Skip to main content
unity catalog pricingunity catalog costis unity catalog freedatabricks unity catalog pricingpredictive optimization costDatabricks Unity Catalog

Unity Catalog Pricing: Is It Free and What Does It Cost?

Unity Catalog is free in the sense that Databricks charges no license, seat or add-on fee for it, but it is not quite free in practice, because the features that make it useful run on Databricks compute that you pay for in DBUs. The catalog itself costs nothing. Predictive optimization, data quality monitoring, queries against system tables, Unity Catalog functions and sharing across regions all show up on the bill, and all of it requires the Premium tier or above.

This guide lists where those costs land and how to find them in your own usage data. It covers Unity Catalog specifically; for Databricks compute pricing in general, including cluster sizing and cloud VM costs, see Databricks pricing explained. Unless stated otherwise, rates are US dollar list prices for Azure Databricks Premium in East US, taken from the Azure Retail Prices API on 16 November 2026. AWS and Google Cloud rates differ, so check the Databricks pricing page for your cloud.

TL;DR

Databricks says Unity Catalog is included at no additional cost, with no separate license or per-seat fee. The costs are indirect. Unity Catalog is not available on the retired Standard tier, so Premium DBU rates are the floor. Predictive optimization, on by default for newer accounts, and data quality monitoring run on serverless compute billed in DBUs. System tables are free to store for their default retention, but querying them uses a SQL warehouse or cluster. Sharing within a region has no egress cost, while cross-region and cross-cloud transfer is billed. The open-source Unity Catalog server has no fee at all, but you operate it and it does not include the managed service's features. Track all of this in system.billing.usage. For a cross-platform comparison, see data catalog pricing.

Is Unity Catalog Free?

The answer from Databricks is direct: its own pricing answer page states that Unity Catalog is free and included with Databricks at no additional cost. There is no license, no per-seat charge and no add-on SKU for the catalog, its grants, its lineage or its audit log. Since November 2023, every new workspace has Unity Catalog enabled automatically.

The same page lists the features that do consume compute: predictive optimization and data profiling, Unity Catalog functions, Genie agents, Databricks SQL warehouses and AI Search indexes. That is the "not quite" part. A catalog with no fee still produces line items, because the governance work it automates has to run somewhere, and on Databricks that somewhere is metered in DBUs.

Where Unity Catalog Costs Land WHERE UNITY CATALOG COSTS LAND Unity Catalog license $0, included with Databricks Premium tier Standard is retired, so Premium DBU rates are the floor Serverless features Predictive optimization, data quality monitoring System tables Free to store; queries run on a warehouse or cluster Sharing egress No egress cost in the same region; cross-region billed Cloud storage Managed tables sit in your storage, billed by your cloud No catalog fee, but four of these five lines appear on the Databricks or cloud invoice every month.
Click to enlarge

Where Unity Catalog Costs Show Up

The table maps each Unity Catalog feature to how it is billed. On your own bill, each line carries a value in the billing_origin_product column of the billable usage system table, such as PREDICTIVE_OPTIMIZATION or DATA_QUALITY_MONITORING; the query further down groups costs by it.

FeatureHow it is billedAzure East US list rate
Catalog, grants, lineage, audit logIncluded, no fee$0
Predictive optimizationServerless jobs compute$0.45 per DBU
Data quality monitoring and data profilingServerless DBUsServerless rate for the product
Querying system tablesWhatever compute runs the query$0.70 per DBU on serverless SQL, $0.55 on SQL Pro
Unity Catalog functionsServerless general computeServerless rate for the caller
Sharing to other regions or cloudsStorage vendor egress, or Databricks for SecureConnectPer GB, by cloud and route
System table retention beyond 395 daysStorage, once configurable retention is generally availableNot yet published

The serverless jobs rate above is the Azure meter named "Premium Automated Serverless Compute". For orientation, the same API lists Premium Jobs Compute on classic clusters at $0.30 per DBU and serverless notebooks ("Interactive Serverless Compute") at $0.95. Serverless rates include the underlying cloud VMs, so they are not directly comparable to classic rates, which come with a separate VM bill.

The Premium Tier Requirement

The oldest hidden cost of Unity Catalog is the tier it needs. Unity Catalog was never available on the Standard tier; Microsoft's Standard tier page lists it as a Premium feature. For years that meant governance had a price: the jump from Standard to Premium DBU rates.

