Join us at FabCon Atlanta from March 16 - 20, 2026, for the ultimate Fabric, Power BI, AI and SQL community-led event. Save $200 with code FABCOMM.
Register now!The Power BI Data Visualization World Championships is back! Get ahead of the game and start preparing now! Learn more
Hello,
I'm an intern that desperately needs some clarification regarding the new CoPilot preview feature of answering data questions. I notice that it only has been able to query data from visuals on my report, but it can't answer data questions from my semantic model. (Example: If I ask about a data table that isn't directly visualized on my report, but is in my semantic model, CoPilot won't generate a response). I'm a bit confused if I am doing something wrong, because Microsoft states that "Copilot will query your model and return the answer to your question in the form of a visual" as long as Q&A is enabled. In that same article, Microsoft provides screenshots of them querying the general pane for CoPilot and asking about their semantic model - they don't seem to be querying a Q&A visual. My understanding is that as long as Q&A is enabled, I should be able to query the CoPilot preview general pane about anything that is in my semantic model - not just my report.
Any help/guidance is appreciated! Thank you.
Solved! Go to Solution.
Hi, @Anonymous
Based on your description, CoPilot seems to have a limitation when querying data that isn't directly visualized in the report. This is a known limitation of the current CoPilot feature set. In the current iteration, CoPilot generates visuals and answers based primarily on the data that has been visualized in the report. This means that if some of the data in the semantic model isn't represented in any of the report's visuals, CoPilot might not be able to query the model directly or provide answers about it.
The documentation you cited emphasizes CoPilot's ability to query your model and return answers in the form of visuals, which is accurate. However, this feature depends to some extent on how the semantic model is used in the report itself. In order to get more comprehensive query capabilities, including data that is not visualized but present in the semantic model, it is important to ensure that the report is structured in such a way as to maximize the reach of CoPilot.
Before using CoPilot, evaluate your data and consider doing some cleansing of your semantic model so that CoPilot can derive insights from it more effectively. This may involve making sure that the model is well-structured and that the relationships between tables are properly defined. For more details, see the documentation:
Update your data model to work well with Copilot for Power BI - Power BI | Microsoft Learn
One way to extend the capabilities of CoPilot queries is to use Q&A visuals in your reports. By adding synonyms for data to your model, you can improve the way CoPilot understands natural language input, improving the quality of the answers provided. For guidance on setting up Q&A visuals, see Synonyms for adding data to your model:
Enhance Q&A with Copilot for Power BI - Power BI | Microsoft Learn
Because CoPilot is in preview, it is constantly learning and improving. Providing feedback on your experience and the limitations you encounter can help shape future enhancements.
How to Get Your Question Answered Quickly
Best Regards
Yongkang Hua
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.
Hi, @Anonymous
Based on your description, CoPilot seems to have a limitation when querying data that isn't directly visualized in the report. This is a known limitation of the current CoPilot feature set. In the current iteration, CoPilot generates visuals and answers based primarily on the data that has been visualized in the report. This means that if some of the data in the semantic model isn't represented in any of the report's visuals, CoPilot might not be able to query the model directly or provide answers about it.
The documentation you cited emphasizes CoPilot's ability to query your model and return answers in the form of visuals, which is accurate. However, this feature depends to some extent on how the semantic model is used in the report itself. In order to get more comprehensive query capabilities, including data that is not visualized but present in the semantic model, it is important to ensure that the report is structured in such a way as to maximize the reach of CoPilot.
Before using CoPilot, evaluate your data and consider doing some cleansing of your semantic model so that CoPilot can derive insights from it more effectively. This may involve making sure that the model is well-structured and that the relationships between tables are properly defined. For more details, see the documentation:
Update your data model to work well with Copilot for Power BI - Power BI | Microsoft Learn
One way to extend the capabilities of CoPilot queries is to use Q&A visuals in your reports. By adding synonyms for data to your model, you can improve the way CoPilot understands natural language input, improving the quality of the answers provided. For guidance on setting up Q&A visuals, see Synonyms for adding data to your model:
Enhance Q&A with Copilot for Power BI - Power BI | Microsoft Learn
Because CoPilot is in preview, it is constantly learning and improving. Providing feedback on your experience and the limitations you encounter can help shape future enhancements.
How to Get Your Question Answered Quickly
Best Regards
Yongkang Hua
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.
The Power BI Data Visualization World Championships is back! Get ahead of the game and start preparing now!
| User | Count |
|---|---|
| 56 | |
| 56 | |
| 35 | |
| 18 | |
| 14 |