If you have ever asked “is Power BI free, and if not, do I need Pro or Premium?” you are not alone — licensing is the single most confusing part of adopting Microsoft’s analytics platform, and the part we most often get asked to untangle for clients. This guide breaks down every Power BI licence for 2026, with verified pricing, worked cost examples, and a decision path so you buy exactly what you need.

Power BI has four commercial licensing options: Power BI Free, Power BI Pro, Power BI Premium Per User (PPU), and capacity-based Premium / Microsoft Fabric. They differ not just in price but in a subtle way that trips up most buyers: the licence you need depends on whether you are building reports, sharing them, or only viewing them.

Power BI licensing at a glance

Here is the whole landscape on one screen. These are Microsoft’s official per-user list prices as of 2026 — both Pro and PPU increased on 1 April 2025 (Pro from $10 to $14, PPU from $20 to $24):

The fourth option, Premium capacity (now delivered through Microsoft Fabric), is priced per capacity rather than per user — covered below, because it is where large organisations often save the most money.

How Power BI licensing actually works

Most licensing confusion disappears once you understand the model behind the prices. Power BI licensing is built on three layers, and every plan is just a different combination of them:

1. The authoring layer — the right to build reports and datasets and publish them to the Power BI service. Free covers authoring to a personal workspace; Pro and PPU cover authoring to shared workspaces.
2. The sharing layer — the right to distribute content to other people. This is what separates Free (no sharing) from Pro (sharing, but viewers must be licensed).
3. The capacity layer — the compute your content runs on. On “shared capacity” (the default for Pro and PPU) you’re on multi-tenant infrastructure and every viewer needs a licence. On “dedicated capacity” (Fabric F SKUs) you rent your own compute, and above F64 your viewers can be unlicensed.

When you buy Power BI, you are really deciding how much of each layer you need. A one-person analyst needs only the authoring layer (Free). A collaborating team needs authoring plus sharing (Pro). A power user needs a bigger engine (PPU). A large enterprise with a huge audience needs dedicated capacity so it stops paying per viewer (Fabric). Hold that mental model and the rest of this guide is just detail.

The concept that unlocks everything: creators vs viewers

Before the tiers, understand the distinction Microsoft’s whole model turns on. Every Power BI user is one of two things:

• Creators build and publish reports and datasets. A creator always needs at least a Pro licence (or PPU).
• Viewers only open and read reports someone else published. Whether a viewer needs a paid licence depends entirely on where the report lives: on shared per-user content, viewers need their own Pro/PPU; on a large Fabric capacity (F64+), viewers can use a free licence.

Power BI Free: what you actually get

Power BI Free (technically a Microsoft Fabric free licence) lets a single person do a surprising amount: connect to data, build rich reports in Power BI Desktop, and publish them to their own personal workspace. For a solo analyst or someone evaluating the tool, it is genuinely capable and costs nothing.

The catch is collaboration. With a Free licence you cannot share reports with colleagues, publish to shared workspaces, or (in most cases) consume content others have published. The moment two people need the same dashboard, Free stops being enough. In practice, Power BI Free is a personal and evaluation tier — not a team tier.

What Free is genuinely good for

• Learning Power BI and building your first reports.
• A one-person analytics function with no sharing need.
• Prototyping a report before you commit to a paid rollout.

Power BI Pro: the standard business licence ($14/user/month)

Power BI Pro is what most organisations actually run on. At $14 per user per month, it unlocks the thing Free is missing — sharing and collaboration. A Pro user can publish to shared workspaces, distribute reports as apps, subscribe colleagues to email updates, and collaborate on datasets.

The rule that catches people out: with Pro, every person who views the report also needs a Pro licence (unless the content sits in a Premium/Fabric capacity). So a report built by one Pro user and viewed by 40 colleagues means 41 Pro licences. That “viewer tax” is the single biggest driver of Power BI cost — and the main reason organisations eventually look at capacity licensing.

Power BI Premium Per User (PPU): $24/user/month

Premium Per User, launched in 2021, sits between Pro and full capacity. At $24 per user per month it includes everything in Pro plus the premium engine features that used to require an expensive capacity: much larger dataset sizes, far more frequent data refreshes (up to 48/day), paginated reports, advanced AI, XMLA endpoints and deployment pipelines.

PPU is ideal when a subset of your users — data scientists, BI developers, finance power users — need enterprise horsepower, but you do not have enough of them to justify a dedicated capacity. One important detail that catches teams out: to share PPU content, the people you share with also need a PPU licence (a Pro licence is not enough to open PPU content). That symmetry is what pushes very large audiences toward capacity licensing instead.

Premium capacity & Microsoft Fabric

Historically, Power BI Premium was sold as dedicated “capacity” (the old P SKUs). Microsoft has since folded this into Microsoft Fabric capacities (F SKUs), bought through Azure and billed by the second based on compute size rather than headcount.

The strategic advantage of capacity is the free viewer model: once you reach the F64 tier, users with only a free licence can view reports hosted on that capacity. For organisations with hundreds or thousands of report consumers, this flips the maths — instead of a Pro or PPU licence for every viewer, you pay for a fixed block of compute and let read-only users in for free. Report authors still need a Pro licence to publish, but the expensive “viewer tax” disappears.

