Databricks Runtime release notes versions and compatibility
This article lists all Databricks Runtime releases and the schedule for supported releases. Each Databricks Runtime version includes updates that improve the usability, reliability, performance, and security of the Databricks platform.
To learn about the Databricks Runtime support lifecycle, generally available releases, and Beta releases, see Databricks support lifecycles. For information on maintenance updates issued for Databricks Runtime releases, see Databricks Runtime maintenance updates.
Supported Databricks Runtime LTS releases
The following table lists supported Databricks Runtime long-term support (LTS) version releases in addition to the Apache Spark version, release date, and end-of-support date. For optimal lifespan, use a Databricks Runtime LTS version.
Note
LTS means this version is under long-term support. See Databricks Runtime LTS version lifecycle.
Version | Variants | Apache Spark version | Release date | End-of-support date |
---|---|---|---|---|
15.4 LTS | - Databricks Runtime 15.4 LTS - Databricks Runtime 15.4 LTS for Machine Learning |
3.5.0 | Aug 19, 2024 | Aug 19, 2027 |
14.3 LTS | - Databricks Runtime 14.3 LTS - Databricks Runtime 14.3 LTS for Machine Learning |
3.5.0 | Feb 1, 2024 | Feb 1, 2027 |
13.3 LTS | - Databricks Runtime 13.3 LTS - Databricks Runtime 13.3 LTS for Machine Learning |
3.4.1 | Aug 22, 2023 | Aug 22, 2026 |
12.2 LTS | - Databricks Runtime 12.2 LTS - Databricks Runtime 12.2 LTS for Machine Learning |
3.3.2 | Mar 1, 2023 | Mar 1, 2026 |
11.3 LTS | - Databricks Runtime 11.3 LTS - Databricks Runtime 11.3 LTS for Machine Learning |
3.3.0 | Oct 19, 2022 | Oct 19, 2025 |
10.4 LTS | - Databricks Runtime 10.4 LTS - Databricks Runtime 10.4 LTS for Machine Learning |
3.2.1 | Mar 18, 2022 | Mar 18, 2025 |
9.1 LTS | - Databricks Runtime 9.1 LTS - Databricks Runtime 9.1 LTS for Machine Learning |
3.1.2 | Sep 23, 2021 | Dec 19, 2024 |
All supported Databricks Runtime releases
The following table lists the Apache Spark version, release date, and end-of-support date for supported Databricks Runtime releases. For optimal lifespan, use a Databricks Runtime LTS version.
MLflow-Databricks Runtime compatibility matrix
This section lists Databricks Runtime ML versions and their respective MLflow versions.
Databricks Runtime ML version | MLflow version |
---|---|
16.0 | 2.15.1 |
15.4 LTS | 2.13.1 |
15.3 | 2.11.3 |
15.2 | 2.11.3 |
14.3 LTS | 2.9.2 |
14.1 | 2.7.1 |
13.3 LTS - 14.0 | 2.5.0 |
12.2 LTS | 2.1.1 |
11.3 LTS | 1.29.0 |
10.4 LTS | 1.24.0 |
9.1 LTS | 1.20.2 |
Feature Engineering compatibility matrix
This section lists Databricks Runtime ML versions and their respective Feature Engineering and Workspace Feature Store client versions.
Databricks Runtime ML version | databricks-feature-engineering version |
databricks-feature-store version |
---|---|---|
16.0 | 0.7.x | None |
15.4 LTS | 0.6.x | None |
15.3 | 0.5.x | None |
15.2 | 0.4.x | None |
14.3 LTS | 0.2.x | None |
14.1 | 0.1.x | 0.15.1 |
13.3 LTS | 0.1.x | 0.14.1 |
12.2 LTS | Not supported | 0.10.0 |
11.3 LTS | Not supported | 0.7.0 (requires MLflow < 2.0) |
10.4 LTS | Not supported | 0.3.8 (requires MLflow < 2.0) |
9.1 LTS | Not supported | 0.3.4 (requires MLflow < 2.0) |
Apache Spark migration guidance
Find Spark-specific migration information in the Apache Spark documentation. The migration information for each Spark version can be found at a URL like the following:
https://spark.apache.org/docs/<version>/migration-guide.html
.
Replace <version>
with the Spark version included in the Databricks Runtime version you’re migrating to. For example, the URL with migration information for Spark 3.5.0, included in Databricks Runtime 14.3 LTS, is https://spark.apache.org/docs/3.5.0/migration-guide.html.
Beta releases
There are no Databricks Runtime Beta releases at this time.
Unsupported releases
For information on unsupported Databricks Runtime version release notes, see End-of-support Databricks Runtime release notes. The unsupported Databricks Runtime versions have been retired and might not be updated.