That choice is now gone. On AWS, Databricks upgraded remaining Standard workspaces to Premium on 1 October 2025. On Azure, new workspaces have had to be Premium since 1 April 2026, and remaining Standard workspaces are being upgraded through a phased rollout that started on 1 October 2026. Premium is therefore the floor for everyone, and Unity Catalog comes with it. If you are still on a Standard workspace, the upgrade is the single largest Unity Catalog-related cost change you will see, and it applies to every workload, not only governance.

Serverless Features That Bill DBUs

Three Unity Catalog features run on serverless compute without anyone starting a cluster, which is convenient and easy to overlook.

Predictive optimization. Predictive optimization runs OPTIMIZE, VACUUM, ANALYZE and index refresh operations on Unity Catalog managed tables automatically, billed through a serverless jobs SKU. Databricks documents it as enabled by default, and its pricing answer page notes that it is on by default for accounts created on or after 11 November 2024. It does not run on external tables. It usually saves money overall, because compacted tables with fresh statistics are cheaper to query, but the maintenance itself is a real line item that grows with the number and churn of managed tables.

Data quality monitoring. Data quality monitoring, which includes data profiling (formerly Lakehouse Monitoring), runs on serverless compute and is billed as serverless DBUs under the DATA_QUALITY_MONITORING billing product. Cost scales with the number of monitored tables, their size and the refresh schedule, so a daily profile on a large table costs far more than a weekly one on a small table.

Unity Catalog functions and Genie. Functions registered in Unity Catalog run on serverless general compute when called, and Genie agents use serverless SQL. Neither is a governance cost as such, but both are billed through features that live in the catalog.

As an illustration only: if predictive optimization used 200 DBUs of serverless jobs compute in a month, at $0.45 per DBU that would be $90. Real consumption depends entirely on your tables, so measure it in the usage table rather than estimating from examples.

System Tables Costs

System tables are where Unity Catalog keeps lineage, audit logs, billing usage, query history and more. Databricks documents that system tables are free to use and that you are charged only for the compute used to query them. Most system tables keep 395 days of data for free, query history keeps 365 days, and the billable usage and pricing tables are kept indefinitely.

Two details matter for budgeting. First, every dashboard, alert or audit report built on system tables runs on a SQL warehouse or cluster, so a busy governance dashboard refreshing every few minutes has a real compute cost. Second, configurable retention is in Beta, with a range from 30 to 3,650 days, and Databricks says data kept beyond 395 days will incur storage charges once that feature is generally available. If compliance requires multi-year audit history, budget for that storage now or copy what you need into your own tables.

Delta Sharing and Egress Costs

Delta Sharing, now documented by Databricks as OpenSharing, is governed through Unity Catalog and has no license fee either. The cost is data transfer. Databricks states that sharing within a region incurs no egress cost, and that storage and network transfer are charged by the storage vendor, or by Databricks when the provider uses SecureConnect. Databricks' data transfer pricing page confirms that egress to a different region, to the internet or to other clouds is billed per GB.

For providers who share widely across regions or clouds, egress can become the largest sharing cost. Databricks' egress guidance suggests keeping recipients in the same region where possible, replicating data to the recipient's region, and using Cloudflare R2 storage, which charges no egress fees (this does not cover view sharing). For Databricks-to-Databricks sharing, compute is charged by Databricks and the cost depends on the recipient type and sharing scope.

External Engine Access Costs

Unity Catalog lets engines outside Databricks read and write its tables through the Iceberg REST Catalog API and through credential vending. In that pattern the external engine reads the files directly from your cloud storage with short-lived credentials, so the query runs on the external engine's compute rather than on Databricks compute. What you still pay for is the external engine itself and your cloud provider's storage requests and any cross-region transfer. This is one way to let a second engine use governed tables without paying for Databricks compute on every read, with the limit that tables protected by row filters or column masks are not vended. The catalog side of that trade-off is covered in the Iceberg catalog comparison.

Open Source vs Managed Unity Catalog

The open-source Unity Catalog server, released under the Apache 2.0 license and hosted by LF AI & Data, has no fee of any kind. The cost is operational: you host it, secure it, back it up and upgrade it, and it is still a pre-1.0 project. It is also a separate implementation from the Databricks-managed service. Features documented for the managed service, such as automatic lineage capture, system tables, predictive optimization and data quality monitoring, are documented for the managed service, not for the open-source server.

In practice the two serve different needs. The open-source server makes sense for teams that want a Unity Catalog-compatible catalog for engines outside Databricks. The managed service makes sense for teams already running on Databricks, where the catalog comes with the platform and its costs show up as the compute described above.

