Google has addressed several issues with the usage limits of its Gemini AI model, which previously caused users to burn through their quotas too quickly. According to Josh Woodward, a Google VP, a bug was causing one or two Omni videos to consume the entire quota. This issue has been resolved, and Ultra members now receive double the number of Omni video generations. Additionally, complex requests to the 3.1 Pro model with large files were using too much quota, so the max consumption per prompt has been capped. The prompt still runs as usual, meaning users get more value from their plan. Other updates include that failed requests are no longer charged, Flash Lite requests are free, and complex features like Deep Research now provide more detailed consumption displays. If users select a specific model, that selection remains consistent across all sessions. At I/O, Google introduced a revamped Gemini app with new features, an agent mode, and a new pricing structure for Gemini AI subscriptions. *Source: [thedecoder](https://the-decoder.com/google-fixes-several-bugs-in-gemini-usage-limits-that-burned-through-quotas-too-fast/)*