Control plane, extended support, Auto Mode fees, EKS Capabilities, provisioned control-plane tiers, Hybrid Nodes, and the line that never appears on an invoice. A worked example for a small team and how to keep each number down.
From the Amazon EKS pricing page, October 2026. EC2, EBS, load balancers, NAT and data transfer are billed as usual on top.
Extended support is on by default. A cluster on Kubernetes 1.33 moved from $0.10 to $0.60 an hour on July 29, 2026, the day its standard support ended, and 1.34 does the same on December 2, 2026. For a company with dev, staging and production clusters plus a couple of experiments that nobody deleted, that is the difference between about $365 and $2,190 a month: for control planes alone.
Book a free 30-minute call. You'll leave knowing what we'd do first and roughly what it costs, whether or not you hire us.
Three clusters in standard support, Auto Mode running about six m5a.xlarge nodes in total, Argo CD as a managed capability with 20 applications: control planes about $219 a month, Auto Mode fees about $90, Argo CD about $44, plus the EC2 instances themselves at the On-Demand rate or less with Savings Plans. The same services on ECS with Fargate or Managed Instances pay nothing for the control plane and nothing for deployment tooling — the compute is comparable.
Kubernetes costs engineering time every four months, when a minor version ships and the oldest supported one ages out: reading deprecations, checking every add-on, CRD and Helm chart against the new version, upgrading the control plane, then the nodes. With Auto Mode and managed add-ons this is days rather than weeks — but it is never zero, and it lands whether or not the team has time. Price it in honestly when comparing against ECS, which has no version to upgrade.
Fewer clusters: namespaces and RBAC separate teams more cheaply than clusters do, unless you need hard isolation.
Stay in standard support: put each cluster's end-of-standard-support date — aws eks describe-cluster-versions returns it — into a scheduled check that opens a ticket ninety days out.
Let Karpenter, inside Auto Mode or run yourself, consolidate nodes and use Spot for interruptible work.
Delete experiment clusters; tag every cluster with an owner and an expiry.
Use Savings Plans for the steady EC2 part — they apply under Auto Mode and Managed Instances too.
Once a month: what changed in tech that month and what it means for you, from cloud and AI to security and the tools you build with. Every item links to its source. No sales pitch, unsubscribe in one click.
Book a free 30-minute call about it. Bring your setup and your bill; you'll leave knowing what we'd do first and roughly what it costs, whether or not you hire us.
Or keep reading
Every AWS Extended Support Charge in 2026–2027: RDS, Aurora, EKS, ElastiCache and OpenSearch
AWS now charges for running old versions on five managed services, each with its own formula: per vCPU for RDS and Aurora, six times the control plane rate for EKS, an 80–160% premium for ElastiCache, and up to the full instance price for old OpenSearch domains. The rates, the dates, and how to find every charge on your bill.
Read
An EKS Reference Architecture for 2026, With Diagrams: The Small-Team Version and the Regulated Version
What a production EKS setup looks like in 2026, layer by layer: edge, entry, cluster, platform add-ons, data, security and observability. A lean version for a small team on Auto Mode, and a regulated version with separate node pools, network policies and audit logging, with the choices behind each box.
Read
As of October 25, 2026. Provided as is, for general information, and accurate to the best of our knowledge on the date shown. Not legal, compliance, tax, financial or other professional advice. Prices, limits and vendor services change; check current vendor documentation before relying on anything here. Terms of use.