Microsoft Power BI Pricing 2026: Pro, PPU & Fabric Costs

Microsoft Power BI pricing starts with a Free account, then Power BI Pro at $14 per user/month paid yearly and Power BI Premium Per User at $24 per user/month paid yearly. Enterprise scale moves into Microsoft Fabric capacity and Power BI Embedded, where pricing is capacity- or consumption-based rather than a fixed per-user list.

Free account

Power BI Free costs $0 and does not require a credit card.

It supports individual creation and exploration, including Power BI Desktop and personal workspace scenarios.

The limitation is organizational sharing: free users cannot simply collaborate in standard Pro workspaces without the appropriate licensing or capacity.

Free is best for learning, evaluation and individual report development.

Power BI Pro: $14/user/month

Power BI Pro costs $14 per user per month paid yearly, equivalent to $168 per user each year.

Pro enables publishing, sharing, workspace collaboration and consumption of reports shared under standard Pro licensing.

Microsoft currently includes Pro with Microsoft 365 E5 and Office 365 E5.

Organizations already buying E5 should therefore check existing entitlement before purchasing duplicate Pro licenses.

Premium Per User: $24/user/month

Premium Per User costs $24 per user per month paid yearly, equivalent to $288 per user each year.

PPU includes all Pro capabilities plus many Premium Power BI features.

Microsoft guidance says PPU can be cost effective when Premium Power BI functionality is needed for fewer than roughly 250 users.

PPU is a Power BI per-user license and does not create Fabric capacity.

What PPU adds

PPU supports larger semantic models, more frequent refresh, XMLA endpoints, deployment pipelines and other premium-scale capabilities.

The Toollers source snapshot records up to one hundred gigabytes per semantic model and up to forty-eight refreshes per day as relevant PPU capabilities.

Exact operational limits can change with Microsoft service updates, so administrators should verify current documentation for production design.

PPU is most useful when a relatively small expert group needs Premium functionality.

Fabric capacity

Microsoft Fabric capacity is purchased as shared compute and is priced separately from Pro and PPU seats.

F SKUs provide capacity units that can run Power BI alongside Fabric workloads such as Data Factory, lakehouse, warehouse, notebooks and real-time intelligence.

Pricing varies by SKU, Azure region, commitment and usage model.

Capacity can become the dominant BI cost in large enterprises even though Pro seat prices are low.

Viewer licensing on Fabric

On Fabric F SKUs smaller than F64, users viewing Power BI content generally still need Pro, PPU or an eligible individual trial.

On F64 and larger capacities, users with a Free license and Viewer role can consume qualifying Power BI content without individual Pro seats.

This can materially change the economics of distributing reports to thousands of readers.

Model creators and viewers separately before comparing capacity with per-user licensing.

Power BI Embedded

Power BI Embedded uses capacity pricing for analytics embedded inside applications and customer experiences.

Cost depends on the selected capacity and deployment pattern rather than a public flat per-viewer price.

Embedded architectures also require development, authentication and tenancy design.

Software companies should include engineering cost and capacity utilization in the total cost model.

Copilot cost implications

Power BI Copilot is not simply included because an author has Pro or PPU.

Microsoft's current guidance requires access to a paid Fabric capacity of F2 or higher or Power BI Premium capacity of P1 or higher, with Copilot enabled and supported regional settings.

Microsoft also supports Fabric Copilot capacity for centralized AI usage billing in some scenarios.

Organizations planning broad AI use should budget capacity consumption in addition to per-user Power BI licensing.

Microsoft 365 E5 economics

Power BI Pro is included with Microsoft 365 E5 and Office 365 E5.

This can make Power BI especially attractive when the organization already licenses E5 broadly.

The effective incremental BI seat cost may be lower than the $14 standalone list price.

PPU, Fabric capacity, embedded analytics and other Microsoft services remain separate.

Sharing costs

A common Power BI budgeting mistake is licensing report authors but not planning for viewers.

Pro-to-Pro sharing is straightforward but can require many paid viewer licenses.

Capacity changes that equation at scale, particularly F64 and above for eligible Free viewers.

Estimate the number of creators, collaborators and read-only consumers separately.

