It moved to Gemini last week at I/O 2026 billing-based usage limits. In the answer For “feedback about hitting limits too quickly,” Google announced some changes today.
The new “computer usage” approach to usage (5 hourly updates until the weekly limit is met) is to take into account the complexity of the prompts, what tools are used and the length of the conversation. Last week, Google noted that “a simple text query uses less computation than a complex video or encoding query.” In the future, Google will allow Gemini app users to purchase additional AI credits.
When using Gemini 3.1 Pro, Gemini leads Josh Woodward shared today Google “limits the amount of quota that a single request can use so you can get more out of the Pro model.” This is the answer to complex instructions with large files that quickly exhaust the limits.
Google clarified that errors do not count against the limits: “If a request fails, you will not be charged. System errors are ours, not yours. Your quota is only used for successful completions.”
Heavy tasks like Deep Search “require more computing,” so Google will provide “more detailed usage breakdowns and notifications to help you maximize your limits.” as it is gemini.google.com/usage the dashboard provides only a high-level overview.
Meanwhile, 3.1 Flash-Lite hints are now “free and won’t count against your quota.” Google also notes how:
When you choose a particular model, we remember that choice for all future sessions. It will only change if you manually adjust it or hit the flap which causes it to automatically revert to a lighter model.
Finally, Google fixed a bug where “just one or two Omni videos” would empty quotas for “certain people”. Google AI Ultra users have now doubled the number of Omni generations.
We’ve fixed this and will continue to look for opportunities to increase the amount of Omni you get.
FTC: We use automatic affiliate links that generate income. More.
