Gemini Omni 1.1 Flash 发布：带来全新视频生成控制功能

Google 发布了 Gemini Omni 1.1 Flash 更新，为开发者提供生成式视频的新控制功能，该模型已通过 Gemini API 在 Google AI Studio 上线，面向专业创作与部署场景 1。在上下文理解方面，新模型可分析最多 10 秒的先前上下文，较此前仅参考最后 1 秒的模型有显著提升 1。

在视频生成控制上，Gemini Omni 1.1 Flash 支持场景延长功能，可按 10 秒增量进行，累计总长度最高 40 秒，并支持首尾帧指定 1。该模型支持在多模态输入中引用最多 3 秒的视频作为参考，并支持输出 1080p 或 4K 分辨率的高清视频 1。其 360p 分辨率预览生成速度比 720p 快最多 60%，成本为 720p 的三分之一 1。

Gemini Omni 1.1 Flash 已面向全球 Google AI Plus、Pro 和 Ultra 订阅用户在 Google Flow 中提供 1。场景延长功能已面向全球 Google AI Plus、Pro 和 Ultra 订阅用户在 Gemini 应用中提供 1。

Google has released an update for Gemini Omni 1.1 Flash, introducing new generative video control features for developers 1. The update allows for scene extension, the specification of first and last frames, 360p quick preview, video reference inputs, and high-definition video output in 1080p or 4K resolution 1. These tools are currently available via the Gemini API in Google AI Studio, catering to professional creation and deployment use cases 1.

The model features a significant upgrade in context analysis, capable of processing up to 10 seconds of prior context compared to the one-second limit of previous models 1. It also introduces a scene extension function that operates in 10-second increments, allowing for a cumulative total length of up to 40 seconds 1. To optimize workflow, a 360p resolution quick preview generates up to 60% faster than 720p at one-third of the cost 1. Furthermore, the system supports referencing up to three seconds of video in multimodal inputs 1. For consumers, Gemini Omni 1.1 Flash is accessible globally to Google AI Plus, Pro, and Ultra subscribers in Google Flow, with the scene extension feature specifically available to these subscribers in the Gemini app 1.