# Power BI Pricing: Pro, Premium Per User and Fabric Licensing

Power BI Pro costs USD 14 per user per month and Premium Per User costs USD 24, both paid yearly at US list prices. A five-person team using Pro costs USD 70 a month; 50 licensed users cost USD 700. Larger audiences need a separate comparison with Microsoft Fabric capacity, because free internal viewing begins at F64 and report authors still need paid licences. [Microsoft Power BI pricing](https://www.microsoft.com/en-us/microsoft-fabric/features/power-bi/pricing).

Canonical URL: https://www.martechdevs.com/blog/power-bi-pricing

By [Jafrin Akter Alo](https://www.martechdevs.com/authors/jafrin-akter-alo)

Data Engineer and Co-founder at [Martechdevs](https://www.martechdevs.com) and [Growth and Analytics Partners](https://www.growthandanalytics.com)

[LinkedIn](https://linkedin.com/in/jafrinalo)

Last updated: 2026-11-11

Next step: [Compare BI licence costs](https://www.martechdevs.com/blog/bi-tool-pricing). Price the same audience across reporting tools.

## What are you billed for in Power BI?

Power BI charges for named user licences, capacity, or both, depending on where reports run and who accesses them. A Pro seat covers a person who creates or consumes shared Pro content. Fabric capacity buys a shared compute pool measured in capacity units, or CUs; it does not automatically replace every user licence.

USD public list prices, before tax. Pro and PPU amounts are monthly equivalents paid yearly. Fabric pricing varies by Azure region and purchase terms. [Power BI plans](https://www.microsoft.com/en-us/microsoft-fabric/features/power-bi/pricing) and [Fabric licensing rules](https://learn.microsoft.com/en-us/fabric/enterprise/licenses).

| Purchase | Price or unit | Who needs it | Main boundary |
| --- | ---: | --- | --- |
| Power BI Desktop and Free | USD 0 | Personal report building and eligible viewing | Free alone does not cover ordinary sharing in a Pro workspace |
| Power BI Pro | USD 14/user/month | Authors and readers in shared Pro workspaces | Check existing Microsoft 365 E5 or Office 365 E5 entitlements first |
| Premium Per User, PPU | USD 24/user/month | Everyone accessing content in a PPU workspace | Includes Pro; buying Pro again is unnecessary |
| Fabric F capacity | Regional CU price | Organization hosting shared compute | Below F64, Power BI readers still need Pro or PPU |
| Power BI Embedded | Capacity price | Application serving embedded analytics | Price the embedding model and concurrency separately |

PPU is an individual licence. Premium capacity is a pool of resources. A team can buy PPU for 12 people without buying a Fabric capacity, but PPU does not supply a Fabric lakehouse, warehouse or notebooks. [Microsoft's PPU guide](https://learn.microsoft.com/en-us/fabric/enterprise/powerbi/service-premium-per-user-faq).

Start with a list of authors, editors and readers, then map those jobs to the workspace mode. Our [data warehouse and reporting work](https://www.martechdevs.com/data-warehouse-cdp) connects that licence decision to the actual model, refresh schedule and access rules. A low seat price helps only when the deployment supports the required work.

## What will Power BI cost a company like yours?

Count the people who need access, subtract licences already included in their subscriptions, and multiply the remaining seats by the appropriate rate. These worked examples use constant user counts, annual Pro or PPU purchases and no negotiated discounts. The Fabric example adds capacity to author licences rather than treating capacity as the whole bill.

Each row is an illustrative deployment, not a recommended package for every company of that size. Source rates are the [US Power BI price list](https://www.microsoft.com/en-us/microsoft-fabric/features/power-bi/pricing). The F64 example uses East US, 64 CUs reserved for one year at USD 938 per CU for the term, from [Microsoft's public retail price data](https://prices.azure.com/api/retail/prices?%24filter=serviceName%20eq%20%27Microsoft%20Fabric%27%20and%20armRegionName%20eq%20%27eastus%27).