Gateway and infrastructure costs

The on-premises data gateway software itself is not the main licensing cost, but operating reliable gateway infrastructure requires servers, monitoring, patching and support.

Organizations may also pay for Azure services, databases, VPN/networking or private connectivity around the BI environment.

These infrastructure costs can exceed a few Pro seats in complex deployments.

Include them when comparing Power BI with fully cloud-native alternatives.

Connector and data-platform costs

Power BI connects to many data sources, but external systems can carry their own database, API or egress charges.

Databricks, Snowflake, Azure SQL, Fabric, Salesforce and other sources have commercial terms separate from Power BI.

Custom connectors may also require engineering effort.

Do not treat connector availability as evidence that the full data stack is included in the $14 Pro rate.

Fabric trial versus Free Power BI

Microsoft Fabric provides trial capabilities separate from the permanent Power BI Free per-user account.

Fabric trials can be useful for testing capacity-backed workloads and non-Power-BI Fabric experiences.

Trial capacity should not be treated as permanent production entitlement.

Keep free-user licensing, individual trials and Fabric capacity trials distinct in procurement documentation.

Cancellation and refunds

Power BI business subscriptions follow Microsoft's commercial billing and cancellation terms rather than one universal consumer refund policy.

Refund or prorated-credit eligibility can depend on purchase channel, billing account type, timing, region and agreement.

Enterprise Agreement, CSP, Microsoft 365 and Azure purchases can therefore have different commercial mechanics.

Use the organization's actual Microsoft contract as the authoritative cancellation source.

Power BI versus Tableau

Tableau Cloud Standard currently lists Creator at $75 per user/month, Explorer at $42 and Viewer at $15, billed annually.

Enterprise Edition is higher, with Creator at $115, Explorer at $70 and Viewer at $35.

Power BI Pro's $14 list price is substantially lower for report creators, especially in Microsoft-centric organizations.

Tableau can still justify its price where visual analytics workflow, platform preference or existing Tableau expertise matters more than seat economics.

Power BI versus Looker

Looker uses platform plus user pricing and currently requires sales engagement for Standard, Enterprise and Embed platform costs.

Each core platform includes a baseline set of Standard and Developer users and then adds user licensing.

Looker emphasizes a centralized semantic layer and Google Cloud integration.

Power BI is easier to price for small authoring teams, while Looker may fit organizations that prioritize governed LookML modeling.

Power BI versus Qlik

Qlik Cloud Analytics currently starts at $300/month for Starter with ten users and ten gigabytes of data for analysis, billed annually.

Standard starts at $825/month and Premium at $2,750/month, with Enterprise quoted separately.

Qlik increasingly prices around data capacity rather than simple users.

Power BI usually has a lower per-author entry price, while Qlik can be attractive for associative analysis and capacity-based scaling.

Power BI versus ThoughtSpot

ThoughtSpot now offers user- and usage-oriented pricing around Essentials, Pro and Enterprise.

Current pricing shows Essentials from $25 per user/month billed annually, while Pro and Enterprise can use broader user or consumption models depending on configuration.

ThoughtSpot differentiates through search and agentic analytics.

Power BI remains cheaper for many authoring scenarios but requires more traditional semantic modeling and DAX expertise.

Pricing recommendation

Start with Free for evaluation and Pro for standard collaborative BI.

Use PPU when a limited expert group needs Premium Power BI features and Fabric workloads are not required.

Evaluate Fabric capacity when viewer scale, Fabric data workloads, Copilot or organization-wide compute economics justify it.

Build the decision around creator, viewer and capacity requirements instead of choosing the cheapest-looking individual license.

Pricing analysis: Licensing design

Power BI licensing decisions should start with who creates content, who shares it, who only views it, and whether the organization will use Fabric capacity. A small authoring team can have very different economics from a company that needs thousands of viewers.

Map roles before buying licenses so the organization does not pay Pro or Premium Per User for people who could legitimately consume content through eligible capacity.

Pricing analysis: Semantic-model governance

Power BI becomes more maintainable when shared semantic models, measures and definitions are governed centrally. Without that discipline, departments can create multiple versions of the same KPI and undermine trust in dashboards.

Define ownership for datasets, measures, refresh schedules, gateways and certification before self-service adoption expands.

