Microsoft Fabric Pricing 2026: The Complete Guide to Costs, F-SKUs, and Licensing

Quick answer: Microsoft Fabric pricing is capacity-based, not per-user. You buy an F-SKU (Fabric capacity) measured in Capacity Units, starting at F2 for around $263 a month on pay-as-you-go and scaling up to enterprise tiers like F64 for roughly $8,410 a month on demand, or about 41 percent less with a one- or three-year reservation. On top of capacity, OneLake storage is billed separately at approximately $0.023 per GB per month, and Power BI Pro licenses (around $14 per user per month) are only required below the F64 tier.

If you have started evaluating Microsoft Fabric for your organization, the first question that comes up is almost always the same: how much does Microsoft Fabric actually cost? Unlike traditional per-user software pricing, Microsoft Fabric pricing works on a capacity-based model, and that shift trips up a lot of buyers who are used to counting seats instead of compute power.

In this guide, we break down Microsoft Fabric pricing, based on how Microsoft structures its official Azure pricing and licensing documentation. We will walk through the F-SKU price tiers, the difference between pay-as-you-go and reserved capacity, OneLake storage pricing, and the licensing rules that decide whether your users need a Power BI Pro seat or not. By the end, you should be able to map your own team size and workload to a realistic monthly number.

What Is Microsoft Fabric and Why Does Its Pricing Model Feel Different?

Microsoft Fabric is Microsoft's unified, AI-powered data and analytics platform. It brings data engineering, data integration, data warehousing, real-time intelligence, and Power BI together under one roof, all built on a single data lake foundation called OneLake.

Because Fabric bundles so many workloads (Lakehouse, Data Warehouse, Data Factory pipelines, Spark notebooks, Real-Time Intelligence, and Power BI) into one product, Microsoft did not price it the old way, per service, per engine. Instead, Microsoft Fabric pricing is built around a single, shared pool of compute called a capacity, measured in Capacity Units (CUs). You buy a capacity, and every Fabric workload you run draws from that same pool.

This is the core idea behind the Microsoft Fabric pricing model: one capacity, one bill, and no separate charges for switching between Data Factory, Spark, or Power BI Premium features, since they are already covered under your F SKU.

How Is Microsoft Fabric Billed?

Microsoft Fabric capacity is sold as an Azure resource, and it is billed on a per-second basis with a one-minute minimum. This is a genuinely important detail for anyone trying to control Microsoft Fabric cost, because it means you are not locked into a flat monthly invoice the way traditional Power BI Premium P-SKUs used to work.

There are two ways to buy Fabric capacity:

• Pay-as-you-go: No commitment, billed per second, and you can pause or resume the capacity whenever you like. This is the most flexible entry point for Microsoft Fabric pay as you go pricing, and it is the option most teams start with during evaluation.

• Reserved capacity: A 1-year or 3-year commitment purchased through Azure. Microsoft states that committing to a reservation can save around 41% compared to pay-as-you-go rates. Microsoft Fabric reserved capacity pricing is the better route once you know your steady-state usage pattern.

There is also a Microsoft 365 route (P-SKUs, the old Power BI Premium capacities), but Microsoft has been consolidating purchase options and is retiring Power BI Premium per-capacity SKUs for new customers. If you are buying fresh today, the Azure F-SKU route is the one Microsoft recommends.

Microsoft Fabric F-SKU Pricing Model Explained

Fabric capacities are sold in Stock Keeping Units labeled F2, F4, F8, and so on, up to F8192. Each F SKU corresponds to a specific number of Capacity Units, and CUs are what actually get consumed when you run a query, a pipeline, a notebook, or a Power BI refresh.

Here is the official SKU-to-CU mapping, along with the Power BI Premium equivalents for teams migrating from an older P-SKU:

This is the backbone of Fabric F SKU pricing and Fabric Capacity Units pricing. The larger the SKU, the more parallel workload the capacity can absorb before it throttles.

Approximate Monthly Microsoft Fabric Capacity Cost (USA, Pay-As-You-Go vs Reserved)

Actual Microsoft Fabric pricing USA figures are region specific, and Microsoft explicitly states that listed prices are estimates converted using monthly exchange rates, so the only way to get an exact, current quote is through the Azure Pricing Calculator or a sales specialist. That said, based on the published Azure rate structure and a 730-hour month, the approximate always-on monthly cost looks roughly like this for US regions:

These numbers are directional, meant to help you size a budget conversation, not a substitute for a live quote. Because pricing changes by region and currency, always confirm the current Microsoft Fabric capacity pricing for your billing region using the official Azure pricing calculator before finalizing a purchase decision.

