Google Cloud spend caps are now generally available

Spend caps on Google Cloud Budgets are now generally available. Once a cap is reached, usage is blocked, and the resources in the project are not deleted. Google Cloud says the changes reflect feedback received since the preview.

A new spend caps API, reached through the existing Cloud Billing Budgets API, lets teams set, update, lift or delete caps in code, including applying a cap when a new project is created. Caps for AI services trigger within minutes using estimated live costs, which can be viewed in Cost Reports. Blocked services resume automatically at the start of the next month, while a manual override remains available.

Google Cloud recommends starting with non-production environments such as developer sandboxes and experimental agent playgrounds. Default alerts go to billing administrators and project owners at 50%, 80% and 100% of the cap, so those roles need to be assigned correctly. Caps halt pay-as-you-go and CUD-covered requests, but Provisioned Throughput requests keep running on reserved capacity.

Developers running agents can attach hard spending limits to sandbox projects and test environments through code instead of setting them by hand in the console. Agents that are deployed on pay-as-you-go or committed-use billing will be paused when a cap is reached, so spending limits should be sized with that in mind. Provisioned Throughput traffic continues to be served after a cap triggers.