Answer five questions. See the F-SKU that actually fits your workload, what it should cost per month, and the red flags most vendors won't mention until after you've signed.
Things most vendors won't flag until your bill shows up. Updates live as you change inputs.
This is a rule-of-thumb sizing tool, not a Microsoft-endorsed calculator. We built it because every prospect asks the same question and no one wants to run the Microsoft capacity calculator themselves.
The CU demand estimate adds up:
data volume CU — ranges from 2 (under 100 GB) to 60 (10 TB+)user CU — ranges from 1 (< 25) to 32 (500+)refresh cadence factor (1.0 daily, 1.6 hourly, 2.4 streaming)Copilot CU (0 / 8 / 24) and mirroring CU (0 / 3 / 8)We then apply a 20% safety buffer, match the result to the smallest Fabric SKU that covers it, and apply an F64 floor whenever Copilot is enabled (because Power BI Copilot requires F64+).
What we're not modeling: Direct Lake vs Import mix, semantic model size per workspace, query complexity, Spark pool cold-starts, Eventhouse ingestion throughput, and cross-region egress. All of these matter — which is why the tool gives you a range, not a guarantee, and why we built Tenant Scan to model them from your actual telemetry.
Prices are pay-as-you-go list prices in USD East US, April 2026. Reserved pricing assumes the standard 1-year commit (~29% off PAYG). Your actual negotiated rate may differ — Microsoft discounts are common above F128.
Your calculator result is an educated guess from five inputs. Tenant Scan is a 2-week deep scan of your real telemetry — workspace sprawl, refresh patterns, query hot spots, Copilot readiness. We tell you the SKU Microsoft won't, and the optimizations that pay for the audit ten times over.
Estimates are directional and based on published Microsoft SKU pricing as of April 2026. Final sizing should be validated against your actual Fabric telemetry. The Power Mates is a Microsoft partner and Dual MVP consultancy — not affiliated with or endorsed by Microsoft.