimate costs based on usage patterns
`azure_region_recommend`	Find cheapest regions for a SKU with savings percentages
`azure_discover_skus`	List available SKUs for a specific service
`azure_sku_discovery`	Intelligent SKU discovery with fuzzy name matching
`get_customer_discount`	Get customer discount information

📋 Installation

📝 New to setup? Check out INSTALL.md for detailed instructions or SETUP_CHECKLIST.md for a step-by-step checklist!
🐳 Prefer Docker? See DOCKER.md for containerized deployment!

Prerequisites

Python 3.10+ (or Docker for containerized deployment)
pip (Python package manager)

Option 1: Docker (Easiest)

docker build -t azure-pricing-mcp .
docker run -i azure-pricing-mcp

Option 2: Automated Setup

# Windows PowerShell
.\scripts\setup.ps1

# Linux/Mac/Cross-platform
python scripts/install.py

Option 3: Manual Setup

# Clone repository
git clone https://github.com/msftnadavbh/AzurePricingMCP.git
cd AzurePricingMCP

# Create virtual environment
python -m venv .venv

# Activate virtual environment
source .venv/bin/activate    # Linux/Mac
.venv\Scripts\activate       # Windows

# Install dependencies
pip install -r requirements.txt

🖥️ VS Code Integration

Step 1: Install GitHub Copilot

Ensure you have the GitHub Copilot extension installed.

Step 2: Configure MCP Server

Create or edit .vscode/mcp.json in your workspace:

Option A: Using Python Virtual Environment

{
  "servers": {
    "azure-pricing": {
      "type": "stdio",
      "command": "/absolute/path/to/AzurePricingMCP/.venv/bin/python",
      "args": ["-m", "azure_pricing_mcp"]
    }
  }
}

Windows users: Use the full path with forward slashes or escaped backslashes:
"command": "C:/Users/YourUsername/Projects/AzurePricingMCP/.venv/Scripts/python.exe"

Option B: Using Docker (stdio) 🐳

{
  "servers": {
    "azure-pricing": {
      "type": "stdio",
      "command": "docker",
      "args": ["run", "-i", "--rm", "azure-pricing-mcp:latest"]
    }
  }
}

Option C: Using Docker (SSE - Server-Sent Events) 🐳

# First, build and run the container with port mapping
docker build -t azure-pricing-mcp .
docker run -d -p 8080:8080 --name azure-pricing azure-pricing-mcp

Then configure .vscode/mcp.json:

{
  "servers": {
    "azure-pricing": {
      "type": "sse",
      "url": "http://localhost:8080/sse"
    }
  }
}

💡 SSE Benefits: Better isolation through Docker, allows multiple clients to connect to the same server instance, and easier to debug with HTTP endpoints.

Step 3: Restart MCP Server

Open Command Palette (Ctrl+Shift+P / Cmd+Shift+P)
Run: MCP: List Servers
Click the refresh/restart button next to azure-pricing

Step 4: Use in Copilot Chat

Open Copilot Chat and ask:

What's the price of Standard_D32s_v6 in East US 2?

You'll see the MCP tools being invoked with real Azure pricing data!

🤖 Claude Desktop Integration

Add to your Claude Desktop configuration file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json

Option A: Using Python

{
  "mcpServers": {
    "azure-pricing": {
      "command": "python",
      "args": ["-m", "azure_pricing_mcp"],
      "cwd": "/path/to/AzurePricingMCP"
    }
  }
}

Option B: Using Docker 🐳

{
  "mcpServers": {
    "azure-pricing": {
      "command": "docker",
      "args": ["run", "-i", "--rm", "azure-pricing-mcp:latest"]
    }
  }
}

💬 Example Queries

Once configured, ask your AI assistant:

Query Type	Example
Basic Pricing	"What's the price of a D4s_v3 VM in West US 2?"
Multi-Node	"Price for 20 Standard_D32s_v6 nodes in East US 2"
Comparison	"Compare VM prices between East US and West Europe"
Cost Estimate	"Estimate monthly cost for D8s_v5 running 12 hours/day"
SKU Discovery	"What App Service plans are available?"
Savings Plans	"Show savings plan options for virtual machines"
Storage	"What are the blob storage pricing tiers?"

