✨ Power BI Modeling MCP Server

The Power BI Modeling MCP Server implements the MCP specification to create a seamless connection between AI agents and Power BI semantic models. This project is in Public Preview and implementation may significantly change prior to our General Availability.

The Power BI Modeling MCP Server brings Power BI semantic modeling capabilities to your AI agents through a local MCP server. This allows developers and AI applications to interact with Power BI models in entirely new ways, from using natural language to execute modeling changes to autonomous AI agentic development workflows.

💡 What can you do?

• 🔄 Build and Modify Semantic Models with Natural Language - Tell your AI assistant what you need, and it uses this MCP server to create, update, and manage tables, columns, measures, relationships, and more... across Power BI Desktop and Fabric semantic models.

• ⚡ Bulk Operations at Scale - AI applications can execute batch modeling operations on hundreds of objects simultaneously — bulk renaming, bulk refactoring, model translations, or model security rules - with transaction support and error handling, turning hours of repetitive work into seconds.

• ✅ Apply modeling best practices - Easily evaluate and implement modeling best practices against your model.

• 🤖 Agentic Development Workflows - Supports working with TMDL and Power BI Project files, enabling AI agents to autonomously plan, create, and execute complex modeling tasks across your semantic model codebase.

• 🔍 Query and Validate DAX - AI assistants can execute and validate DAX queries against your model, helping you test measures, troubleshoot calculations, and explore your data

📹 Watch the video for an end-to-end demo.

[!WARNING]

• Use caution when connecting an AI Agent to a semantic model. The underlying LLM may produce unexpected or inaccurate results, which could lead to unintended changes. Always create a backup of your model before performing any operations.
• LLMs might unintentionally expose sensitive information from the semantic model, including data or metadata, in logs or responses. Exercise caution when sharing chat sessions. See Data Privacy and LLM Providers.
• The Power BI Modeling MCP server can only execute modeling operations. It cannot modify other types of Power BI metadata, such as report pages or semantic model elements like diagram layouts.
• The AI model you select directly influences the quality and relevance of the responses you receive. For the best results, choose a deep-reasoning model such as GPT-5 or Claude Sonnet 4.5. You can find more details about available models in the GitHub Copilot AI model comparison.

📦 Installation

The easiest way to install this MCP Server is by using the Visual Studio Code extension extension together with GitHub Copilot. However, you can also manually install it in any other MCP client.

Visual Studio Code (Recommended)

1. Install Visual Studio Code.

2. Install the GitHub Copilot Chat extension.

3. Install the Power BI Modeling MCP Visual Studio Code extension.

   

4. Open GitHub Copilot chat and confirm the powerbi-modeling-mcp is available and selected.

   

[!NOTE] If you do not see powerbi-modeling-mcp in the available tool list, verify that the MCP servers in Copilot option is enabled in Copilot settings on GitHub.com. For enterprise accounts, this option is disabled by default and must be enabled by an administrator.

Manual

This MCP Server can also be configured across other IDEs, CLIs, and MCP clients.

Node Package Executor (NPX) (requires Node.js)

Add the JSON configuration to your MCP client. Node will automatically download the MCP server from the @microsoft/powerbi-modeling-mcp npm package.

{
	"powerbi-modeling-mcp": {
			"type": "stdio",
			"command": "npx",
			"args": [
				"-y",
				"@microsoft/powerbi-modeling-mcp@latest",
				"--start"				
			]
		}	
}

Manual download

1. Download the VSIX package for the version you want using the URL below:
   • Template: https://marketplace.visualstudio.com/_apis/public/gallery/publishers/analysis-services/vsextensions/powerbi-modeling-mcp/[version]/vspackage?targetPlatform=[platform]
   • Example (version 0.1.9, platform win32-x64): https://marketplace.visualstudio.com/_apis/public/gallery/publishers/analysis-services/vsextensions/powerbi-modeling-mcp/0.1.9/vspackage?targetPlatform=win32-x64
2. Rename the downloaded .visx file to .zip
3. Unzip the contents to a folder of your choice, for example: C:\MCPServers\PowerBIModelingMCP
4. Run \extension\server\powerbi-modeling-mcp.exe
5. Copy the MCP JSON registration from the console and register it in your preferred MCP client tool.

Example of config that should work in most MCP clients:

{
	"powerbi-modeling-mcp": {
		"type": "stdio",
		"command": "C:\\MCPServers\\PowerBIModelingMCP\\extension\\server\\powerbi-modeling-mcp.exe",
		"args": [
			"--start"                
		],
		"env": {}			
	}	
}

🚀 Get started

First, you must connect to a Power BI semantic model, which can reside in Power BI Desktop, Fabric workspace or in Power BI Project (PBIP) files.

