# Power BI Modeling MCP Server [Official] [Health: Active]

**Category:** 📊 Data Visualization  
**Repository:** https://github.com/microsoft/powerbi-modeling-mcp.git  
**GitHub Stars:** 1276  
**npm Downloads (last month):** 41832  
**Views:** 14  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/power-bi-modeling-mcp-server

## Description
The Power BI Modeling MCP Server brings Power BI semantic modeling capabilities to your AI agents.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "power-bi-modeling-mcp-server": {
    "command": "npx",
    "args": ["-y","@microsoft/powerbi-modeling-mcp@latest"]
  }
}
```

## Documentation

## What Power BI Modeling MCP Server does

Power BI Modeling MCP Server exposes Power BI semantic modeling operations to MCP-compatible AI agents. After connecting to a model, an agent can create, update, and manage tables, columns, measures, relationships, and security rules. It also supports executing and validating DAX queries, which can help test measures, investigate calculations, and inspect model behavior.

The server is intended for authoring rather than general Power BI metadata management. Its supported scope covers semantic model operations, including models associated with Power BI Desktop, Fabric workspaces, and Power BI Project files using TMDL. It does not handle report pages or diagram layouts.

## How it works

The local server communicates with an MCP client over stdio. An agent first establishes a connection to one of three model sources:

- A semantic model open in Power BI Desktop
- A semantic model in a Fabric workspace
- A semantic model definition inside a Power BI Project folder

Once connected, prompts can describe modeling changes in natural language. Operations can also be applied in batches across many objects, supporting tasks such as renaming, refactoring, translations, and model security updates. Transaction support and error handling are available for bulk modeling operations.

TMDL and PBIP support enables agent workflows that plan and apply changes against semantic model files. The server also includes an `accept_eula` tool for interactive license acceptance.

## Setup and configuration

The recommended setup uses Visual Studio Code with the GitHub Copilot Chat extension and the Power BI Authoring MCP extension. After installation, open Copilot Chat and ensure `powerbi-modeling-mcp` is enabled in the available tools. GitHub Copilot MCP server access may need to be enabled in Copilot settings; enterprise accounts may require an administrator to enable it.

For other clients, Node.js can launch the package through `npx`:

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

Review and accept the Power BI Authoring MCP Server EULA before using the npm package. Interactive sessions can authorize the `accept_eula` tool. Unattended setups can add `--accepteula` to the arguments or set `PBI_MODELING_MCP_ACCEPT_EULA` to `true` after accepting the terms.

A VSIX package can also be downloaded, extracted, and launched with its bundled `powerbi-modeling-mcp.exe`, then registered as a stdio server in the chosen client.

## Tools and capabilities

- Create, update, and manage semantic model tables, columns, measures, and relationships
- Apply bulk renames, refactoring, translations, and security-rule changes
- Execute and validate DAX queries against a connected model
- Evaluate and apply modeling best practices
- Work with TMDL and Power BI Project model files
- Connect to Power BI Desktop and Fabric semantic models
- Accept the EULA interactively through the `accept_eula` tool

## Limitations and notes

Back up a model before allowing an AI agent to make changes. Generated operations may be inaccurate or unexpected, and model data or metadata could appear in logs or responses. Take care when sharing chat sessions or connecting models that contain sensitive information.

The server is limited to modeling operations. Report pages and diagram layouts are outside its supported scope. Agent response quality also depends on the selected AI model; the project recommends a deep-reasoning model for these workflows.

_Full upstream README: https://allmcps.com/mcp/power-bi-modeling-mcp-server/readme_