Power BI MCP Server

Power BI MCP Server 🚀

🎥 Live Demo

Power BI MCP Server Demo

Transform your Power BI experience - ask questions in natural language and get instant insights from your data.

A Model Context Protocol (MCP) server that enables AI assistants to interact with Power BI datasets through natural language. Query your data, generate DAX, and get insights without leaving your AI assistant.

✨ Features

🔗 Direct Power BI Connection - Connect to any Power BI dataset via XMLA endpoints
💬 Natural Language Queries - Ask questions in plain English, get DAX queries and results
📊 Automatic DAX Generation - AI-powered DAX query generation using GPT-4o-mini
🔍 Table Discovery - Automatically explore tables, columns, and measures
⚡ Optimized Performance - Async operations and intelligent caching
🛡️ Secure Authentication - Service Principal authentication with Azure AD
📈 Smart Suggestions - Get relevant question suggestions based on your data

🎥 Demo

Power BI MCP Demo

Ask questions like "What are total sales by region?" and get instant insights from your Power BI data.

🚀 Quick Start

Prerequisites

Python 3.8 or higher
Windows with ADOMD.NET or Docker on Linux (container includes the runtime)
SQL Server Management Studio (SSMS) or ADOMD.NET client libraries (Windows only)
Power BI Pro/Premium with XMLA endpoint enabled
Azure AD Service Principal with access to your Power BI dataset
OpenAI API key (optional for natural language features)

Installation

Clone the repository
git clone https://github.com/yourusername/powerbi-mcp-server.git cd powerbi-mcp-server
Install dependencies
pip install -r requirements.txt
Configure environment variables
cp .env.example .env # Edit .env with your credentials
Test the connection
python quickstart.py

Configure with Claude Desktop

Add to your Claude Desktop configuration file:

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

{
  "mcpServers": {
    "powerbi": {
      "command": "python",
      "args": ["C:/path/to/powerbi-mcp-server/src/server.py"],
      "env": {
        "PYTHONPATH": "C:/path/to/powerbi-mcp-server",
        "OPENAI_API_KEY": "your-openai-api-key"
      }
    }
  }
}

Docker

Build the container image:

docker build -t powerbi-mcp .

Run the server:

docker run -it --rm -e OPENAI_API_KEY=<key> powerbi-mcp

The container listens on port 8000 by default. Override the host or port using environment variables or command-line arguments:

docker run -it --rm -e OPENAI_API_KEY=<key> -p 7000:7000 powerbi-mcp \
  python src/server.py --host 0.0.0.0 --port 7000

The server exposes a Server-Sent Events endpoint at /sse. Clients should connect to this endpoint and then POST JSON-RPC messages to the path provided in the initial endpoint event (typically /messages/).

The container includes the .NET runtime required by pythonnet and pyadomd. It sets PYTHONNET_RUNTIME=coreclr and DOTNET_ROOT=/usr/share/dotnet so the .NET runtime is detected automatically. Environment variables mirror those in .env.example; pass them with -e VAR=value or provide a .env file in the build context.

📖 Usage

Once configured, you can interact with your Power BI data through Claude:

Connect to Your Dataset

Connect to Power BI dataset at powerbi://api.powerbi.com/v1.0/myorg/YourWorkspace

Explore Your Data

What tables are available?
Show me the structure of the Sales table

Ask Questions

What are the total sales by product category?
Show me the trend of revenue over the last 12 months
Which store has the highest gross margin?

Execute Custom DAX

Execute DAX: EVALUATE SUMMARIZE(Sales, Product[Category], "Total", SUM(Sales[Amount]))

🔧 Configuration

Required Credentials

Power BI XMLA Endpoint
- Format: powerbi://api.powerbi.com/v1.0/myorg/WorkspaceName
- Enable in Power BI Admin Portal → Workspace Settings
Azure AD Service Principal
- Create in Azure Portal → App Registrations
- Grant access in Power BI Workspace → Access settings
OpenAI API Key (optional)
- Needed only for natural language features
- Endpoints that rely on GPT models are hidden if this key is not set
- Get from OpenAI Platform
- Model used: gpt-4o-mini (200x cheaper than GPT-4)

