How to create a report page with Copilot in Power BI

João Barros 06 de July de 2026 4 min read

Building a report page from scratch can be slow, especially while you are still exploring the data and deciding which visuals make sense. Copilot in Power BI helps you get started much faster: you describe what you want in natural language and it generates a complete page, with visuals already connected to your model. In this practical guide you will learn how to create a report page with Copilot in Power BI, from the prerequisites to validating the result.

Prerequisites

An up-to-date version of Power BI Desktop (Copilot is also available when editing reports in the Power BI Service).
Admin, member or contributor access to a workspace on a paid Fabric F2 or higher capacity, or Power BI Premium P1 or higher.
Copilot enabled at the tenant level by your Fabric administrator.
A data model already loaded, with tables, relationships and clear column names.
Note: trial capacities are not supported; you need a paid capacity.

Step 1: Confirm the capacity and enable Copilot

Copilot only works when the report is associated with an eligible capacity. A Power BI Pro or Premium Per User (PPU) license on its own is not enough. Ask your administrator to confirm that a paid F2+ (Fabric) or P1+ (Premium) capacity exists in a supported region and that the tenant setting that allows Copilot is enabled. If the Copilot button appears greyed out, it usually means the capacity or the tenant permission is missing.

Step 2: Open the Copilot pane

Open a report that already has a data model loaded. On the Home tab of the ribbon, click the Copilot button. The pane opens on the right. If prompted, choose a workspace that sits on an eligible capacity — that workspace provides the compute for Copilot, even when you are working in Desktop.

Step 3: Describe the page you want to create

In the pane, you can let Copilot suggest topics based on your model or write your own description. The more specific your request — metrics, dimensions and time period — the better the result. A good example request is:

Create a sales overview page with total revenue, revenue by month, revenue by product category and the top 5 customers by revenue.

Send the request and wait a few seconds. Copilot creates a new page with the suggested visuals, ready to review.

Step 4: Review and refine the result

The generated page is a starting point, not a finished product. Check that each visual uses the correct measure and field, and adjust whatever you need by asking for changes in natural language. For example:

Change the revenue-by-month chart to a line chart and add a slicer by year.

You can repeat this process several times: change chart types, add filters or remove visuals you do not need. Treat Copilot as a colleague who produces a quick first draft that you then polish.

Verify the result

Before you share the page, confirm that the numbers are correct. Compare a known value — for example, total revenue for the year — with a figure you have already validated elsewhere. Also check that the page-level filters are what you expect and that the visual titles clearly describe what they show. Copilot builds the visual, but responsibility for the accuracy of measures and filters remains yours.

Conclusion

In just a few steps, Copilot in Power BI turns a natural-language description into a working report page, saving the initial effort of dragging fields and choosing visuals. From here, try refining the same page with increasingly precise requests and compare the result with a page you would build by hand. Which approach gives you more control — and when do you prefer each one?