Autopilot prices intent; Standard prices capacity

For general-purpose Autopilot workloads, requested CPU, memory and ephemeral storage drive Pod-based billing. Specific hardware uses node-based billing plus an Autopilot premium; Standard bills the underlying nodes.

Question	Autopilot general purpose	Autopilot specific hardware	Standard
Billing unit	Running Pod requests	Provisioned node hardware	Compute Engine node resources
Idle node capacity	Abstracted from tenant	Visible in node cost	Paid by cluster owner
Main optimization	Right-size requests	Pack and release hardware	Bin-pack nodes and right-size requests
Control premium	Less tuning, more guardrails	Managed nodes with hardware choice	Most control and operational work

In every mode, inaccurate resource requests become a cost and scheduling problem.

GKE cost model is easiest to understand by separating the Kubernetes contract from the Google Cloud implementation. The cluster management fee applies across modes. General-purpose Autopilot uses Pod-based billing tied to requests, while hardware-specific Autopilot workloads and Standard node pools use node-oriented economics. The Kubernetes objects stay familiar, but GKE supplies controllers, infrastructure and safe defaults around them. This is why a team can move from an on-premises cluster without rewriting every workload, while still needing to redesign networking, identity and operational ownership for the cloud environment.

A useful inspection step is `kubectl top pods -A --containers`. Read the output as evidence, not as a ritual: first confirm the desired object exists, then look at status conditions, events and the Google Cloud resource it represents. In production, capture the expected result in a runbook or automated check so an operator can distinguish slow reconciliation from a configuration error.

Production gotcha: Never quote a fixed price from a slide; region, machine type, discounts and pricing policy change. Use the current calculator for decisions. The safe habit is to verify quotas, regional availability and feature support against current Google Cloud documentation before rollout. The final architecture slide explains editions and features.