Power BI XMLA Endpoint Update

Microsoft has announced a significant change for all Fabric capacities: starting 9 June 2025, all capacity SKUs will support XMLA read/write operations by default. This update removes administrative hurdles for semantic model owners and co-authors but comes with important implications—particularly that models modified via XMLA endpoints can no longer be downloaded to Power BI Desktop. This article explores what the XMLA endpoint enables, its use cases, compatible third-party tools, and critical considerations for Power BI professionals.

Understanding the XMLA Endpoint: A Gateway to Advanced Capabilities

The XML for Analysis (XMLA) endpoint serves as a communication protocol between client applications and the Microsoft Analysis Services engine that powers semantic models in Power BI. This protocol—the same one used in SQL Server Analysis Services and Azure Analysis Services—opens Power BI semantic models to a broader ecosystem of tools and capabilities. At its core, the XMLA endpoint allows Power BI semantic models to be accessed, queried, and modified by external applications. Previously, capacities defaulted to read-only mode, limiting what semantic model owners could accomplish without administrator intervention. The upcoming change shifts the default to read/write mode, empowering data professionals to work more effectively with their models.

Key Benefits and Use Cases

The XMLA endpoint unlocks several powerful capabilities for Power BI professionals:

• Advanced Semantic Modelling: Beyond what's possible in Power BI Desktop, XMLA enables fine-grained control over model properties, complex calculations, and metadata management. Data professionals can implement sophisticated modelling patterns typically reserved for enterprise-grade Analysis Services deployments.
• Streamlined DevOps and Automation: With write access through the XMLA endpoint, organisations can automate semantic model deployments across development, testing, and production environments. This capability is crucial for maintaining consistency and implementing proper change management processes.
• Enhanced Refresh Control: The XMLA endpoint removes the 48-refreshes-per-day limitation and timeout constraints of the standard Power BI refresh mechanism. It also provides granular control over incremental refresh partitions, allowing selective updates of historical data without reloading everything.
• Co-authoring and Collaboration: Multiple professionals can collaborate on complex semantic models using specialised tools, improving team productivity and enabling division of labour for large modelling projects.
• Diagnostics and Optimisation: XMLA connectivity provides access to detailed diagnostic information through Dynamic Management Views (DMVs) and SQL Server Profiler. This visibility helps identify performance bottlenecks and optimise complex semantic models.

Third-Party Tools That Leverage XMLA

The true power of XMLA connectivity comes through the ecosystem of tools it supports. Here are the most significant ones:

• Tabular Editor: This lightweight, open-source tool has become a favourite among data modellers for its intuitive interface and powerful features for working with tabular models. It excels at bulk operations, display folder management, and DAX editing that would be laborious in Power BI Desktop. It comes in 2 flavours; version 2 which is free and the more fully featured paid version 3.
• SQL Server Management Studio (SSMS): This familiar tool to database professionals allows for scripting metadata, executing TMSL (Tabular Model Scripting Language) commands, and running DAX and MDX queries against Power BI semantic models.
• DAX Studio: The premier tool for DAX development and optimisation, DAX Studio offers advanced query analysis, debugging features, and performance metrics that help professionals write efficient and effective DAX.
• ALM Toolkit: Focused on application lifecycle management, this tool enables comparison and merging of semantic models across environments, simplifying deployment processes and ensuring consistency.
• PowerShell with Analysis Services cmdlets: For automation scenarios, PowerShell cmdlets offer programmatic control over semantic model management tasks like deployments and refreshes.
• Visual Studio with Analysis Services projects: Enterprise-grade modelling capabilities come to Power BI through Visual Studio, allowing teams to leverage established development practices for semantic models.

Critical Considerations: The One-Way Street

While the XMLA endpoint provides substantial benefits, it introduces a significant limitation that demands careful consideration: once a Power BI semantic model is modified through the XMLA endpoint, it can no longer be downloaded back to Power BI Desktop. This limitation creates what's effectively a one-way street for model development. When you choose to leverage the advanced capabilities of XMLA tools, you commit to continuing development through those tools rather than returning to the familiar Power BI Desktop environment.

This restriction exists because tools like Tabular Editor and SSMS can implement model features and properties that Power BI Desktop doesn't fully support or represent in its interface. Once these features are applied, the model becomes incompatible with the Desktop development experience. For organisations adopting XMLA-based workflows, this necessitates:

1. Maintaining backup copies of original PBIX files before XMLA modifications
2. Establishing clear development processes that account for this one-way transition
3. Training team members on external tools that will become primary development environments
4. Documenting which models have been modified via XMLA and are no longer Desktop-compatible

We've been caught out on this (only once!) when consolidating models for customers. The process can be reversed but it is not supported and very much down to luck whether it will work or not. 

The Default Change: What It Means for Your Organisation

The shift to read/write as the default setting significantly impacts how organisations work with Power BI. Prior to this change, capacity administrators needed to explicitly enable write operations, creating a barrier that often prevented semantic model owners from leveraging advanced tools. With the upcoming change, all capacity SKUs—including Power BI Premium, Premium Per User, and Fabric—will have XMLA read/write operations enabled by default. This change applies to all capacities using the default configuration as of 9 June 2025.

For security-conscious organisations that prefer maintaining read-only access, capacity administrators must proactively configure their environments before the change takes effect. This requires switching the XMLA Endpoint setting to OFF and then back to read-only mode, which removes the capacity from the default configuration path.

Closing notes:

The XMLA endpoint represents both an opportunity and a commitment for Power BI professionals. It unlocks enterprise-grade semantic modelling capabilities and connects Power BI to a broader ecosystem of professional tools, significantly expanding what's possible with the platform. However, it also introduces a point of no return for model development—once modified via XMLA, a model must continue its life cycle outside Power BI Desktop.

As Microsoft makes XMLA read/write the default configuration, organisations should evaluate their governance policies, development processes, and training needs. With proper planning, the XMLA endpoint can transform Power BI from a self-service BI tool to an enterprise analytics platform with unmatched flexibility and power.

Remember to maintain original PBIX files before committing to XMLA-based development, establish clear guidelines for when and how to leverage these capabilities, and ensure your team understands the implications of this powerful but consequential feature.