Google Cloud spend caps are now generally available

Google Cloud has made spend caps on Cloud Budgets generally available, letting customers block further usage of a service once a spending threshold is hit, without deleting any resources. The feature adds three updates on top of the earlier preview.

First, a new spend caps API, built on the existing Cloud Billing Budgets API, lets teams programmatically create, update, or remove caps, including automatically applying one whenever a developer spins up a new sandbox project. Second, live cost reports now show estimated costs for AI services within minutes of usage, so teams can see exactly what triggered an enforcement instead of waiting hours for billing data to settle. Third, spend caps now reset automatically every month, so a blocked service resumes on its own at the start of the next billing period, though a manual-override option still exists for teams that require an audit before resuming.

Google recommends applying strict, low-dollar caps to sandbox and experimental AI projects first, and notes that default alerts go to billing admins and project owners at 50%, 80%, and 100% of a cap. Spend caps block pay-as-you-go and committed-use-covered requests, but never halt Provisioned Throughput, which keeps serving traffic on reserved capacity.

Why it matters

It gives cloud teams an automated way to cap runaway AI spending without the usual risk of resources getting deleted.

Source: Google Cloud