Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 

Power BI is turning 10! Let’s celebrate together with dataviz contests, interactive sessions, and giveaways. Register now.

Reply
Avih
Frequent Visitor

Connected models in Desktop still create separate semantic models when published?

Hi everyone,

I have 3 Power BI models - 2 of them are connected to the semantic model of the 3rd one in Desktop. When I published all three to the service, I saw that each created its own semantic model (so 3 total).

I was expecting only the "main" model to create a semantic model, with the other two staying connected to it. Is this normal behavior?

What's the best way to handle this if I want to avoid duplicate semantic models in the service?

Thanks!

Edit:
This is what I did - Get Data -> Power BI semantic model, than I connected to the one in the service. and it shows me that the Storage Mode is DirectQuery instead of Live Connection and when I publish it, it creates a semantic model.

7 REPLIES 7
Poojara_D12
Super User
Super User

Hi @Avih 

Yes, what you're experiencing is the expected behavior in Power BI. When you publish a report from Power BI Desktop—even if it connects to another semantic model (dataset) in Desktop—Power BI Service treats the published file as a standalone report and automatically creates a new semantic model associated with it. This happens because Power BI Desktop doesn’t truly support live connections to Power BI Service datasets; it connects to local copies of those datasets during development. When published, these connections don’t carry over as live links but are instead embedded as either static imports or re-published semantic models, resulting in duplication.

To avoid creating duplicate semantic models in the service, the best practice is to develop and maintain your centralized or “main” semantic model separately—publish it once to the Power BI Service—and then, for all dependent reports, do not build them in Desktop by connecting to a local version of that semantic model. Instead, after publishing the main model, create the dependent reports directly in the Power BI Service or in Power BI Desktop by using the “Power BI datasets” connector, which connects live to the published dataset in the service. This way, when you publish those reports, you can choose to publish just the report (without its own dataset), thereby preserving a single source of truth and avoiding dataset duplication. You can do this by publishing a report file created from a live connection (which shows "Live connection" mode in Desktop) to the same workspace as the main dataset or a different one, depending on your governance setup.

 

Did I answer your question? Mark my post as a solution, this will help others!
If my response(s) assisted you in any way, don't forget to drop me a "Kudos"

Kind Regards,
Poojara - Proud to be a Super User
Data Analyst | MSBI Developer | Power BI Consultant
Consider Subscribing my YouTube for Beginners/Advance Concepts: https://youtube.com/@biconcepts?si=04iw9SYI2HN80HKS

Hi @Avih ,

If these solutions related to create separate semantic models when published  helped resolve your issue which was posted earlier.

we’d greatly appreciate it if you could mark this as the Accepted Answer it helps others facing a similar challenge find it quickly.

If there’s anything else you’d like to clarify or if you need further assistance, please don’t hesitate to share  we’re here to help!

Thanks once again for your valuable contribution to the Microsoft Fabric Community!

BhavinVyas3003
Solution Sage
Solution Sage

Yes, connected models in Desktop will still create separate semantic models when published.
To avoid this, always create thin reports by connecting to the published semantic model in the Service, not another local PBIX.


Thanks,
Bhavin
Problem solved? Hit “Accept as Solution” and high-five me with a Kudos! Others will thank you later!

This is what I did - Get Data -> Power BI semantic model, than I connected to the one in the service. and it shows me that the Storage Mode is DirectQuery instead of Live Connection and when I publish it, it creates a semantic model.

You're might be seeing extra semantic models because your report is using DirectQuery instead of a true Live Connection this usually happens when the "DirectQuery for Power BI datasets and AS" preview feature is enabled, allowing composite models. To avoid this and ensure publishing creates only a thin report (without a new semantic model), disable that feature in File → Options → Preview features, restart Power BI Desktop, then create a new PBIX using Get Data → Power BI semantic models without adding any other sources this will maintain a true Live Connection.


Thanks,
Bhavin
Problem solved? Hit “Accept as Solution” and high-five me with a Kudos! Others will thank you later!

Unfortunatly, this feature doesn't exist.

Hi @Avih ,


Apologies for the inconvenience you're experiencing with the semantic models after publishing. can you try below steps :

 

  • When you publish a PBIX connected to another semantic model via Desktop, it still creates its own dataset in the Power BI Service.
  • This is expected behavior Power BI doesn't maintain the local semantic model connection during publishing.
  • Each of your 3 PBIX files ends up creating its own dataset (semantic model), even if 2 of them were dependent in Desktop.
  • To avoid this, publish only the main PBIX file that contains the full data model.
  • For the other two PBIX files, open them in Desktop and remove their existing model.
  • Then, go to Home > Get Data > Power BI Datasets and connect to the semantic model from the Power BI Service.
  • These become “thin reports” that rely on the shared dataset in the Service no duplication of semantic models.
  • When you publish the thin reports, only the report is published  not another dataset.
  • If you edit or extend the model in the dependent PBIX (creating a composite model), it will still create a new dataset on publish.

Reference : Separate reports from models in Power BI Desktop - Power BI | Microsoft Learn


Hope this helps !!

If this post was helpful, please consider marking Accept as solution to assist other members in finding it more easily.




Helpful resources

Announcements
June 2025 Power BI Update Carousel

Power BI Monthly Update - June 2025

Check out the June 2025 Power BI update to learn about new features.

June 2025 community update carousel

Fabric Community Update - June 2025

Find out what's new and trending in the Fabric community.