At last week's Google I/O 2026 event, Gemini officially switched to a compute-resource-based usage quota system. But the change quickly sparked heavy backlash—many users reported that their quotas drained fast when processing complex videos or coding prompts, with even a single Deep Research task consuming a large chunk of their allowance. In response to these complaints, Google has already started fine-tuning the new system.
Josh Woodward, head of Gemini, confirmed today that Google is placing a cap on per-prompt quota consumption to give users more sustainable availability on the Pro model. Specifically, for high-demand tasks involving large files or heavy computation—such as Deep Research—the system will apply finer-grained quota controls. Google also pledged to introduce more detailed usage breakdowns and notification features, helping users clearly see where each quota unit goes. Currently, the gemini.google.com/usage dashboard offers only a rough overview, but future updates will significantly enhance visualization and alert capabilities.
