Gemini API Adds Spend Caps
Google added spend caps and cost-transparency tools to the Gemini API as generative-AI operational costs climb, and is clarifying free-tier data-use terms that may permit submitted data for product improvement. The changes reflect growing developer demand for predictable billing and clearer data governance in production AI deployments. ( )
Google published the change on March 16, 2026 in a blog post authored by Jason Stephen and Nabila Babar that outlines Project Spend Caps and Usage Tier updates. (blog.google)) Project Spend Caps are set per project inside Google AI Studio’s Spend tab and the product note warns the enforcement signal can lag by roughly 10 minutes, with developers held responsible for any overages during that window. (blog.google)) Google says the Usage Tiers now use lower spend qualifications and automatic upgrades so accounts with a stronger payment history move to higher quotas faster, and each tier carries a system-defined monthly billing-account cap separate from custom project caps. (blog.google)) Public documentation and independent observers list the preset billing-account caps as Tier 1 ≈ $250, Tier 2 ≈ $2,000, and Tier 3 from about $20,000 up to $100,000+, and Google’s docs indicate those tier caps begin being enforced on April 1, 2026. (zidooka.com)) Google’s Gemini API terms state that when using Paid Services the company does not use submitted prompts or responses to improve its products and processes paid-data under a Data Processing Addendum, while unpaid/free quota remains subject to product-improvement usage per the published terms. (ai.google.dev)) The timing follows a 2025 billing incident in which a configuration bug generated erroneous charges—reports described some developers being billed tens of thousands of dollars—which industry coverage and developer pressure helped push Google to add hard budget controls. (ppc.land))
