Google Gemini's New Usage Quota System Explained
Original: How Google’s New Gemini Rates Work and How to Track Your Usage
Why This Matters
Google's compute-based quota model signals a broader industry shift toward resource-aware AI billing structures.
Google has overhauled its Gemini AI usage metering system, shifting from a per-request count to a compute-power-based model. This affects all tiers: Free, AI Plus ($8/mo), AI Pro ($20/mo), and AI Ultra ($100–$200/mo), with 2x, 4x, and up to 20x more quota respectively.
Google has revamped how it measures Gemini AI usage across all subscription tiers. Instead of counting individual requests, the system now tracks the computing resources each prompt consumes. This means a complex video generation or code-writing task uses more quota than a simple weather query, regardless of request count.
The four available plans in the US are: Free (standard limits, unspecified), AI Plus at $8/month (2x standard), AI Pro at $20/month (4x standard), and AI Ultra at either $100 or $200/month (5x or 20x AI Pro limits). All tiers have access to the full model lineup—Flash-Lite, Flash, and Pro—with more advanced models consuming more quota per use.
Google's support documentation notes that 'access is subject to change or may be limited based on testing, experimentation or availability,' meaning daily limits may vary. The AI model selected and the complexity/length of prompts are the two primary factors determining usage consumption. Users can select their preferred model directly from the prompt interface and track remaining quota through Google's usage dashboard.