Databricks Release Notes

Unity Gateway estimates external model spend using each provider's published list prices. If your organization has negotiated different pricing with a provider, you can now configure a price multiplier on the model provider service so that spend estimates reflect those rates.

Price multipliers only apply to external models accessed through model provider services, not Databricks-hosted partner models.

See Configure a price multiplier.

Configuring telemetry for your Databricks apps is now generally available. Databricks Apps telemetry collects traces, logs, and metrics and persists them to Unity Catalog tables using the OpenTelemetry (OTel) protocol. After you enable app telemetry, Databricks automatically captures system logs and usage events, and you can add custom instrumentation using the OpenTelemetry SDK for your framework.

See Configure telemetry for Databricks Apps.

Databricks Apps is now on by default for workspaces with the compliance security profile enabled.

See Databricks Apps.

A Genie One skill can now include multiple files and sub-folders, up to three folders deep and 50 files or folders per skill. Skills can load text files only. See Skills.

Unity Gateway now supports all compliance security profile standards supported by Databricks and is available by default for workspaces with the compliance security profile enabled.

For a full list of supported compliance standards, see Compliance security profile.

Metastore-level ABAC policies are in Beta in Unity Catalog. You can now attach an ABAC policy (row filter, column mask, GRANT, or DENY) at the metastore level, so a single policy applies across every catalog in the metastore, including catalogs created later. This gives governance teams one place to enforce organization-wide data protection rules without replicating a policy for each catalog.

For details, see Metastore-level ABAC policies (Beta).

You can now use base environments to manage Python dependencies on classic compute. You can use Databricks-provided environments or custom workspace-level environments. See Manage dependencies using environments.

On September 19, 2026, Anthropic models will be removed from Databricks on AWS GovCloud and AWS GovCloud DoD workspaces that have the DoD Impact Level 5 (IL5) compliance standard selected. After this date, IL5 workspaces can't create or use endpoints that serve Anthropic models. Migrate any affected endpoints to another supported model before September 19, 2026.

See DoD IL5.

Genie One now supports account-only users, so users in a registered identity provider can access account-level Genie One without being assigned to a workspace or granted consumer access. When enabled, eligible users are enrolled automatically. Unity Catalog continues to govern what each user can see. See Account-only user access to Genie One.

Databricks now automatically deletes a service principal's OAuth client secret after 90 days without use, consistent with unused access token cleanup. If a deleted secret is still needed, create a new OAuth secret.

Managed agent memory and managed agent sessions are in Beta. Databricks now provides two fully managed, Lakebase-backed stores for agent state, both usable from agents built on any framework:

• Managed agent memory gives agents durable, long-term memory that persists across conversations, with semantic search. See Managed agent memory.
• Managed agent sessions store durable conversation history that an agent reads at the start of a turn and appends to as it runs. See Managed agent sessions.

You can now use a single VARIANT function to mask columns of type ARRAY and MAP, in addition to STRUCT and other types. This requires Databricks Runtime 19 or above. See Masking multiple column types with a single function.

Sharing and reading foreign Delta tables with OpenSharing is now generally available. Providers can share Delta-format tables federated from OneLake, a Hive metastore, or AWS Glue, without copying data into Databricks. See Add foreign Delta tables to a share.

Data Classification can now scan Unity Catalog views in addition to tables, classifying columns for sensitive data using the same detection engine. This is a Beta feature. See Classify views.

You can now manage Lakebase snapshots programmatically. The snapshots API lets you create, get, list, and delete point-in-time snapshots of a project branch, and restore a snapshot by creating a new branch from it. This API is in Beta and available to all Lakebase workspaces. See Create a snapshot.

Window measures support unitless offset and trailing/leading ranges on a consecutive integer order column. Use a numeric index for non-calendar periods such as fiscal weeks or 4-4-5 periods. See Window measures on a numeric index column.

When a Genie response includes a chart, the visualization now appears as an image in the Slack reply. If you installed the Databricks Genie app for Slack previously, reinstall it to enable this. See Visualizations.