Power BI vs Tableau: Licensing, DAX and SQL (2026)

Quick verdict

Short answer

Choose Power BI if your organisation runs on Microsoft 365, Azure or Microsoft Fabric, your authors use Windows, and you want a strong central semantic model at a per-user price that is low for the category. Choose Tableau if visual exploration is the priority, your authors include macOS users, you need a self-hosted server on Linux, or you are invested in Salesforce. For SQL users the practical difference is the modelling layer: Power BI expects you to learn DAX and Power Query, while Tableau calculations stay closer to spreadsheet and SQL expressions, with level of detail expressions for the harder cases.

How we know: This comparison is research-based: licensing, prices, platforms, connectivity modes, SQL access, calculation languages and embedding were checked against Microsoft's Power BI pricing page and Microsoft Learn, and against Tableau's help, pricing and technical specification pages and Salesforce's Tableau Next page, in October 2026. We have not built reports in either product for this page and make no performance claims.

Power BI is Microsoft's business intelligence product. Authors build a semantic model (tables, relationships and DAX measures, with data shaped in Power Query) and reports on top of it, usually in the free Power BI Desktop app, then publish to the Power BI service, which is now a workload inside Microsoft Fabric. Sharing needs Power BI Pro or Premium Per User (PPU) licences, or content hosted on a Fabric capacity of F64 or larger for free-licence viewers. An on-premises option, Power BI Report Server, also exists.

Tableau is a visual analytics platform owned by Salesforce. Authors build workbooks in Tableau Desktop or in the browser and publish them to Tableau Cloud (SaaS) or Tableau Server (self-hosted on Windows or Linux). Licences are role based: Creator, Explorer and Viewer, plus a Display role for passive screens, with usage-based and capacity-based options documented as well. Tableau Next, a newer agentic analytics product built on the Salesforce Platform, is sold through the Tableau+ bundle.

Side by side

Aspect	Power BI	Tableau
Owner and ecosystem	Microsoft; part of Microsoft Fabric and Microsoft 365	Salesforce; Tableau Next runs on the Salesforce Platform and Data 360
Free option	Free Power BI Desktop; a free account can create but not share	Tableau Public for public data; trials of paid products
Paid licensing	Per user (Pro, Premium Per User) or Fabric F SKU capacity	Per role (Creator, Explorer, Viewer) in Standard or Enterprise editions, annual contracts; usage-based and capacity-based options
Desktop authoring	Power BI Desktop: Windows only	Tableau Desktop: Windows and macOS
Self-hosting	Power BI Report Server (Windows), licensed through SQL Server core licences or Fabric F64+ reserved capacity	Tableau Server on Windows or Linux
Modelling language	DAX measures plus Power Query (M) for data preparation	Tableau calculated fields, table calculations and LOD expressions; Tableau Prep for data preparation
SQL access	Native SQL statement in many connectors; Import, DirectQuery, Direct Lake and composite models	Custom SQL (a single SELECT) on most database connections; live connections or extracts
Semantic model reuse	Published semantic models reusable by other reports; XMLA endpoint on Premium, PPU and Fabric	Published data sources; Tableau Semantics with Tableau Next
Embedding	Power BI Embedded (A SKUs) or F and P capacities; app-owns-data needs no end-user licences	Embedding API v3 for Tableau Cloud, Server and Public; viewers licensed by role or usage
Main trade-off	Low entry price and deep Microsoft integration, but Windows-only authoring and a steep DAX learning curve	Flexible visual exploration and cross-platform authoring, but higher per-role list prices

Key differences

Licensing: per user and capacity versus per role

Power BI has three ways in. A free account lets someone build reports in My workspace but not share them. Pro is the per-user licence for publishing and sharing; Premium Per User adds larger models (Microsoft lists 100 GB), 48 refreshes a day and other Premium features. Organisations with many consumers buy a Fabric capacity (F SKUs): on F64 and larger, viewers with a free licence can consume content, while on smaller capacities every viewer still needs Pro or PPU. Microsoft is retiring the older Power BI Premium P SKUs in favour of F SKUs. Creators always need Pro or PPU to publish.

Tableau licenses by role. Creator includes Tableau Desktop and Tableau Prep Builder and can connect to new data and publish data sources; Explorer can author and edit in the browser using published data; Viewer can interact with dashboards, comment, subscribe and export. Tableau's licensing documentation also lists a Display role for broad passive audiences, usage-based licensing charged per analytical impression, and capacity-based Viewer blocks from July 2026. Roles are sold in Standard and Enterprise editions, and Tableau+ bundles Tableau Cloud with Tableau Next.

In our assessment, Power BI is usually the cheaper route for organisations that already pay for Microsoft 365 E5 (which includes Pro) or a Fabric capacity, while Tableau's role model gives more control over who can author, at a higher list price per seat. Model the cost with your own numbers of creators and viewers rather than relying on headline prices.