| Company situation | Explicit purchase assumption | Monthly equivalent | Annual amount |
| --- | --- | ---: | ---: |
| Five-person startup | All five use Pro; no existing entitlement | USD 70 | USD 840 |
| 20-person SaaS | Two authors and 18 readers, all on Pro | USD 280 | USD 3,360 |
| 50-person B2B SaaS | Ten authors and 40 readers, all on Pro | USD 700 | USD 8,400 |
| 200-person business with one specialist reporting group | 25 people use PPU; the other 175 receive no Power BI access through this purchase | USD 600 | USD 7,200 |
| 1,000-person organization | 20 Pro authors, 980 Free viewers; reports assigned to an adequately sized F64 reservation | USD 5,282.67 | USD 63,392 |

The last row is `64 × 938 + 20 × 14 × 12 = USD 63,392` per year. Its displayed monthly equivalent is rounded. It excludes storage, networking, extra capacity and implementation. The 980 viewers need the correct viewer access, and all relevant content must be on the qualifying capacity. [Fabric licensing](https://learn.microsoft.com/en-us/fabric/enterprise/licenses) and [reservation coverage](https://learn.microsoft.com/en-us/azure/cost-management-billing/reservations/fabric-capacity).

A company of 200 people does not automatically need 200 licences. The PPU row funds a 25-person audience only. Giving the other 175 people access to those PPU reports would increase that audience to 200 and the licence amount to `200 × 24 = USD 4,800` a month.

### When does F64 look cheaper than Pro seats?

For the same 20 authors, compare the additional viewer licences with the capacity amount. The example F64 reservation is `64 × 938 ÷ 12 = USD 5,002.67` a month. At USD 14 each, 358 additional Pro viewers cost USD 5,012, just above that capacity amount. This is a licence-cost crossover under these assumptions, not a sizing recommendation.

Before using that crossover, establish whether F64 can serve the models, refreshes and simultaneous queries. Subtract already included Pro licences, add other Fabric workloads, and include storage. A capacity that must be doubled to meet demand changes the result. Microsoft recommends testing capacity needs against actual workloads in its [licensing and implementation guidance](https://learn.microsoft.com/en-us/power-bi/guidance/powerbi-implementation-planning-subscriptions-licenses-trials).

## What makes a Power BI bill jump?

The biggest jumps come from expanding the licensed audience, moving shared reports into PPU workspaces, and adding capacity for workloads that user licences cannot cover. The relevant count is everyone who needs access to that content. A premium feature used by one author can therefore change the cost for an entire reporting group.

Three checks catch most planning mistakes:

- **A PPU workspace expands the PPU audience.** Moving a report used by 100 Pro users into PPU means budgeting for 100 PPU licences if all need continued access. At list price, the difference is USD 1,000 a month before existing upgrade entitlements. [PPU access requirements](https://learn.microsoft.com/en-us/fabric/enterprise/powerbi/service-premium-per-user-faq).
- **Refresh and model requirements can force a different purchase.** Pro allows up to eight scheduled refreshes per day; PPU supports up to 48. Diagnose the refresh duration and source bottleneck before buying more refresh slots. [Power BI refresh guidance](https://learn.microsoft.com/en-us/power-bi/connect-data/refresh-data).
- **Copilot adds a capacity question.** Pro or PPU alone does not meet Copilot's capacity requirement. Microsoft lists paid F2 or higher, or qualifying P1 or higher, plus tenant and region requirements. F2 access to Copilot does not give Free users the F64 viewing entitlement. [Copilot requirements](https://learn.microsoft.com/en-us/power-bi/create-reports/copilot-introduction).

Capacity overage can also add usage charges if enabled. Azure describes a 24-hour limit, checked every five minutes, and warns that billed overage can exceed that limit. Treat it as an operating control with a boundary, rather than a promise that the invoice cannot pass a number. [Fabric capacity pricing](https://azure.microsoft.com/en-us/pricing/details/microsoft-fabric/).

## Does adding people or adding usage cost more?

In Pro and PPU deployments, more licensed people directly increase the subscription. In capacity deployments, more work can require a larger capacity or incur enabled overage charges. Adding a Free viewer to a qualifying F64 workspace has no separate Pro seat charge, but that person's queries still consume resources.

