Fabric, Power BI Embedded and Power BI Premium Capacity

The reports in the Embedsy Portal run on a Microsoft capacity: the workspaces that hold them must be assigned to one. The portal embeds reports from any capacity type, but it can only start and pause a capacity that is an Azure resource, which means a Fabric (F SKU) or Power BI Embedded (A SKU) capacity. For most new portals we recommend a Fabric capacity.

Which capacity to choose

• Microsoft Fabric (F SKU) - bought in Azure, billed by the hour (pay-as-you-go) or prepaid as a reservation. Starts at F2, smaller than any A SKU, and also runs Fabric items such as lakehouses, warehouses and pipelines.
• Power BI Embedded (A SKU) - bought in Azure, billed by the hour with no commitment. It runs Power BI content only, so it suits a capacity that exists purely to serve embedded reports. A SKUs are not being retired; see Are Power BI Embedded A SKUs being retired?
• Power BI Premium (P and EM SKUs) - bought through Microsoft 365 with a monthly or yearly commitment. They embed reports, but they can't be paused and the portal can't manage them. Microsoft is retiring the P SKUs and moves those customers to F SKUs.

Each A SKU has an F SKU with the same compute: A1 = F8, A2 = F16, A3 = F32, A4 = F64, A5 = F128 and A6 = F256 (Capacity and SKUs in Power BI embedded analytics).

Pricing

Prices differ by SKU, region and payment model. Our blog post about capacity pricing compares them.

A pay-as-you-go capacity costs nothing while it is paused, and the Embedsy Portal can pause it for you. With Auto Managed Capacity switched on in Global Settings, the portal starts the capacity when a signed-in user opens the portal or a report, and pauses it once nobody has interacted with a report for the inactivity time you set (10 minutes by default, at least 5). Optional schedules (in UTC) keep it running at set times, for example during working hours. The first person to open a report after a pause waits while the capacity starts. For this to work, the service principal needs access to the capacity in Azure.

Monitoring

Whatever capacity you use, check its usage regularly: one that is too small slows reports down, one that is too large costs more than it needs to. See Fabric and Power BI Embedded capacity monitoring for the tools and how we can help.