Skip to content

Cloud pricing term

What is Scale to zero?

Scale to zero means a service can reduce its running compute instances to none when there is no work.

Also called: scaling to zero, zero replicas, sleep when idle, serverless idle scaling.

The expensive mistake

Pricing an intermittent workload as 730 always-on hours, or assuming every plan with usage billing actually reaches zero. Minimum instances, unsupported shapes, open connections, background work, persistent storage, and request patterns can keep costs above zero.

Where it appears on your bill

Compute duration can fall toward zero, but storage, public IP, platform minimums, request charges, and minimum-instance idle rates may remain.

Worked example

For the recorded Cloud Run 1 vCPU, 2 GiB instance-based shape, 250 business hours cost $14.58 after the captured free allowances, while 730 always-on hours cost $52.596. Reaching zero outside business hours saves $52.596 - $14.58 = $38.016 in this scenario. Storage and other services are separate.

Example prices checked:

Primary sources