Two teams with 100 readers can therefore have different needs. One opens a cached sales report each morning. The other filters large models throughout the day while data refreshes run. Headcount determines the seat calculation; query concurrency, model size and refresh behavior determine capacity pressure. Microsoft's [implementation planning guide](https://learn.microsoft.com/en-us/power-bi/guidance/powerbi-implementation-planning-subscriptions-licenses-trials) separates these decisions.

DirectQuery can also move part of the cost to the underlying database. Report interactions query the source rather than relying exclusively on an imported copy. Include that warehouse usage in the estimate, especially when many visuals refresh together. [Power BI storage and refresh modes](https://learn.microsoft.com/en-us/power-bi/connect-data/refresh-data).

## Which Power BI plan fits your situation?

Start with Pro for ordinary internal sharing, PPU for a limited audience that needs its additional features, and Fabric when the workload or distribution model justifies capacity. Keep the least expensive option that passes the access and performance checks. A product name containing Premium is not enough to establish that fit.

| Situation | First option to test | What could change it |
| --- | --- | --- |
| One analyst, private reports | Desktop and Free | Sharing or collaboration requirements |
| Small internal dashboard audience | Pro | Required PPU features or model limits |
| Limited specialist audience needing PPU capabilities | PPU for that full audience | Broader distribution or non-Power-BI Fabric workloads |
| Large internal audience | Pro versus properly sized F64+ | Existing E5 entitlements, peak load and annual commitment |
| Customer-facing analytics application | Embedded architecture | Authentication model, tenant isolation and concurrency |

The choices follow Microsoft's [licence planning guidance](https://learn.microsoft.com/en-us/power-bi/guidance/powerbi-implementation-planning-subscriptions-licenses-trials). For a lighter reporting alternative, compare [Data Studio and Power BI](https://www.martechdevs.com/blog/data-studio-vs-power-bi). For cross-vendor seat calculations, use our [BI pricing comparison](https://www.martechdevs.com/blog/bi-tool-pricing).

![Choose an embedded or internal reporting model, compare Pro, PPU and Fabric, then test access and peak load. Failed checks return to the requirements before approving annual cost.](https://www.martechdevs.com/assets/blog/power-bi-pricing.svg)

Diagram connections:

- Define audience and required report features to External application audience?. 
- External application audience? to Price the embedded architecture: Yes. 
- External application audience? to Need premium features or Fabric workloads?: No. 
- Need premium features or Fabric workloads? to Price Pro after checking existing licences: No. 
- Need premium features or Fabric workloads? to Compare PPU audience cost with Fabric capacity: Yes. 
- Price the embedded architecture to Access and peak load checks pass?. 
- Price Pro after checking existing licences to Access and peak load checks pass?. 
- Compare PPU audience cost with Fabric capacity to Access and peak load checks pass?. 
- Access and peak load checks pass? to Adjust model capacity or licence scope: No. 
- Adjust model capacity or licence scope to Define audience and required report features. 
- Access and peak load checks pass? to Approve full annual cost and ownership: Yes

## How do you keep Power BI costs in check?

Control the licence audience and remove avoidable workload before increasing capacity. Review the reports people actually use, their existing Microsoft entitlements, the workspaces they access and the model behind each report. Then measure the effect of each change on access, refresh time and capacity consumption, so savings preserve the reporting people need.

1. Check Microsoft 365 E5 and Office 365 E5 assignments before buying standalone Pro. The [Power BI price page](https://www.microsoft.com/en-us/microsoft-fabric/features/power-bi/pricing) lists those inclusions.
2. Separate a small PPU reporting group from ordinary Pro reports when the business permits it. Record which models each report depends on; a report built on a PPU model still has PPU access implications. [PPU content rules](https://learn.microsoft.com/en-us/fabric/enterprise/powerbi/service-premium-per-user-faq).
3. Remove unused columns, excessive history and repeated transformations before sizing up. Verify the resulting model against finance and operational totals. Our [marketing data modelling guide](https://www.martechdevs.com/blog/data-warehouse-design) explains how to keep useful detail without copying every source field into every report.
4. Set refresh frequency from a business need. A daily decision rarely justifies reloading its entire history every 30 minutes. Check incremental refresh and source query behavior before adding refresh slots. [Refresh planning](https://learn.microsoft.com/en-us/power-bi/connect-data/refresh-data).
5. Pause pay-as-you-go development capacity when it is genuinely unused. Pausing affects content availability and bills remaining smoothed usage; it does not erase completed work. [Pause and resume rules](https://learn.microsoft.com/en-us/fabric/enterprise/pause-resume).

