Kubeflow AI Reference Platform 26.03 delivers key improvements in scalability, security, and operational efficiency. It reduces per-namespace overhead, enhances multi-tenant configurations, and increases reliability for large-scale Kubernetes deployments.

This release adopts a calendar-based versioning model (Year.Month.Patch), with two primary releases annually and optional patches. Community support is best-effort for approximately six months, with additional commercial support options available. Regular upgrades are recommended to take advantage of continuous security and performance enhancements.

Highlight features

  • Kubernetes 1.34+
  • Kubeflow Pipelines 2.16.0, Spark operator 2.5.0 Model registry v0.3.5, Kserve Web Application v0.16.1
  • Compatibility of Kubeflow Pipelines v1 and v2 with PSS restricted
  • Extended Kserve test with authentication and authorization from inside and outside the cluster as well as non-knative / raw deployments
  • Simplified installation and automatic installation of the right Kustomize and Kubectl versions
  • Installation steps tested and based on our CI, easier in-place updates (optimized PDBs)
  • Cleanup of all synchronization steps for faster releases / updates of dependencies
  • Knative 1.20, cert-manager 1.19.4, Oauth2-proxy v7.14.3, Dex 2.45.0
  • Fix networkpolicies for cert-manager, knative-serving, istio-system, dex, oauth2-proxy

Kubeflow Platform (Manifests & Security)

The Kubeflow Platform Working Group focuses on simplifying Kubeflow installation, operations, and security. See details below.

Manifests:

  • Documentation updates that make it easier to install, extend and upgrade Kubeflow
  • For more details and future plans please check 26.06 roadmap.
Notebooks Dashboard Pipelines Katib Trainer KServe Model Registry Spark SDK
1.10.0 1.10.0 2.16.0 0.19 2.1.0 0.17.0 0.3.7 2.5.0 0.4.0
Kubernetes Kind Kustomize Cert Manager Knative Istio Dex OAuth2-proxy
[1.35+] (https://github.com/kubernetes/kubernetes/releases/tag/v1.35.3) 0.30.1 5.8.1 1.20.2 1.21.0 1.29.1 2.45.1 7.15.1

Security:

Pipelines

Model Registry

Training Operator (Trainer) & Katib

Spark Operator

KServe

Kubeflow SDK

Dashboard and Notebooks

How to get started with 26.03

Visit the Kubeflow AI Reference Platform 26.03 release page or head over to the Getting Started and Support pages.

Join the Community

We would like to thank everyone who contributed to Kubeflow 26.03, and especially Tarek Abouzeid for his work as the v26.03 Release Manager. We also extend our thanks to the entire release team and the working group leads, who continuously and generously dedicate their time and expertise to Kubeflow.

Release team members : Tarek Abouzeid, Anya Kramar, Andy Stoneberg, Humair Khan, Matteo Mortari, Adysen Rothman, Jon Burdo, Milos Grubjesic, Vraj Bhatt, Dhanisha Phadate, Alok Dangre

Working Group leads : Andrey Velichkevich, Julius von Kohout, Mathew Wicks, Matteo Mortari

Kubeflow Steering Committee : Andrey Velichkevich, Julius von Kohout, Yuan Tang, Johnu George, Francisco Javier Araceo

You can find more details about Kubeflow distributions here.

Want to help?

The Kubeflow community Working Groups hold open meetings and are always looking for more volunteers and users to unlock the potential of machine learning. If you’re interested in becoming a Kubeflow contributor, please feel free to check out the resources below. We look forward to working with you!