Environment Variables

Create a .env file (OpenAI settings are optional):

# OpenAI Configuration (optional)
OPENAI_API_KEY=your_openai_api_key_here
OPENAI_MODEL=gpt-4o-mini  # Defaults to gpt-4o-mini

# Optional: Default Power BI Credentials
# These values are used when the `connect_powerbi` action does not supply
# tenant_id, client_id or client_secret.
DEFAULT_TENANT_ID=your_tenant_id
DEFAULT_CLIENT_ID=your_client_id
DEFAULT_CLIENT_SECRET=your_client_secret

# Logging
LOG_LEVEL=INFO

🏗️ Architecture

powerbi-mcp-server/
├── src/
│   └── server.py          # Main MCP server implementation
├── docs/                  # Documentation
├── examples/              # Example queries and use cases
├── tests/                 # Test suite
├── .env.example          # Environment variables template
├── requirements.txt      # Python dependencies
├── quickstart.py        # Quick test script
└── README.md           # This file

Key Components

PowerBIConnector - Handles XMLA connections and DAX execution
DataAnalyzer - AI-powered query generation and interpretation
PowerBIMCPServer - MCP protocol implementation

🔐 Security Best Practices

Never commit credentials - Use .env files and keep them in .gitignore
Use Service Principals - Avoid personal credentials
Minimal permissions - Grant only necessary access to datasets
Rotate secrets regularly - Update Service Principal secrets periodically
Use secure connections - Always use HTTPS/TLS

🧪 Testing

Run the test suite:

python -m pytest tests/

Test specific functionality:

python tests/test_connection.py
python tests/test_dax_generation.py

🤝 Contributing

We welcome contributions! Please see CONTRIBUTING.md for details.

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

📊 Performance

Connection time: 2-3 seconds
Query execution: 1-5 seconds depending on complexity
Token usage: ~500-2000 tokens per query with GPT-4o-mini
Cost: ~$0.02-0.06 per day for typical usage

🐛 Troubleshooting

Common Issues

ADOMD.NET not found
- For Windows, install SQL Server Management Studio (SSMS)
- On Linux, use the provided Docker image which bundles the cross-platform ADOMD.NET runtime
Connection fails
- Verify XMLA endpoint is enabled in Power BI
- Check Service Principal has workspace access
- Ensure dataset name matches exactly
Timeout errors
- Increase timeout in Claude Desktop config
- Check network connectivity to Power BI

See TROUBLESHOOTING.md for detailed solutions.

📝 License

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

🙏 Acknowledgments

Anthropic for the MCP specification
Microsoft for Power BI and ADOMD.NET
OpenAI for GPT models
The MCP community for inspiration and support

📬 Support

📧 Email: sulaimanahmed013@gmail.com
💬 Issues: GitHub Issues
📚 Docs: Full Documentation

This server cannot be installed

security - not tested

license - permissive license

quality - not tested

How are these scores calculated?

hybrid server

The server is able to function both locally and remotely, depending on the configuration or use case.

Enables AI assistants to interact with Power BI datasets through natural language, allowing users to query data, generate DAX, and get insights without leaving their AI assistant.

Related MCP Servers

MySQL MCP Server
designcomputer
A
security
A
license
A
quality
Allows AI assistants to list tables, read data, and execute SQL queries through a controlled interface, making database exploration and analysis safer and more structured.
Last updated -
1
732
Python
MIT License
Metabase MCP Server
imlewc
-
security
F
license
-
quality
Enables AI assistants to interact with Metabase databases and dashboards, allowing users to list and execute queries, access data visualizations, and interact with database resources through natural language.
Last updated -
56
JavaScript
adx-mcp-server
pab1it0
A
security
A
license
A
quality
AI assistants to query and analyze Azure Data Explorer databases through standardized interfaces.
Last updated -
4
45
Python
MIT License
Metabase Server MCP
berry-street
-
security
F
license
-
quality
Enables AI assistants to interact with Metabase, providing access to dashboards, questions, databases, and tools for executing queries and viewing data through natural language.
Last updated -
JavaScript

View all related MCP servers

Power BI MCP Server