Kubernetes cost calculator
Estimate your monthly and annual Kubernetes bill on AWS, GCP or Azure. It covers nodes, load balancer, API gateway, egress and managed services, broken down by component. Free, no signup, and it runs entirely in your browser.
Breakdown
- Compute (nodes)
- Cluster management
- Load balancer
- Request processing
- Egress
- Managed database
- Managed cache
3 quick wins to cut this
Cheaper compute could save up to $0 / month.
You’re already on spot, the cheapest compute tier. The wins below still apply to egress and right-sizing.
- Spot / preemptible nodes Run stateless and batch workloads on spare capacity for up to ~70% off compute.
- Committed-use / reserved Commit to steady-state nodes for ~40% off, without changing anything technical.
- Right-size & cut egress Most clusters are over-provisioned and leak egress. Matching nodes to real load and tightening data transfer is usually the fastest win.
Want the real number, and a plan to cut it?
This is a ballpark. Send me your actual setup and you’ll get a precise figure plus a concrete plan to bring it down. First call is 30 minutes, no charge.
Email me →Assumptions. Figures are representative on-demand list prices (Linux, general-purpose instances, a common EU/US region) and are estimates only. Your real bill depends on region, commitments, detailed data-transfer and service tiers. Rates last updated 2026-06-24.
Estimating LLM API spend too? Try the LLM cost calculatorFrequently asked questions
How is Kubernetes cost calculated?
The bulk is compute: number of nodes × the node’s hourly price × hours running per month. On top of that come the managed control plane (cluster fee), load balancer and request processing for ingress/API gateway, internet egress per GB, and any managed databases or caches. This tool sums those components for AWS, GCP or Azure.
Why is my AWS / GCP / Azure bill higher than this estimate?
Real bills add things a ballpark can’t: cross-AZ and inter-region data transfer, storage and snapshots, NAT gateways, logging and monitoring, support plans, and premium service tiers. Egress in particular surprises people. Treat this as a floor, not a ceiling.
How do I reduce Kubernetes costs?
In order of impact: right-size nodes to real CPU/memory usage, move stateless and batch work to spot/preemptible instances, buy committed-use/reserved capacity for steady-state load, and cut egress by keeping traffic in-region and caching at the edge. Together these often cut a bill by 30-60%.
Which is cheaper for Kubernetes: AWS, GCP or Azure?
Raw compute list rates are broadly comparable. GCP often edges ahead on automatic sustained-use discounts, while AWS and Azure tend to win on commitment-based savings plans. The bigger differentiators are usually egress pricing and how well your workload fits spot capacity, which is why a per-workload estimate matters more than the sticker rate. Switch providers in the calculator above to compare your own setup.
Is this calculator accurate?
It’s a good directional estimate, not a quote. The rates are representative on-demand list prices, refreshed quarterly. For an exact number tied to your workload and region, the fastest path is a short call.