How to Monitor Unity Catalog Costs

The system.billing.usage table records every DBU with a billing_origin_product, and joining it to system.billing.list_prices gives list-price cost, as Databricks shows in its serverless cost monitoring guide. A simple monthly view of the governance-related products looks like this:

SELECT
  date_trunc('month', u.usage_date) AS month,
  u.billing_origin_product,
  SUM(u.usage_quantity) AS dbus,
  SUM(u.usage_quantity * p.pricing.default) AS list_cost
FROM system.billing.usage u
JOIN system.billing.list_prices p
  ON u.sku_name = p.sku_name
 AND u.cloud = p.cloud
 AND u.usage_start_time >= p.price_start_time
 AND (p.price_end_time IS NULL OR u.usage_start_time < p.price_end_time)
WHERE u.billing_origin_product IN (
  'PREDICTIVE_OPTIMIZATION', 'DATA_QUALITY_MONITORING',
  'DATA_CLASSIFICATION', 'FINE_GRAINED_ACCESS_CONTROL', 'DATA_SHARING')
GROUP BY ALL
ORDER BY month, list_cost DESC;

List cost is a ceiling: negotiated discounts and commit plans lower the real number. Run the query on a small serverless SQL warehouse, or schedule it as a job, since the query itself is billed like any other.

Unity Catalog Pricing vs Other Catalogs

Catalog pricing models differ more than catalog features do. Microsoft Purview bills its Unified Catalog per governed asset per day. The AWS Glue Data Catalog is free for the first million objects and requests, then $1 per 100,000 objects per month. Snowflake bills its Horizon Iceberg REST endpoint at 0.5 credits per million API calls, with billing scheduled to begin in the second half of 2026. Unity Catalog is the only one of these with no catalog meter at all, which makes it cheap inside Databricks and irrelevant outside it.

That last point is the real cost question for most teams. A free catalog that covers one platform still leaves Snowflake, Fabric, operational databases and BI tools to be governed somewhere else. For the trade-offs between Unity Catalog and Purview, see Unity Catalog vs Microsoft Purview, and for third-party catalogs, see data catalog pricing.

Where Dawiso Fits

Dawiso does not monitor Databricks costs; that stays with the billing system tables and your cloud cost tools. What Dawiso adds is the governance that a platform catalog does not cover. The Dawiso Databricks connector reads Unity Catalog through its system tables and REST API, bringing in catalogs, schemas, tables, views, metric views, volumes, functions and ML models, with owners, tags, column masks and row filters, and column-level lineage from system.access.column_lineage.

Dawiso then puts that metadata next to other platforms in one data catalog, with a business glossary and stewardship on top. Dawiso is priced per user, not per governed asset or per API call, so the cost of documenting more tables does not grow with the size of the estate.

Conclusion

Unity Catalog pricing is simple at the top and detailed underneath. The catalog has no fee. The costs are the Premium tier it requires, serverless DBUs for predictive optimization and data quality monitoring, compute for querying system tables, egress for sharing across regions, and storage for long retention once that is billed. Measure them with system.billing.usage, decide which automated features earn their keep, and plan for the governance work that sits outside Databricks. For an overview of the catalog itself, see Databricks Unity Catalog.

Sources

Prices change and vary by cloud and region. All sources were read on 16 November 2026.

See it in action

Data & Analytics Catalog

Create a unified view of your data assets and gain insights faster with automated data discovery.

A cookie a day keeps bad UX away.

We use cookies to personalize content, ads and to analyze our traffic. We also share information about your use of our site with our advertising and analytics partners who may combine it with other information that you've provided to them or that they've collected from your use of their services. By clicking "Accept All", you allow us to use cookies for analytics and ads via Google Tag Manager. You can also customize cookies.

Customize Consent Preferences

We use cookies to personalize content, ads and to analyze our traffic. We also share information about your use of our site with our advertising and analytics partners. Privacy Policy

Necessary cookies allow core website functionality such as user login and account management. The website cannot be used properly without strictly necessary cookies.

Functionality cookies are used to remember visitor information on the website, eg. language, timezone, enhanced content.

Analytics cookies are used to see how visitors use the website, eg. analytics cookies. Those cookies cannot be used to directly identify a certain visitor.

We use Microsoft Clarity to see how you use our website (including heatmaps and session replays) so we can improve it. By using our site, you agree that we and Microsoft can collect and use this data. See our Privacy Policy for details.

Advertisement cookies are used to identify visitors between different websites, eg. content partners, banner networks. Those cookies may be used by companies to build a profile of visitor interests or show relevant ads on other websites.