Pricing analysis: DAX learning curve

Basic reporting can feel familiar to Excel users, but advanced DAX, filter context and dimensional modeling require real expertise. Organizations should budget training time instead of assuming drag-and-drop visuals eliminate the modeling learning curve.

Strong model design often improves both performance and maintainability more than adding visual complexity.

Pricing analysis: Performance planning

Large semantic models, DirectQuery patterns, high-cardinality columns and inefficient DAX can create slow reports. Capacity can help at scale, but poor models can still waste compute.

Use Performance Analyzer, model-size optimization and representative production data before deciding that licensing or capacity alone is the performance bottleneck.

Pricing analysis: Fabric architecture

Microsoft Fabric can consolidate data engineering, warehousing, lakehouse, real-time analytics and Power BI on one capacity platform. That can simplify a Microsoft-centric data estate but also changes the cost model from per-user BI toward shared compute capacity.

Decide which workloads actually need Fabric before treating the broader platform as a mandatory Power BI expense.

Pricing analysis: Copilot governance

Copilot in Power BI requires eligible paid capacity and administrator configuration, and AI consumption can create additional capacity cost. Organizations should define which users can access it, which workspaces are eligible and how generated content is reviewed.

Good semantic-model descriptions and governed business definitions are especially important because AI output depends heavily on the quality of the underlying model.

Pricing analysis: Connector governance

Power Query supports a very broad connector ecosystem across files, databases, Azure, SaaS services and custom connectors. Connectivity breadth is useful, but every source introduces credentials, refresh, privacy and ownership requirements.

Maintain an approved connector and gateway strategy so self-service analytics does not create unmanaged data access.

Pricing analysis: Sharing architecture

The biggest licensing surprises often appear when reports need to be distributed broadly. Pro workspaces, Premium Per User workspaces and Fabric/Premium capacities have different viewer requirements.

Test the exact sharing scenario—including external users and free viewers—before buying hundreds of seats.

Pricing analysis: Security review

Power BI inherits Microsoft Entra identity, tenant controls, encryption, information protection and broader Microsoft governance capabilities. Enterprise buyers should still review tenant settings, external sharing, gateway architecture, sensitivity labels and capacity administration.

Security is not only a Microsoft setting; report authors and dataset owners can still expose inappropriate data if row-level security and permissions are poorly designed.

Pricing analysis: Decision framework

Compare BI platforms on modeling, visualization, governance, sharing, embedded use, AI, ecosystem fit and total three-year cost. Power BI is often inexpensive per author, but enterprise capacity and data-platform choices can dominate final cost.

The best BI platform is the one that produces trusted, maintainable analytics with a licensing model the organization can govern.

Pricing analysis: Licensing design

Power BI licensing decisions should start with who creates content, who shares it, who only views it, and whether the organization will use Fabric capacity. A small authoring team can have very different economics from a company that needs thousands of viewers.

Map roles before buying licenses so the organization does not pay Pro or Premium Per User for people who could legitimately consume content through eligible capacity.

Pricing analysis: Semantic-model governance

Power BI becomes more maintainable when shared semantic models, measures and definitions are governed centrally. Without that discipline, departments can create multiple versions of the same KPI and undermine trust in dashboards.

Define ownership for datasets, measures, refresh schedules, gateways and certification before self-service adoption expands.

Pricing analysis: DAX learning curve

Basic reporting can feel familiar to Excel users, but advanced DAX, filter context and dimensional modeling require real expertise. Organizations should budget training time instead of assuming drag-and-drop visuals eliminate the modeling learning curve.

Strong model design often improves both performance and maintainability more than adding visual complexity.

Pricing analysis: Performance planning

Large semantic models, DirectQuery patterns, high-cardinality columns and inefficient DAX can create slow reports. Capacity can help at scale, but poor models can still waste compute.

Use Performance Analyzer, model-size optimization and representative production data before deciding that licensing or capacity alone is the performance bottleneck.

Continue reading about Microsoft Power BI

• Microsoft Power BI tool profile
• Microsoft Power BI Review: Features, Pricing, Pros and Cons
• Microsoft Power BI Alternatives 2026: 4 BI Platforms