Before approving a purchase:

## Checklist

- [ ] List authors and readers for each report, including external audiences.
- [ ] Confirm current Pro and PPU entitlements and renewal commitments.
- [ ] Check workspace mode and every dependent semantic model.
- [ ] Test the largest model, required refresh schedule and peak concurrency.
- [ ] Reconcile source totals and confirm each audience sees only its permitted data.
- [ ] Include annual licences, capacity, storage, source-query costs and ownership.

Use the [marketing data audit skill](https://www.martechdevs.com/skills/marketing-data-audit) to examine the data and reporting evidence behind that last set of checks. It supports the model and source assessment; the licence purchase still needs the actual Microsoft subscription terms.

## What should you watch for in the contract?

Confirm the payment term, seat-change rules, capacity region and what happens when a reservation ends. A monthly price displayed with yearly payment is an annual purchase. Treat free trials and startup credits as temporary funding, then calculate the next full year using the same workload and the applicable renewal prices.

For capacity, a reservation commits spend for one or three years and covers matching compute usage. Storage and networking remain separate. Unused reserved capacity does not become a refund simply because a development resource was paused; match the reservation to steady demand. [Fabric reservation terms](https://learn.microsoft.com/en-us/azure/cost-management-billing/reservations/fabric-capacity).

For user licences, ask the supplier when you can reduce purchased quantities and whether your existing agreement includes an upgrade path to PPU. Keep the answers in the quote. Avoid treating a licence reassignment, a seat cancellation and a subscription cancellation as the same operation.

Microsoft has a [startup programme with Azure credits](https://www.microsoft.com/en-us/startups). Do not subtract those credits from a Power BI quote until the approved offer confirms the eligible product, subscription and expiry. The recurring budget should still show the full bill after the benefit ends. In the 20-user Pro example, an unchanged list-price year is USD 3,360; a promotional first year does not change that baseline.

## How do you read the Power BI rate card?

Read the price alongside the workspace and access entitlement. Pro's USD 14 and PPU's USD 24 are per-user monthly equivalents paid yearly. Fabric's F number denotes capacity units. The price of the capacity and the right of a Free user to view its Power BI content are separate questions.

Keep these translations beside the quote:

- **Paid yearly:** multiply the monthly seat amount by 12 for the annual commitment.
- **PPU includes Pro:** use the full PPU price or the eligible quoted upgrade, without adding a second full Pro licence.
- **F64:** 64 CUs, with the documented Free-viewer entitlement when the content and access configuration qualify.
- **Reservation:** a commitment applied to matching capacity usage, excluding storage and networking.
- **Embedded:** an application distribution model with its own capacity and identity design.

Use the [Power BI pricing page](https://www.microsoft.com/en-us/microsoft-fabric/features/power-bi/pricing), [Fabric price selector](https://azure.microsoft.com/en-us/pricing/details/microsoft-fabric/) and [embedding licence guide](https://learn.microsoft.com/en-us/power-bi/developer/embedded/embedded-capacity) together. If you retrieve rates programmatically, Microsoft's [Retail Prices API guide](https://learn.microsoft.com/en-us/rest/api/cost-management/retail-prices/azure-retail-prices) explains regional meters and terms; a reservation amount is a term price even when its meter has an hourly unit label.

## What should you do before choosing Power BI licences?

Price one real reporting workflow from source to reader before purchasing organization-wide licences. Record the model, refresh frequency, people, access rules and annual commitment, then test the busiest realistic usage. That gives you a defensible Pro, PPU or capacity decision and exposes costs that a seat-only comparison would miss.

If HubSpot supplies the data, start with our [HubSpot and Power BI integration guide](https://www.martechdevs.com/blog/hubspot-power-bi-integration). Use the [BI cost comparison](https://www.martechdevs.com/blog/bi-tool-pricing) to compare the same audience across tools. For the model, pipelines and reporting implementation, see our [data warehouse and CDP service](https://www.martechdevs.com/data-warehouse-cdp).

