How to Edit Fabric Semantic Models with Power BI Desktop (Connect vs Edit Explained)

Author: Manuel Quintana, Pragmatic Works Editing and maintaining semantic models in Microsoft Fabric is a common task for Power BI professionals, but knowing how and where to make those changes can have a significant impact on scalability, governance, and collaboration. In this training session, Manuel Quintana from Pragmatic Works walks through the key considerations when editing Fabric semantic models, with a clear focus on the differences between using Connect versus Edit in Power BI Desktop.

While Microsoft Fabric provides a powerful, browser‑based modeling experience, there are scenarios where Power BI Desktop offers advantages, especially for advanced modeling, centralized measure management, and preview features. This blog summarizes the main lessons from the video and helps clarify which approach is best for different use cases.

Creating a Starting Semantic Model in Fabric

Manuel begins by demonstrating how easy it is to create a semantic model in Fabric using a Lakehouse. For demonstration purposes, he uses the built‑in sample Lakehouse available to all Fabric users. Once the workspace is Fabric‑enabled, a semantic model can be created directly from the Lakehouse by selecting tables and applying basic naming conventions.

At this stage, the goal is not deep modeling but simply having an existing semantic model that can later be updated or enhanced. Once created, the model opens in the browser and supports common modeling actions such as:

• Defining relationships
• Creating DAX measures
• Hiding columns from report view
• Adjusting formatting for reporting

However, Fabric semantic models currently do not support calculated columns in the browser, which leads to the need for Power BI Desktop in certain scenarios.

Using Power BI Desktop: Connect vs Edit

One of the most important takeaways from this session is the distinction between the Connect and Edit options when working with Fabric semantic models in Power BI Desktop.

Connect: Report‑Only Changes

When using the Connect option, Power BI Desktop connects to the semantic model in Direct Lake mode for the purpose of building reports. This allows report creators to:

• Create visuals
• Add report‑specific measures
• Format measures for display

However, any measures created using Connect exist only within that specific report. They are not saved back to the semantic model in the Fabric service. As Manuel demonstrates, a measure such as Total Sales may appear in the report but will not be visible to other users who connect to the same semantic model.

This approach is ideal when:

• Measures are highly specific to a single report
• No centralized reuse is required
• The goal is visualization, not model maintenance

Edit: Centralized Model Management

For updating and maintaining the semantic model itself, Manuel strongly recommends using the Edit option in Power BI Desktop. When selecting Edit, Power BI Desktop opens directly into the modeling experience with no report or data view.

In this mode:

• Changes are applied directly to the semantic model
• All updates automatically sync to the Fabric service
• Measures become available to all report creators

This live editing experience ensures that commonly used measures, relationships, and model changes are centralized and reusable across the organization.

Why Desktop Editing Can Be Safer and Faster

Manuel also highlights practical advantages of editing semantic models in Power BI Desktop. While the Fabric browser experience is powerful, browser refreshes or accidental reloads can result in lost DAX expressions. Power BI Desktop provides a more stable environment where work remains visible even during connectivity interruptions.

Additionally, many users find the desktop interface more responsive and familiar, especially for complex modeling tasks.

Unlocking Preview Features and Advanced Capabilities

Another key reason to use Power BI Desktop is access to preview features that may not yet be available in the Fabric service. Manuel specifically references user‑defined DAX functions, which are currently available only through Power BI Desktop when enabled as a preview feature.

Through Desktop editing, users can:

• Create user‑defined functions
• Leverage advanced modeling tools
• Deploy preview features directly to Fabric semantic models

This makes Power BI Desktop the preferred option when working with cutting‑edge functionality or advanced modeling scenarios.

Final Recommendations

Manuel concludes by emphasizing the importance of choosing the right approach based on intent:

• Use Connect for report‑specific visuals and calculations
• Use Edit for centralized measures, relationships, and model governance

Understanding this distinction helps teams build scalable, maintainable semantic models while still supporting flexible reporting workflows. By combining Fabric’s browser experience with Power BI Desktop’s advanced modeling capabilities, organizations can get the best of both worlds.

For more in‑depth training on Microsoft Fabric, Power BI, and semantic modeling best practices, explore Pragmatic Works’ on‑demand learning platform and YouTube channel.

Don't forget to check out the Pragmatic Works' on-demand learning platform for more insightful content and training sessions on Power BI and other Microsoft applications. Be sure to subscribe to the Pragmatic Works YouTube channel to stay up-to-date on the latest tips and tricks.