Google Adjusts Gemini Usage Limits After User Complaints About Quota Depletion

At the 2026 Google I/O conference, the Gemini app switched to a compute-based usage limit system, replacing the previous fixed message quota. However, the new rules quickly sparked user backlash—prompts involving large files or complex tasks such as video analysis and Deep Research rapidly consumed the daily allocation. In response, Gemini head Josh Woodward announced several adjustments on May 28 aimed at improving Pro model availability and consumption transparency.

Under the new plan, Google will cap the quota consumed by a single prompt, preventing complex tasks from hogging too many resources at once. Additionally, prompts for the Gemini 3.1 Flash-Lite model are now free and will not count toward user quotas. Woodward also stated that more detailed usage breakdowns and notifications will be provided to help users understand exactly which actions consume quota. Furthermore, failed requests will not be counted toward the quota, a clarification that was previously lacking. The current gemini.google.com/usage dashboard only shows an overview, but it will be upgraded to a more granular dashboard in the future.

These changes come as users are still grappling with the new compute-based quota system. Although Google has not yet announced a specific launch date for pay-as-you-go AI credits, it is clear that flexible paid options for heavy users will be a key focus going forward. This swift response shows Google is working to balance compute costs with user experience, but finding a better solution between high-intensity tasks and broad accessibility remains a critical challenge for the Gemini ecosystem in the long run.