Google Cloud Added Deferred-Execution Agent Pricing at Half the Inference Cost and Pooled Quota Across Business Apps and IDEs
Google Cloud published new agent billing controls on 2026-08-26: a pay-as-you-go Gemini Enterprise consumption edition with no base subscription, project-wide pooled quotas that let business apps, Antigravity and custom agents draw from one allowance, and deferred execution pricing where flagged workloads run in off-peak windows at "up to half the inference cost" while bypassing quota limits. Flexible Savings Plans give 10% off a one-year and 20% off a three-year spend commitment with no minimum, drawing against an existing Google Cloud EA. Deferred execution is the first mainstream enterprise agent SKU priced on latency tolerance rather than seats or tokens, which is a spot-instance model applied to inference.
↳ Follow the thread