What Is F64 in Microsoft Fabric, and Why Does Everyone Talk About It?

If there is one number in Microsoft Fabric pricing that you should remember, it is F64. The F64 SKU is the licensing threshold where Fabric behavior changes in a way that directly affects your Microsoft Fabric licensing cost.

On any capacity smaller than F64, every person who wants to view or consume Power BI content still needs an individual Power BI Pro (or PPU) license. Once your organization is on F64 or larger, users with only a free Microsoft Fabric license can view Power BI content, as long as they have a viewer role on the workspace. Creators and publishers still need Pro or PPU, but read-only consumers do not.

This is exactly why F64 is so often positioned as the "enterprise" tier: for organizations with a large audience of report viewers, the savings on Pro licenses alone can offset a meaningful chunk of the jump from F32 to F64.

Microsoft Fabric OneLake Storage Pricing

Compute (your F SKU) and storage are billed separately in Microsoft Fabric. All Fabric data, whether it sits in a Lakehouse, a Warehouse, or a Mirrored database, lives in OneLake, Fabric's unified, open-format data lake. OneLake storage cost does not draw from your Capacity Units; it is billed as its own pay-as-you-go meter, similar to how Azure Data Lake Storage is billed.

Fabric offers different OneLake storage tiers:

• OneLake hot storage – the default tier for actively used data, priced per GB per month.

• OneLake cool storage and OneLake cold storage – lower-cost tiers for less frequently accessed data.

• OneLake cache – used for KQL cache storage and Data Activator retained data, billed at a different, higher per-GB rate than standard hot storage.

• OneLake BCDR (Business Continuity and Disaster Recovery) storage – an optional geo-redundant tier for hot, cool, and cold data, priced separately from standard storage.

Published community and partner estimates, consistent with Microsoft's documented rate structure, put OneLake hot storage at roughly $0.023 per GB per month in most US regions, which works out to about $23 per TB per month. That means storing 5 TB of data in OneLake would run somewhere around $115 a month, on top of your capacity charge. As always, treat this as an approximation and check the live Azure pricing page for your exact region and currency.

One genuinely useful cost break: Fabric's Mirroring feature gives you free OneLake storage for database replicas, up to a limit tied to your purchased SKU. If you buy F64, for example, you get up to 64 TB of free Mirroring storage before OneLake storage charges kick in on that replicated data. This is a meaningful lever if your use case involves continuously mirroring source databases into Fabric.

Does Microsoft Fabric Include OneLake Storage?

Short answer: no, not for free beyond the specific exceptions above. Your F SKU price covers compute, meaning the processing power behind pipelines, Spark notebooks, Warehouse queries, and Power BI refresh operations. OneLake storage is billed on its own separate meter, based on how many gigabytes you actually store, regardless of which SKU you are on.

The one exception worth remembering is Power BI's native storage, which remains free up to the limit tied to your Power BI plan, for example 10 GB per user under Power BI Pro. Data stored in OneLake specifically for Power BI import semantic models is included in your existing Power BI licensing cost, not billed again separately.

Is Microsoft Fabric Free?

Microsoft Fabric offers a 60-day free trial that provisions a fixed trial capacity, functionally equivalent to an F64, for each participating business user. This lets you explore every Fabric workload, Lakehouse, Warehouse, Data Factory, Real-Time Intelligence, and Power BI, without buying a capacity first.

Once the trial period ends, or once you move beyond evaluation into production use, you need to purchase a paid F SKU to keep working. So Microsoft Fabric is free to try, but not free to run in production. There is no permanently free production tier for anything beyond the individual Power BI Free license, which only lets users create content in their own personal workspace.

Is Microsoft Fabric Priced Per User?

No, and this is the single biggest mental shift for teams coming from traditional software licensing. Microsoft Fabric capacity pricing is not per-user. You buy a shared pool of compute (an F SKU), and any number of people in your organization can use Fabric workloads against that same capacity, subject to their individual license type.

