New Delhi: Google made AI-related announcements at I/O 2026, including more Search AI integration, Gemini 3.5, and the latest Gemini Spark AI personal assistant. Google has also revealed Gemini Omni, an AI model that can create anything from any input, starting with video. This platform-first model, Gemini Omni Flash, is now rolling out to the Gemini applications, Google Flow, and YouTube Shorts.
Gemini Omni is the next step up from Nano Banana. It also enables users to combine images, audio, video, and text to generate high-quality videos. These videos can then be edited via natural language. While this sounds similar to Veo 3.1, Gemini Omni seems to be capable of taking this concept to the next level.
Google also outlined how Omni can better understand gravity, fluid dynamics, and kinetic energy to make the AI content it also generates look more realistic. All the content created with Omni is tagged with Google’s SynthID digital watermark to suggest that it features AI and was created with the Gemini Omni.
Gemini Omni also reached the point where it has gotten rid of the uncanny valley look that realistic AI videos almost always feature, but Google also seems to think it can offer its lofty claims about the platform. Gemini Omni demo video was definitely impressive, but it is still unclear what actual videos will look like when users start creating AI-generated and augmented videos with Omni Flash.
Gemini Omni Flash is available to all Google AI Plus, Pro, and Ultra-tier subscribers globally and is rolling out to YouTube Shorts and the YouTube Create App later this week.
