Stay up to date with the latest scikit-learn updates, new features, and improvements. Here's what changed in 2026.
learn has seen a strong release cadence through 2024â
4 (January 2024) introduced native missing
value support in decision trees and random forests, TunedThresholdClassifierCV for post
hoc decision
Version
5 (June 2024) graduated metadata routing from experimental, expanded Array API support to more estimators for GPU
Version
6 (December 2024) delivered experimental support for free
threaded CPython (PEP 703) enabling true multi
threaded parallelism without the GIL, further broadened Array API coverage for hardware
accelerated backends, added real
time validation via dataclass
Across these releases, the metadata routing API has matured significantly, allowing users to route sample weights, groups, and other metadata through nested pipelines and cross
The project continues to invest in making scikit
learn the bridge between classical ML and modern hardware through the Array API initiative.
scikit-learn is actively developing new features. Check back regularly or follow their official channels for the latest updates.
Follow their release notes for instant updates
Test the latest updates in your workflow
Help shape future development with your input
Try out all the new features and improvements that make scikit-learn even more powerful.
Changelog last updated March 2026