Sample Response

Standard_D32s_v6 in East US 2:
- Linux On-Demand: $1.613/hour → $23,550/month for 20 nodes
- 1-Year Savings:  $1.113/hour → $16,250/month (31% savings)
- 3-Year Savings:  $0.742/hour → $10,833/month (54% savings)

⚠️ Retirement Status Warnings

The server automatically checks VM SKUs against Microsoft's official retirement documentation and warns you when querying SKUs that are:

Status	Icon	Description
Retirement Announced	⚠️	SKU has a planned retirement date - migrate soon
Retired	🚫	SKU is no longer available for new deployments
Previous Generation	ℹ️	Newer versions available - consider upgrading

Example Warning Output

⚠️ RETIREMENT WARNING: Lsv2-series
   Status: Retirement Announced
   Retirement Date: 11/15/28
   Recommendation: Migrate to Lsv3, Lasv3, Lsv4, or Lasv4 series
   Migration Guide: https://learn.microsoft.com/azure/virtual-machines/...

ℹ️ PREVIOUS GENERATION: Ev3-series
   Status: Newer versions available
   Recommendation: Consider upgrading to Ev5 or Ev6 series

The retirement data is fetched dynamically from Microsoft's official documentation and cached for 24 hours.

🧪 Testing

Verify Installation

# Run the server directly (should start without errors)
python -m azure_pricing_mcp

# Run tests
pytest tests/

Test MCP Connection in VS Code

Open Command Palette → MCP: List Servers
Verify azure-pricing shows 8 tools
Open Copilot Chat and ask a pricing question

🤝 Contributing

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

Development Setup

# Fork and clone the repository
git clone https://github.com/YOUR_USERNAME/AzurePricingMCP.git
cd AzurePricingMCP

# Create development environment
python -m venv .venv
source .venv/bin/activate

# Install dependencies
pip install -e ".[dev]"

# Make your changes
# ...

# Test your changes
pytest tests/

Contribution Guidelines

Fork the repository
Create a branch for your feature (git checkout -b feature/amazing-feature)
Commit your changes (git commit -m 'Add amazing feature')
Push to your branch (git push origin feature/amazing-feature)
Open a Pull Request

Code Style

Follow PEP 8 guidelines
Add type hints for function parameters and return values
Include docstrings for public functions
Run black, ruff, and mypy before submitting
Test your changes with pytest tests/

📁 Project Structure

AzurePricingMCP/
├── src/
│   └── azure_pricing_mcp/
│       ├── __init__.py       # Package initialization
│       ├── __main__.py       # Module entry point
│       ├── server.py         # Main MCP server implementation
│       ├── handlers.py       # Tool handlers
│       ├── client.py         # Azure Pricing API client
│       ├── config.py         # Configuration constants
│       ├── formatters.py     # Response formatting
│       ├── models.py         # Data models
│       ├── tools.py          # Tool definitions
│       └── services/         # Business logic services
│           ├── pricing.py    # Pricing operations
│           ├── retirement.py # VM retirement tracking
│           └── sku.py        # SKU discovery
├── scripts/
│   ├── install.py            # Installation script
│   ├── setup.ps1             # PowerShell setup script
│   └── run_server.py         # Server runner
├── tests/                    # Test suite
├── docs/                     # Additional documentation
├── requirements.txt          # Python dependencies
├── pyproject.toml            # Package configuration
├── README.md                 # This file
├── QUICK_START.md            # Quick start guide
└── INSTALL.md                # Detailed installation guide

🔌 API Reference

This server uses the Azure Retail Prices API:

https://prices.azure.com/api/retail/prices

No authentication required - The Azure Retail Prices API is publicly accessible.

📚 Additional Documentation

QUICK_START.md - Step-by-step setup guide
INSTALL.md - Detailed installation instructions
DOCKER.md - Docker containerization guide 🐳
USAGE_EXAMPLES.md - Detailed usage examples and API responses
SETUP_CHECKLIST.md - Installation verification checklist

⚠️ Troubleshooting

Tools not appearing in VS Code

Check Python syntax: Run python -m azure_pricing_mcp to check for errors
Verify path: Use absolute paths in .vscode/mcp.json
Restart server: Command Palette → MCP: List Servers → Restart

"No module named 'mcp'"

# Ensure you're in the virtual environment
source .venv/bin/activate
pip install mcp>=1.0.0

Connection errors

Check your internet connection
The Azure Pricing API may rate-limit requests (automatic retry is built-in)

📄 License

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

🙏 Acknowledgments

Original Author: @charris-msft
Current Maintainer + Version 2.3: @msftnadavbh
Contributors:
- @notoriousmic - Testing infrastructure and best practices
Model Context Protocol - The protocol that makes this possible
Azure Retail Prices API - Microsoft's public pricing API
All open-source contributors

📬 Support

Issues: GitHub Issues
Discussions: GitHub Discussions

Made with ❤️ for the Azure community