Rohan Paul@rohanpaul_ai · XAI 摘要
Google针对Gemini应用配额问题推出多项修复。主要修复了一个导致1-2个Omni视频即耗尽部分用户配额的bug，Ultra用户的Omni生成次数提升至2倍。Pro提示词现在设有每提示配额上限，失败的请求不再计入配额。Flash-Lite模型免费使用且不占用配额。此外，系统将记住用户选择的模型，除非用户主动更改或达到配额上限。更详细的用量分析和提醒功能即将推出。
Google fixed some issues around Gemini's usage/quota limits.
- The biggest issue was that 1 or 2 Omni videos could drain limits for some users because of a bug； Ultra users now get 2X Omni generations - Pro prompts now have a per-prompt quota cap - Failed requests won't count against quota - Flash-Lite is free and won't use quota - Granular usage breakdowns/alerts are coming - Selected model will stay remembered unless you change it or hit a cap
We've heard your feedback about hitting limits too quickly on @GeminiApp. We're rolling out several fixes to make your quota stretch further and feel more predi...