Azure previews Argo CD in AKS portal
Argo CD arrives in AKS portal public preview to simplify GitOps for multi‑cluster AI workloads
A laptop displaying code and a notepad sit on a conference table facing a wall monitor showing a network diagram. © The GPU Trade Inc 2026
Microsoft announced a public preview of an Argo CD extension for Azure Kubernetes Service (AKS), bringing a guided GitOps experience into the AKS portal and Azure Arc‑enabled clusters. This move aims to reduce setup friction for teams managing containerized workloads at scale.
Argo CD is an open‑source GitOps continuous delivery tool that treats Git as the single source of truth for Kubernetes configuration. Microsoft will deliver Argo CD as a managed cluster extension, so teams can install and operate it as an Azure resource rather than hand‑rolling upstream installs.
The new portal experience offers a guided flow to enable Argo CD on an AKS cluster, configure identity and registry integrations, and expose the Argo CD UI for monitoring and application management. Microsoft’s announcement and Azure updates list the portal entry as a public preview item dated May 2026.
Microsoft highlights enterprise identity and security as a core benefit. The AKS engineering team published a step‑by‑step guide showing how Argo CD running as an AKS extension can integrate with Microsoft Entra ID for single sign‑on and workload identity federation. That integration removes some long‑lived secrets and ties Argo CD permissions to corporate identities.
The extension is delivered with Azure‑hardened images and an opt‑in automatic patching model, which Microsoft says reduces CVE exposure and simplifies lifecycle maintenance for operators running production clusters. Those hardening and patching features are part of the preview offering.
On feature parity, Microsoft says the AKS Argo CD extension remains aligned with upstream Argo CD and supports production patterns such as high availability, hub‑and‑spoke/topology for multi‑cluster fleets, and ApplicationSet for large‑scale application delivery. That makes it suitable for organizations operating many clusters from a central GitOps control plane.
For teams running AI workloads, the timing matters. AKS has been positioning itself for GPU‑accelerated and LLM inference workloads in recent months, with blog posts and guidance on scaling multi‑node LLM inference and running agentic workloads on AKS. Microsoft and partners say a managed Argo CD experience can speed deployments and policy‑driven updates across fleets of inference and agent clusters.
Operationally, Microsoft provides several ways to enable the extension today. You can install Argo CD on AKS using the Azure CLI or as a cluster extension resource (ARM/Bicep/Terraform), and the official tutorials note that the extension uses the community Helm chart starting at the 1.0.0‑preview release. Portal management of the extension is being rolled out as part of the preview experience.
Microsoft’s documentation and tutorials include practical steps and sample Terraform code to automate the Entra ID integration and install the Argo CD extension. The AKS engineering walkthrough emphasizes automating identity, ingress, and registry permissions to avoid manual, error‑prone setup.
A few caveats: Argo CD for AKS is in public preview, which means it’s not yet generally available and preview terms apply. Some configuration keys changed in the preview (the extension now uses the community Helm chart), so teams upgrading from an earlier private preview or OSS installs should follow the migration guidance in Microsoft’s tutorials.
For platform teams running multi‑cluster LLM inference or fleets of agent nodes, the promise is simpler onboarding and lifecycle operations. Managed Argo CD reduces manual patching, centralizes identity, and pairs with AKS features for GPU and autoscaling to help maintain consistent delivery workflows across clusters. Still, operators should validate performance, RBAC, and networking for AI inference topologies during the preview.
Microsoft is asking customers to try the preview and provide feedback through the usual Azure channels; documentation and tutorials are already live. For those ready to experiment, the quickest path is enabling the extension with Azure CLI or Terraform and following the AKS engineering blog’s Entra ID walkthrough.