• For Power BI Desktop:

  hljs language-css

  Copy

  Connect to '[File Name]' in Power BI Desktop
  

• For Semantic Model in Fabric Workspace:

  hljs language-css

  Copy

  Connect to semantic model '[Semantic Model Name]' in Fabric Workspace '[Workspace Name]'
  

• For Power BI Project files:

  hljs language-sql

  Copy

  Open semantic model from PBIP folder '[Path to the definition/ TMDL folder in the PBIP]'
  

Once the connection is established, you can use natural language to ask the AI agent to make any modeling changes. To get started, try one of the following scenarios.

Example scenarios

[!TIP] The scenarios above are just examples. This MCP server equips your agents with modeling tools for any type of model change, and with the right prompt and context, you can automate virtually any modeling task.

Confirmation prompts

This MCP Server supports the Elicitation MCP protocol, requiring user approval for the following actions:

• Before the first modification made to a semantic model.
• Before the first query executed against a semantic model.

[!TIP] You can configure the MCP to skip these confirmations by using the --skipconfirmation option.

🛠️ Available tools

[!NOTE]

• This project is in Public Preview and tools may significantly change prior to our General Availability.
• You can ask AI to explain what tools are available and show examples of how to use them. For example: Tell me with some examples what I can do with powerbi-modeling-mcp

▶️ Available prompts

This MCP server includes built-in prompts to help you get started. In Visual Studio Code, access the available MCP prompts by typing the / command in the chat.

[!NOTE] Some prompts also attach resources that provide important context for the LLM.

⚙️ Settings

The MCP server supports several command line options and environment variables:

For Visual Studio Code, you can set the command line options and environment variables in the User Settings:

Open Visual Studio Code user settings and search for @ext:Microsoft.powerbi-modeling-mcp.

For manual installations, you can configure command-line options using the args property and set environment variables using the env property in the MCP Server registration JSON:

{
"servers": {
		"powerbi-modeling-mcp": {
			"command": "[Path To MCP Server folder]\\powerbi-modeling-mcp.exe",
			"args": [
				"--start"
                , "--skipconfirmation"
				, "--authmode=interactive"    				
			],
			"env": {
				"PBI_MODELING_MCP_ACCESS_TOKEN": "[ACCESS_TOKEN]"
			},
			"type": "stdio"
		}
	}
}

💬 Feedback and Support

• Check the Troubleshooting guide to diagnose and resolve common issues.
• We're building this in the open. Your feedback is much appreciated, and will help us shape the future of the Power BI Modeling MCP server.
  • 👉 Open an issue in the public GitHub repository - we’d love to hear from you!

Considerations and limitations

• This MCP server follows the same rules and behaviors as modeling operations performed by External Tools. Refer to the documentation for more information.

Data Privacy and LLM Providers

The MCP server runs locally and connects to your semantic model using your existing credentials and permissions — it does not bypass any Power BI security controls. However, data retrieved by the MCP server (metadata, schemas, or query results) is sent to the MCP client (e.g., VS Code) and may then be forwarded to the configured LLM provider as part of the conversation context.

In short: AI assistance does not expand data access, but it may transmit accessed data to the LLM provider. Governance efforts should focus on your organization's AI data-handling policies and the LLM provider's terms of service, rather than on Power BI security controls alone.

If you use GitHub Copilot, refer to the GitHub Copilot model hosting documentation for details on how data is shared with external AI model providers.

Security

Your credentials are always handled securely through the official Azure Identity SDK - we never store or manage tokens directly.

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 Modeling MCP Server, 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.

Permissions and Risk

MCP clients can invoke operations based on the user’s Fabric RBAC permissions. Autonomous or misconfigured clients may perform destructive actions. You should review and apply least-privilege RBAC roles and implement safeguards before deployment. Certain safeguards, such as flags to prevent destructive operations, are not standardized in the MCP specification and may not be supported by all clients.

Data Collection

The software may collect information about you and your use of the software and send it to Microsoft. Microsoft may use this information to provide services and improve our products and services. There are also some features in the software that may enable you and Microsoft to collect data from users of your applications. If you use these features, you must comply with applicable law, including providing appropriate notices to users of your applications together with a copy of Microsoft's privacy statement. You can learn more about data collection and use in the help documentation and our privacy statement. Your use of the software operates as your consent to these practices.

Compliance Responsibility

This MCP server may be installed, used and share data with third party clients and services, such as third party LLMs that operate outside Microsoft compliance boundaries. You are responsible for ensuring that any integration complies with applicable organizational, regulatory, and contractual requirements.

Third Party Components

This MCP server may use or depend on third party components.  You are responsible for reviewing and complying with the licenses and security posture of any third-party components.

Export Control

Use of this software must comply with all applicable export laws and regulations, including U.S. Export Administration Regulations and local jurisdiction requirements.

No Warranty / Limitation of Liability

This software is provided “as is” without warranties or conditions of any kind, either express or implied. Microsoft shall not be liable for any damages arising from use, misuse, or misconfiguration of this software.

Code of Conduct

This project has adopted the Microsoft Open Source Code of Conduct. For more information, see the Code of Conduct FAQ or contact open@microsoft.com with any additional questions or comments.