## Frequently asked questions

### Can I use Power BI Desktop without paying?

Yes. Power BI Desktop supports free report creation, and a Free account supports personal use within its permissions. Sharing confidential reports with colleagues introduces licence and workspace requirements. Start by separating personal analysis from an ongoing shared reporting service, then price the people who need the latter. [Power BI plans](https://www.microsoft.com/en-us/microsoft-fabric/features/power-bi/pricing).

### Does one PPU licence let everyone view my reports?

No. Content in a PPU workspace normally requires PPU access for every consumer. Buying PPU for the author alone does not fund the audience. Check the semantic model as well as the report workspace, because moving a report does not remove a PPU dependency underneath it. [Microsoft PPU access rules](https://learn.microsoft.com/en-us/fabric/enterprise/powerbi/service-premium-per-user-faq).

### Is Microsoft Fabric F2 enough for free Power BI viewers?

No. F2 can supply Fabric capacity, but the Free-user viewing threshold for internal Power BI content is F64 or higher. On smaller F capacities, viewers generally need Pro or PPU. Copilot's minimum eligible capacity is a separate requirement, so an F2 Copilot deployment does not establish free internal report distribution. [Fabric licensing](https://learn.microsoft.com/en-us/fabric/enterprise/licenses).

### Can Publish to web replace paid licences for private reports?

No. Publish to web makes the content publicly accessible without authentication, including potentially underlying detail. It is appropriate only for information approved for public release. Use authenticated sharing or a properly designed embedded application for confidential customer or company reports, and include the corresponding licensing in the budget. [Microsoft's Publish to web warning](https://learn.microsoft.com/en-us/power-bi/collaborate-share/service-publish-to-web).

### Does a Fabric reservation include OneLake storage?

No. A Fabric capacity reservation covers matching capacity compute, while storage and networking charges remain separate. Estimate data retention and relevant data movement alongside the reservation. Also check expiry and renewal settings: continuing workloads can return to pay-as-you-go billing after a reservation ends unless an applicable replacement reservation covers them. [Reservation coverage](https://learn.microsoft.com/en-us/azure/cost-management-billing/reservations/fabric-capacity).

## Sources

1. [Microsoft Power BI pricing](https://www.microsoft.com/en-us/microsoft-fabric/features/power-bi/pricing)

2. [Microsoft Fabric licences and capacity](https://learn.microsoft.com/en-us/fabric/enterprise/licenses)

3. [Power BI Premium Per User](https://learn.microsoft.com/en-us/fabric/enterprise/powerbi/service-premium-per-user-faq)

4. [Microsoft Fabric pricing](https://azure.microsoft.com/en-us/pricing/details/microsoft-fabric/)

5. [Microsoft Fabric East US retail rates](https://prices.azure.com/api/retail/prices?%24filter=serviceName%20eq%20%27Microsoft%20Fabric%27%20and%20armRegionName%20eq%20%27eastus%27)

6. [Azure Retail Prices API](https://learn.microsoft.com/en-us/rest/api/cost-management/retail-prices/azure-retail-prices)

7. [Fabric capacity reservations](https://learn.microsoft.com/en-us/azure/cost-management-billing/reservations/fabric-capacity)

8. [Power BI licence implementation planning](https://learn.microsoft.com/en-us/power-bi/guidance/powerbi-implementation-planning-subscriptions-licenses-trials)

9. [Data refresh in Power BI](https://learn.microsoft.com/en-us/power-bi/connect-data/refresh-data)

10. [Copilot for Power BI requirements](https://learn.microsoft.com/en-us/power-bi/create-reports/copilot-introduction)

11. [Pause and resume Fabric capacity](https://learn.microsoft.com/en-us/fabric/enterprise/pause-resume)

12. [Power BI embedded capacity](https://learn.microsoft.com/en-us/power-bi/developer/embedded/embedded-capacity)

13. [Publish to web](https://learn.microsoft.com/en-us/power-bi/collaborate-share/service-publish-to-web)

14. [Microsoft for Startups](https://www.microsoft.com/en-us/startups)