Where per-user cost does enter the picture is on the Power BI side. Depending on your SKU size:

• On F2 through F32, every person who needs to view Power BI content in a shared workspace needs a Power BI Pro or PPU license (an individual, per-user cost).

• On F64 and above, viewers can consume Power BI content with a free Fabric license; only creators and publishers need Pro or PPU.

So the honest answer is: Fabric capacity itself is a flat, shared-pool cost, but your total Microsoft Fabric licensing cost is capacity plus whatever per-user Power BI licenses your workspace configuration still requires.

Real-World Microsoft Fabric Cost Scenarios

To make Microsoft Fabric pricing 2026 more concrete, here are three simplified scenarios based on organization size. Treat the dollar figures as planning estimates, not quotes.

Small team or pilot project: An F4 reserved capacity (~$310/month) plus 10 Power BI Pro licenses (~$100–140/month, depending on current Pro pricing) plus a modest 500 GB of OneLake storage (~$12/month) lands in the neighborhood of $420–460 per month. This fits a small analytics team piloting Lakehouse and Power BI together.

Mid-market organization: An F32 reserved capacity (~$2,480/month), around 100–150 Power BI Pro licenses for creators and viewers, and 2–5 TB of OneLake storage puts most mid-market deployments somewhere in the $4,000–5,500 per month range, depending on headcount and data volume.

Enterprise deployment: An F64 reserved capacity (~$4,960/month) removes the need for Pro licenses on pure viewers, so even with 500+ report consumers, the incremental license cost stays low, since only creators need Pro or PPU. Add 10–20 TB of OneLake storage (~$230–460/month), and most enterprise deployments land in the $5,500–11,000 per month range before factoring in F128 or larger for very high concurrency.

How to Optimize Your Microsoft Fabric Cost

A few practical levers genuinely move the needle on your bill:

1. Pause capacities when they are not in use: Fabric lets you pause and resume pay-as-you-go capacities from the Azure portal, and you stop paying for compute the moment you pause, though OneLake storage charges continue.

2. Move to reserved pricing once your usage is predictable: The roughly 41% discount on 1-year or 3-year reservations is significant at any SKU size above F8.

3. Right-size before you scale up: Use the Fabric Capacity Metrics app to see how close you are running to your CU ceiling before jumping to the next SKU.

4. Reassess the F64 threshold early: If you expect more than roughly 50–80 report viewers, run the math on Pro license costs at F32 versus the jump to F64. The free-viewer benefit at F64 often pays for itself faster than expected.

5. Use tiered OneLake storage: Move cold, rarely accessed data out of hot storage to reduce the per-GB rate you are paying.

6. Turn on Fabric capacity overage carefully: It prevents throttling during demand spikes, but it is billed at pay-as-you-go rates even on a reserved capacity, so monitor it rather than leaving it unmanaged.

Final Thoughts

Microsoft Fabric pricing is fundamentally different from the per-user software licensing most teams are used to, and that is exactly why it confuses so many first-time buyers. Once you separate the three cost layers, Fabric capacity (your F SKU), OneLake storage, and Power BI per-user licensing, the model actually becomes easier to plan around than traditional Power BI Premium ever was.

The practical starting point for most organizations is simple: estimate your workload with a small F SKU or the free trial, watch your Capacity Unit consumption using the Fabric Capacity Metrics app, and only commit to reserved pricing once you have a real usage pattern to plan against. Getting the F64 decision right, in particular, can be the difference between an efficient rollout and an overpriced one.

We, at Dynamics Square, have spent more than 15 years helping organizations across the USA, UK, Canada, and Australia plan, license, and implement Microsoft's data and business application platforms the right way. With over 500 successful Dynamics 365 and Microsoft ecosystem implementations and a team of 150+ Microsoft-certified consultants, we help you translate Microsoft Fabric's capacity-based pricing into a plan that actually fits your data volume, team size, and budget, instead of a guess. Since 2011, our focus has stayed on getting the licensing and architecture decisions right the first time, so you are not stuck rebudgeting six months into a Fabric rollout.

If you are weighing Microsoft Fabric pricing against your current Power BI Premium setup, or trying to decide between F32 and F64 for your organization, Connect with our team can walk through the numbers with you and map out a capacity plan built around your actual usage, not a generic estimate.