For AI Ultra subscribers, Google is also doubling the number of Omni videos they can generate.
Using Gemini 3.1 Pro with complex prompts or large files exhausted usage limits quickly. To address this, Google says it is now capping how much quota a single prompt can consume. This should help you get more out of your overall usage limit.
In case Gemini throws an error, it won’t count against your usage quota. Only successful completions will be counted.
Since Deep Research and other tasks use more tokens and compute than a simple text prompt. To ensure you can get more out of your five-hour and weekly quotas, Google will now provide a more detailed usage breakdown and notifications.
Gemini Flash-Lite usage remains free
Going forward, Josh says Gemini 3.1 Flash-Lite prompts will not count against your usage limit. So, you can at least continue working until your usage quota resets again.
Google is also making another quality-of-life improvement: once you select a Gemini model, the app will remember your choice and use it as the default for all future sessions. It’s only when you reach your usage limit that Gemini will switch to a lighter model.
Since rolling out compute-based usage limits for Gemini, Google has already tripled the usage limits twice following widespread criticism from users about hitting their quotas too quickly. The latest round of changes should help users get more out of ther usage limits