Fabric capacity tiers, from F2 to F2048

Fabric capacities come in a range of sizes, each defined by its capacity units (CUs) — the amount of compute available. The SKU number is the CU count: F2, F4, F8, F16, F32, F64, F128, and up through F2048. Two things matter about this ladder:

• F64 is the magic threshold. Only at F64 and above do free-licensed users gain the right to view Power BI content on that capacity. Below F64, viewers still need Pro — so a small F2 or F8 capacity does not remove the viewer tax; it just gives you dedicated compute.
• Capacity is billed through Azure, per second, with the option to pause. If your reporting is only used during business hours, you can pause capacity overnight and at weekends to cut cost — something per-user licensing can never do. You can also reserve capacity for a further discount over pay-as-you-go.

This is why capacity licensing is so powerful at scale but wasteful at small scale: an organisation with 40 users would pay far more for an F64 than for 40 Pro seats, while an organisation with 2,000 viewers would pay a fraction of the per-user cost. The break-even is the whole game, and it’s specific to your numbers.

Copilot, AI and other add-ons

Two things worth budgeting for beyond the base licence:

• Copilot in Power BI — the generative-AI features run on Fabric capacity, so accessing Copilot generally requires a Fabric capacity (historically F64+), not just a per-user licence.
• AI Builder / premium AI — heavy AI workloads can consume capacity units and may need additional capacity.

For most standard reporting, you won’t need these on day one — but if AI-assisted analytics is on your roadmap, it pushes you toward capacity licensing sooner.

Nonprofit & education pricing

Microsoft offers discounted and, in some cases, donated licensing to eligible nonprofits and academic institutions through its nonprofit and education programmes. If you qualify, Power BI can cost dramatically less than commercial list price — so eligible organisations should always check programme eligibility before buying at standard rates.

Trials: test before you buy

You don’t have to commit blind. Microsoft offers a 60-day free trial of Power BI Pro, and Premium Per User has its own trial, so you can validate that a tier does what you need before paying for it. This matters more than it sounds: the difference between Pro and PPU is about engine capability (dataset size, refresh frequency), and the cleanest way to know whether you’ll hit Pro’s limits is to build your real model and watch it. We routinely run a short proof-of-value on a trial licence before recommending a tier, precisely so the licensing decision is evidence-based rather than a guess.

Migrating from Premium P SKUs to Fabric

If you bought Power BI Premium capacity in the past, you’ll have a P SKU (P1, P2, P3). Microsoft has unified Premium capacity into Microsoft Fabric F SKUs, and the two map roughly by capacity units — P1 ≈ F64, P2 ≈ F128, P3 ≈ F256. There are two practical implications:

• The free-viewer threshold is unchanged. P1 already gave you free viewing; its Fabric equivalent (F64) does the same. You don’t lose that benefit by moving.
• Fabric unlocks more than Power BI. The same capacity now also powers data engineering, data warehousing, real-time analytics and Data Factory workloads. If you’re going to use those, consolidating onto Fabric capacity can replace several separate tools — but if you only use Power BI, you’re simply on the modern equivalent of what you had.

Timing your move around your existing agreement’s renewal, and right-sizing the F SKU to your actual usage rather than a like-for-like swap, is where the savings (or the overspend) hide.

Free vs Pro vs Premium: full comparison

Model-size and refresh figures are Microsoft’s published limits and can change; treat them as current guidance, not contractual values.

Which Power BI licence am I? A simple decision path

Worked example: licensing a 300-person rollout

Common Power BI licensing mistakes

• Buying Pro for hundreds of view-only users. Past a few hundred viewers, Fabric F64 with free viewing is usually cheaper. Model it before you renew.
• Paying for Pro you already have. If your organisation is on Microsoft 365 E5, Power BI Pro is already included — audit your seats first.
• Over-buying PPU. PPU is for genuine power users; giving it to everyone is expensive and rarely necessary.
• Forgetting the sharing symmetry. PPU content can only be opened by PPU users — plan the whole audience, not just the author.
• Ignoring the nonprofit/education programmes. Eligible organisations routinely overpay by buying at commercial rates.

Governance: keeping Power BI costs under control

Licensing is a moment-in-time decision; cost control is ongoing. The organisations that keep Power BI affordable as it spreads tend to do a few things deliberately:

• Track your creator-to-viewer ratio over time. The number that justified Pro seats last year may justify a Fabric capacity this year as your audience grows. Review it at each renewal.
• Reclaim unused licences. Pro and PPU seats assigned to people who’ve moved on or stopped using Power BI are pure waste — audit assignments quarterly.
• Put reports in the right workspace. Hosting a widely-viewed report on a Fabric capacity (rather than per-user shared capacity) is what actually triggers the free-viewer saving. A capacity you’re paying for delivers nothing if your key reports still live on shared capacity.
• Use a centre of excellence. A small governing group that owns standards, security and licensing decisions prevents the sprawl of duplicate workspaces and ungoverned datasets that quietly inflate both cost and risk.

None of this is complicated, but it rarely happens by accident. Building a light governance rhythm into your BI operating model is the difference between Power BI costs that stay predictable and ones that creep every quarter.

How PraviMinds helps

As a Microsoft Dynamics 365 and Power Platform partner, we help organisations right-size their Power BI licensing — auditing existing Microsoft 365 entitlements, modelling per-user vs capacity break-even, and building the governance so costs stay predictable as adoption grows. We do this alongside the actual Power BI implementation, so the licensing decision is grounded in how your reports will really be built and consumed. If you are unsure whether you are over- or under-licensed, we will map it in a short call.

Frequently asked questions