Google Tries to Tame the Beast of Unpredictable AI Spending

Google has introduced new pricing options for its Gemini Enterprise AI platform in an effort to help enterprises manage their costs and gain better visibility into their AI spending. The company has added a pay-as-you-go model, which allows customers to only pay for the compute and tokens they consume, as well as Flexible Savings Plans (FSPs) that offer discounts of 10% for one-year commitments and 20% for three-year commitments.

The pay-as-you-go model is designed to be more flexible than traditional software licenses, which can lock customers into fixed costs even if they don't use the full capacity. Google says this new pricing option will help enterprises lower their barriers to adoption and experiment with AI workloads without committing to a base subscription.

However, experts warn that the pay-as-you-go model also brings its own set of trade-offs, particularly around predictability. One user request can trigger an opaque chain of model calls, reasoning steps, and tool invocations, leading to unpredictable consumption costs.