Authoring platforms and deployment

Power BI Desktop runs on Windows only: Microsoft lists Windows 10 or Windows Server 2016 or later as the minimum, and recommends a client version of Windows. It is updated monthly and only the latest version is supported. Mac users can author some content in the Power BI service in a browser, including DirectQuery semantic models, or use Windows through Azure Virtual Desktop or Windows 365, which Microsoft supports; other virtualised setups such as Citrix published apps are not supported.

Tableau Desktop runs on Windows and macOS, including Apple silicon; Tableau's technical specifications list macOS Ventura to Tahoe for recent releases. Self-hosted Tableau Server runs on Windows or on Linux (Tableau lists RHEL, Ubuntu LTS, Oracle Linux, Rocky Linux, AlmaLinux and Amazon Linux among supported distributions). Power BI's on-premises option, Power BI Report Server, is a Windows server product with its own optimised version of Power BI Desktop.

Data connections and SQL access

Power BI offers several storage modes. Import loads data into the in-memory engine and is Microsoft's recommended default; DirectQuery sends queries to the source at report time; Direct Lake reads Fabric lakehouse and warehouse tables; composite models and hybrid tables mix them. DirectQuery has documented limits, including a one million row cap on intermediate results (raisable on Premium capacities) and a four-minute query timeout in the service. For SQL users, connectors such as SQL Server, PostgreSQL, MySQL, Oracle, Snowflake, BigQuery and Redshift accept a native SQL statement under Advanced options, and Power Query's View Native Query shows the SQL that transformations fold into.

Tableau connects live or through extracts. Custom SQL is available for most database connections, with documented rules: it must be a single SELECT statement, Tableau wraps it in an outer query to add filters and grouping (which can affect performance), and parameters can replace only literal values, not table names. In our view, analysts who think in SQL often find Tableau's connect-and-drag workflow quicker to start with, and Power BI's Power Query more capable for repeatable data shaping.

Calculation languages: DAX and Power Query versus Tableau calculations

Power BI measures are written in DAX, a formula language that evaluates in a filter context. CALCULATE, which changes that context, is the function most SQL users need to learn first. This example from Microsoft's DAX reference returns revenue for blue products only:

-- DAX (Power BI measure)
Blue Revenue =
CALCULATE ( SUM ( Sales[Sales Amount] ), 'Product'[Color] = "Blue" )

Data preparation happens before the model, in Power Query (the M language), usually through the editor rather than hand-written code. Tableau calculated fields use a smaller expression language with aggregate functions, and level of detail (LOD) expressions control the grain of an aggregation independently of the view. The FIXED example from Tableau's help computes total sales per region whatever is on the canvas:

-- Tableau calculated field (FIXED level of detail expression)
{FIXED [Region] : SUM([Sales])}

A SQL reader will recognise the LOD as roughly a grouped subquery joined back to the detail rows. In our assessment, DAX has the steeper learning curve but gives one reusable, governed set of measures; Tableau calculations are easier to start with but tend to live in individual workbooks or published data sources.

Sharing, embedding and reuse

Power BI shares through workspaces and apps in the service, Teams, SharePoint and PowerPoint. For embedding in your own product, Microsoft documents embed for your customers (app owns data), which needs a capacity (A, EM, P or F SKU) but no Power BI licences for end users, and embed for your organisation, where viewers need Pro or PPU unless the capacity is F64 or larger. On Premium, PPU and Fabric, the XMLA endpoint exposes semantic models to Excel, SSMS, Tabular Editor, DAX Studio and third-party tools.

Tableau shares through projects on Tableau Cloud or Server, subscriptions and Tableau Mobile. The Embedding API v3 embeds views and web authoring in web pages through a <tableau-viz> web component or JavaScript, for Tableau Cloud, Server and Public; embedded viewers need a licence or a usage-based arrangement.

Ecosystems: Microsoft Fabric versus Salesforce

Power BI is one workload of Microsoft Fabric, so reports can read OneLake tables in Direct Lake mode and share capacity with Spark, the Fabric warehouse and pipelines. It also fits Microsoft 365 licensing and Microsoft Entra ID. See Databricks vs Microsoft Fabric for how Fabric capacity and throttling work.

Tableau sits in Salesforce. Salesforce describes Tableau Next as an agentic analytics platform built on the Salesforce Platform and Data 360, with Tableau Semantics as its semantic layer, generally available through the Tableau+ SKU. Tableau Cloud and Server continue alongside it. If your customer data already lives in Salesforce, that integration is a reason to favour Tableau; if it lives in Azure and Microsoft 365, the Fabric integration favours Power BI.

Pricing and licensing

Power BI. Microsoft's Power BI pricing page, checked 7 October 2026, lists Power BI Pro at USD 14.00 per user per month, paid yearly (Pro is also included in Microsoft 365 E5 and Office 365 E5). Premium Per User is a higher per-user tier, and Fabric capacity is sold separately through Azure, pay-as-you-go or reserved, at regional prices. Power BI Desktop and a free account cost nothing, but a free account cannot share. Prices exclude tax and may vary by country.

