Move off Power BI Premium without overpaying for Fabric
Microsoft no longer sells Power BI Premium P SKUs. When your agreement term ends, workspaces have to move to a Fabric F SKU. The default advice is “P1 becomes F64” — but a right-sized, reserved capacity is often cheaper, and a rushed move risks throttled or rejected workloads.
After a P SKU subscription ends: 30-day grace period, throttling from day 31, all operations rejected from day 91.
What happens when a P SKU term ends
Each P SKU subscription ends at the end of its current agreement term. After that the capacity enters a 30-day grace period; from day 31 interactive operations are throttled, and from day 91 all operations are rejected. Data is retained but inaccessible until workspaces are moved to an F SKU.
Customers with an active Enterprise Agreement can keep renewing existing P capacity annually until the EA ends; customers on expiring agreements cannot buy new P capacity. There is no automatic conversion — workspaces are reassigned to the new capacity in the Fabric admin portal or with Microsoft's capacity migration notebook.
F64 is the same size as P1 and keeps the free-viewer benefit, which is why it is the default mapping. But F SKUs can be reserved for a year at roughly 40.5% below pay-as-you-go, and many P1 estates use far less than their capacity at peak. Picking the size from measured usage — not from the old SKU name — is where the savings are.
Signs this applies to you
- Your P SKU renewal or Enterprise Agreement end date is within the next 6–12 months
- Nobody has looked at the Capacity Metrics app in months
- Capacity was upgraded in the past to “fix” slow reports
- You run Power BI Report Server on Premium licensing and need to keep it
- Finance is asking whether Fabric will cost more than Premium
How the migration sprint works
- 1
Measure current usage
Pull several weeks of Capacity Metrics data: peak interactive load, background refresh load, throttling events and the largest models' memory.
- 2
Right-size and price it
Recommend an F SKU from measured usage, compare reserved vs pay-as-you-go, and check the viewer-licensing impact (at list price a reserved F64 equals about 358 Pro licences at $14).
- 3
Fix what would force a bigger SKU
Colliding refresh schedules, oversized models and expensive DAX often decide the SKU. Fixing the top few can drop a whole size.
- 4
Reassign and validate
Move workspaces in waves, verify refreshes, gateways, Power BI Report Server licensing and embedded apps, and watch the first weeks of capacity metrics.
What you get
- Capacity usage analysis and recommended F SKU with reasoning
- Reserved vs pay-as-you-go cost comparison and licensing impact
- Workspace migration runbook with rollback steps
- Workspaces reassigned and validated in agreed waves
- Post-migration capacity health check
Fixed-Scope Engineering Sprint — Scoped and priced after a short call. Priced by number of capacities and workspaces. Start with the free capacity estimator; a Migration & Right-Sizing Report is coming soon for teams that want the plan without the delivery. See all pricing.
Related reading
Frequently asked questions
- Is P1 to F64 always the right move?
- F64 is the same size and keeps free viewers, so it is the safe default. It is not always the cheapest: with measured usage, a reserved F64 is roughly 40% below pay-as-you-go, and estates that only needed P1 for licensing can sometimes split dev/test onto a smaller pay-as-you-go capacity that is paused out of hours.
- Do we have to move every workspace on the same day?
- No. Workspaces are reassigned individually, so a migration can run in waves — lowest-risk workspaces first — before the P SKU term ends.
- What about Power BI Report Server?
- Power BI Report Server remains available with a Fabric capacity reservation or SQL Server Enterprise Edition with Software Assurance. Confirm which applies to you before the Premium term ends.
- How long does a migration take?
- It depends mainly on the number of capacities and workspaces and how much usage data already exists. You get a timeline in the written scope; the date that matters is your P SKU term end.
Sources: Power BI Premium to Fabric migration overview (Microsoft Learn).