Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
Important
This feature is in preview.
The Power BI MCP servers enable AI agents to interact with Power BI through natural language. Both the local MCP server and the remote MCP server implement the Model Context Protocol, each exposing specialized tools for different scenarios:
- Power BI MCP server (remote): Query data and generate insights from existing models
- Power BI MCP server (local): Build and modify semantic models programmatically
Choose the local server for development and model management workflows, or the remote server for data analysis and insights scenarios.
Remote vs. local MCP Server
| Feature | Remote MCP Server | Local MCP Server |
|---|---|---|
| Installation | No installation required | Requires Visual Studio Code or Node.js 20.0+ (npx) |
| Transport | Streamable HTTP | stdio |
| Authentication | Microsoft Entra ID (OAuth) | Microsoft Entra ID, Service Principal |
| Hosting | Fabric hosted service | Runs locally on your machine |
| Configuration | Minimal mcp.json |
Environment-specific setup |
| Tools | Query tools | Tools for metadata read/write, query, database operations |
| Status | Public preview | Public preview |
Power BI MCP server (remote)
The remote Power BI MCP server is a hosted endpoint that enables AI agents to query Power BI semantic models. It generates and executes DAX queries using Copilot's intelligence, enabling natural language conversations with your data.
Key capabilities
- Custom agent development - Build agents that answer user questions by querying semantic models
- Schema-aware querying - Agents automatically learn model structure to generate accurate queries
- Copilot-powered DAX generation - Leverage the same query generation engine used in Copilot for Power BI
- Flexible LLM integration - Use any LLM provider supported by your MCP client
Common scenarios
The remote MCP server is ideal for enhancing personal AI assistants like GitHub Copilot in VS Code. The server uses the authenticated user's permissions to execute queries, ensuring proper security and data access controls.
Example use cases:
- Ask GitHub Copilot about last quarter's sales trends from your semantic model
- Generate ad-hoc analyses through conversational queries
- Explore data relationships and patterns using natural language
Get started: Set up the remote MCP server
Power BI MCP server (local)
The local Power BI MCP server runs locally and provides AI agents with comprehensive semantic modeling capabilities. It enables developers and AI applications to interact with Power BI models through natural language, from simple property updates to complex agentic development workflows. The server runs on your machine and acts on the models in Power BI Desktop, Fabric workspaces or on Power BI Project files in so changes follow your existing source control and review workflows.
Key capabilities
- Natural language model editing - Create, update, and manage tables, columns, measures, and relationships across Power BI Desktop and Fabric semantic models using conversational commands
- Bulk operations at scale - Execute batch operations on hundreds of objects simultaneously with transaction support and error handling
- Best practice application - Evaluate and implement modeling best practices easily
- Agentic development workflows - Work with TMDL and Power BI Project files, enabling AI agents to autonomously plan and execute complex modeling tasks
- DAX query validation - Execute and validate DAX queries to test measures, troubleshoot calculations, and explore data
Common scenarios
The local MCP server is ideal for semantic model authors and developers who want to accelerate model development.
Example use cases:
- Create or update tables, columns, measures, and relationships through natural language prompts
- Apply modeling best practices across an existing semantic model in bulk
- Refactor TMDL or Power BI Project files as part of an agentic development workflow
- Author and validate DAX measures interactively while building a model
Get started: Set up the local MCP server
About the Model Context Protocol
The Model Context Protocol (MCP) is an open standard that defines how AI assistants interact with external tools and data sources in a structured, secure way.
MCP architecture components:
- Host - Applications that run MCP clients (for example, VS Code)
- Client - Components within a host that connect to MCP servers and consume their capabilities (for example, GitHub Copilot)
- Server - Local or remote programs that expose tools, resources, and prompts to clients
For example, when you use GitHub Copilot in VS Code with a Power BI MCP server, VS Code is the host, Copilot is the client, and Power BI provides the server.
Learn more: Model Context Protocol specification
Considerations
MCP as a phenomenon is very novel and cutting-edge. As with all new technology standards, consider doing a security review to ensure any systems that integrate with MCP servers follow all regulations and standards your system is expected to adhere to. This includes not only the Power BI MCP servers, but any MCP client/agent that you choose to implement down to the model provider.
You should follow Microsoft security guidance for MCP servers, including enabling Entra ID authentication, secure token management, and network isolation. Refer to Microsoft Security Documentation for details.
Next steps
Choose your path based on your scenario: