Google has introduced a new usage model for the Gemini app based on compute consumption, responding to user feedback regarding rapid depletion of limits. This revised system will refresh every five hours until the weekly limit is reached, considering prompt complexity and tool usage. Gemini's lead, Josh Woodward, mentioned that the company is implementing caps on the quota used by individual prompts to enhance the experience for users of the Gemini 3.1 Pro version.
As part of the updates, users will have the option to purchase pay-as-you-go AI credits in the future. Google emphasized that errors do not impact usage limits; unsuccessful requests will not be charged against a user's quota. Additionally, the company plans to enhance the current usage dashboard, which only provides a basic overview, by offering detailed breakdowns and notifications to help users manage their consumption effectively.
In a notable change, prompts categorized as 3.1 Flash-Lite will now be free and excluded from quota calculations. Users of Google AI Ultra will see an increase in Omni generations, addressing previous issues where a limited number of Omni videos were causing quota drains. This series of adjustments aims to improve user experience and maximize functionality within the Gemini platform.