Google added a pay-as-you-go option to the Gemini Enterprise app, dropping the base subscription fee so companies pay only for whatever compute and tokens their teams burn, at standard model API rates, with per-seat plans still available alongside it. Agent workloads forced the change, since one request can set off a chain of model calls and subagents unrelated to how many people hold licenses. Administrators can now cap monthly spend on a project, get alerts as it hits 50%, 80%, and 100% of that budget, and have agent API calls stop at the ceiling unless they let overages run. Committing to a monthly spend cuts token costs by 10% over one year or 20% over three, and Antigravity and Android Studio AI now draw on the same subscription and a daily quota pooled across the project.