Tableau. Tableau publishes per-role list prices on its pricing page (Creator, Explorer and Viewer, billed annually); we could not load that page to confirm the current figures, so check them there. Creator is the most expensive role and Explorer and Viewer cost less per user, Enterprise edition costs more, every deployment needs at least one Creator, and Tableau Cloud and Tableau Server plans require annual contracts. Tableau+ and the usage-based and capacity-based options are quoted by sales. Prices exclude tax; confirm current figures and your region's currency on Tableau's page before budgeting.

Pricing checked on the vendors' official pages on 7 October 2026. Prices change; confirm before buying.

Where each one leads

Power BI strengths

Low per-user list price, and Pro is included in Microsoft 365 E5
One governed semantic model with DAX measures, reusable across reports and through the XMLA endpoint
Power Query provides repeatable data shaping with native SQL support in many connectors
Part of Microsoft Fabric, with Direct Lake access to OneLake tables
Embedding for customers needs a capacity but no end-user licences

Tableau strengths

Tableau Desktop runs on Windows and macOS
Self-hosted Tableau Server runs on Linux as well as Windows
Visual, drag-and-drop exploration with LOD expressions for controlling aggregation grain
Flexible licensing: roles, plus Display, usage-based and capacity-based options
Integration with Salesforce through Tableau Next and Data 360

Limitations

Power BI limitations

Power BI Desktop is Windows only; Mac authors need the browser or a Windows virtual desktop
DAX filter context has a steep learning curve for SQL users
Free-licence viewing requires an F64 or larger capacity; below that every viewer needs Pro or PPU
DirectQuery has documented limits, including a one million row intermediate result cap

Tableau limitations

Higher list price per author than Power BI Pro, with annual contracts only
Custom SQL must be a single SELECT and is wrapped in an outer query
Calculations are scattered across workbooks unless you manage published data sources carefully
Tableau Next and Tableau Semantics require the Tableau+ bundle

When to choose each

Choose Power BI if

Your organisation is standardised on Microsoft 365, Azure or Microsoft Fabric
You want a central semantic model of governed measures that many reports reuse
Cost per user matters and you have many consumers, or already own E5 or a Fabric capacity
You need to embed reports in a customer-facing app without licensing each end user

Choose Tableau if

Some of your authors work on macOS
Visual exploration by analysts matters more than a central modelling layer
You need to self-host the BI server on Linux
Your organisation is invested in Salesforce and Data 360

When neither is right

You want governed metrics defined in code and version control: look at Looker's LookML model; see Power BI vs Looker and Tableau vs Looker.
You need free, self-hosted dashboards over SQL databases: see Power BI vs Metabase, Metabase vs Apache Superset and the best open source BI tools.
You only need simple shared reports on Google data sources: Google's free Data Studio (formerly Looker Studio) may be enough; see Looker vs Looker Studio.

Final recommendation

Bottom line

Power BI is the default choice for Microsoft-centred organisations: a low per-user price, a strong semantic model and tight integration with Fabric, as long as your authors are on Windows and someone on the team learns DAX properly. Tableau earns its higher price where visual exploration, macOS authoring, Linux self-hosting or Salesforce integration matter. For SQL-heavy teams, both accept custom SQL, but neither replaces a well-modelled warehouse; see data warehouse vs database for the layer underneath.

Frequently asked questions

Does Power BI Desktop run on a Mac?

No. Microsoft lists Windows 10 or Windows Server 2016 or later as the minimum. Mac users can author some content in the Power BI service in a browser, or run Power BI Desktop in Azure Virtual Desktop or Windows 365, which Microsoft supports. Tableau Desktop runs natively on both Windows and macOS.

Is Power BI cheaper than Tableau?

Compare the current list prices directly: Microsoft lists Power BI Pro at USD 14 per user per month paid yearly (checked 7 October 2026), and Tableau lists its Creator role, which every Tableau deployment needs at least one of, on its own pricing page, which we could not load to confirm. Total cost depends on your mix of authors and viewers, Fabric capacity, Microsoft 365 E5 (which includes Pro) and negotiated discounts.

Can I write SQL in Power BI and Tableau?

Yes. Power BI connectors such as SQL Server, PostgreSQL, MySQL, Oracle and Snowflake accept a native SQL statement under Advanced options. Tableau supports custom SQL on most database connections, limited to a single SELECT statement. Calculations on top of the data use DAX in Power BI and Tableau's calculation language in Tableau.

Do Power BI viewers need a paid licence?

Usually. Viewers need Pro or Premium Per User unless the content is on a Fabric capacity of F64 or larger, where free-licence users can view it. When you embed reports in your own app for customers (app owns data), end users need no Power BI licence, but you need a capacity.

What is Tableau Next?

Salesforce describes Tableau Next as an agentic analytics platform built on the Salesforce Platform and Data 360, with Tableau Semantics as its semantic layer. It is generally available through the Tableau+ bundle and sits alongside Tableau Cloud and Tableau Server rather than replacing them.

Sources

Checked October 2026.

How we research comparisons: our editorial method.