Red Hat OpenShift AI Self-Managed 3.3

Release notes

Features, enhancements, resolved issues, and known issues associated with this release

Assess and plan for migration from OpenShift AI 2.25.4 (and later) to 3.3 (latest)

Assess and plan for migration to Red Hat OpenShift AI Self-Managed 3.3 (latest)

Get started with projects, workbenches, and pipelines in OpenShift AI

Get set up to create projects, launch workbenches, and deploy your first model on OpenShift AI

OpenShift AI tutorial - Fraud detection example

Use OpenShift AI to train and deploy an example fraud detection model

Prepare your platform and hardware for Red Hat AI

Review compatibility matrices, accelerator support, deployment targets, and update policy prior to installation

Deploy or decommission OpenShift AI on your cluster

Install via Operator or CLI, enable required components, verify the deployment, and cleanly uninstall when needed

Deploy or decommission OpenShift AI in disconnected environments

Install via Operator or CLI, enable required components, verify the deployment, and cleanly uninstall when needed

Manage migration from OpenShift AI 2.25.4 to 3.3.2

How to initiate migration to Red Hat OpenShift AI Self-Managed 3.3.2

Provision secure workbenches and custom images for teams

Use CRDs or dashboard to publish images and provision resourced workbenches

Administer OpenShift AI platform access, apps, and operations

Administer access, apps, resources, and accelerators; maintain logging, audit, and backups

Deliver consistent ML features to models with Feature Store

Use Feature Store to define, store, and serve reusable machine learning features to models

Provision hardware configurations and resources for projects

Enable supported hardware configurations for your workloads

Configuring your model-serving platform

Configure your model-serving platform in Red Hat OpenShift AI Self-Managed

Build AI/Agentic Applications with Llama Stack

Operate Llama Stack: activate the operator and expose OpenAI‑compatible RAG APIs

Configure user access, storage, and telemetry in OpenShift AI

As an administrator, configure user access, customize the dashboard, and manage specialized resources for data science and AI engineering projects

Managing and monitoring models

Manage and monitor models in Red Hat OpenShift AI Self-Managed

Choose production‑ready OpenShift AI APIs

Plan which APIs to build on and how to upgrade with minimal risk by mapping each OpenShift AI endpoint to a support tier that defines stability and deprecation timelines

Manage and govern model catalog sources

Manage and govern model catalog sources in Red Hat OpenShift AI Self-Managed

Provision and secure access to model registries

Use the OpenShift AI dashboard to create registries, set access with RBAC groups, and manage model and version lifecycle so teams can register, share, and promote models to serving with traceability

Register, version, and promote models with the model registry

Store, version, and promote models with metadata for cross‑project sharing and traceability

Discover, evaluate, register, and deploy models from the model catalog

Use the model catalog to discover, evaluate, register, and deploy models for rapid customization and testing

Experimenting with models in the gen AI playground

Using the AI playground to experiment with RAG using models from your catalog

Accelerate data processing and training with distributed workloads

Distribute data and ML jobs for faster results, larger datasets, and GPU‑aware auto‑scaling and monitoring

Connect your workbench to S3-compatible object storage

Create a connection, configure an S3 client, and list, read, write, and copy objects from notebooks

Organize projects, collaborate in workbenches, and deploy models

Organize projects, collaborate in workbenches, build notebooks, train/deploy models, and automate pipelines

Use the Red Hat data science IDE images effectively

Launch a workbench, pick an IDE, and develop with prebuilt images or custom environments

Build, schedule, and track machine learning pipelines

Define KFP‑based pipelines, version and schedule runs, and track artifacts in S3‑compatible storage

Enable and manage connected applications from the OpenShift AI dashboard

Enable applications, connect with keys, remove unused tiles, and access Jupyter from the dashboard

Customize models to build generative AI applications

Customize AI models that are specific to your domain-specific use case, from setting up your development environment to building and deploying models for use in generative AI applications

Evaluating AI systems with LM-Eval

Configure LMEvalJobs, select tasks, run evaluations, and retrieve metrics to compare model performance

Ensuring AI safety with guardrails

Orchestrate detectors to filter LLM inputs/outputs, auto‑configure security, and expose guarded endpoints

Monitoring your AI Systems

Monitor model bias and data drift by configuring metrics, thresholds, and visualizations in OpenShift AI

Deploy large models using the single-model serving platform (KServe RawDeployment)

Deploy models with KServe—choose RawDeployment or Knative, set resources and runtimes, and expose authenticated endpoints

Govern LLM access with Models-as-a-Service

Govern LLM access with Models-as-a-Service in Red Hat OpenShift AI Self-Managed

Red Hat OpenShift AI lifecycle

Red Hat OpenShift AI Supported Configurations