Google has unveiled a significant new family of generative AI models, dubbed Gemini Omni. The inaugural model, Omni Flash, currently focuses on generating AI videos, utilizing diverse inputs such as text, photographs, existing videos, and audio. Looking ahead, Google's ambition for Omni, as detailed in a blog post, is to evolve into a system capable of "creating anything from any input," thereby explaining its comprehensive "Omni" designation.
Google is positioning Omni Flash as the video equivalent of its highly successful Nano Banana image generation model, which has facilitated the creation of over 50 billion images since its launch last year. A notable feature of Omni Flash will be the ability to integrate a user's likeness into videos. While this prospect might not appeal to everyone, Nicole Brichtova, who heads the product team for Omni, informed The Verge that Google has observed significant user engagement with Nano Banana for similar likeness insertions into images.
Dumitru Erhan, Senior Research Director at Google DeepMind, confirmed to The Verge that Gemini Omni Flash will initially enable the generation of video and audio clips up to 10 seconds in length. Google is actively developing capabilities to extend this duration.
While Google already offers Veo, a text-to-video generation model, Omni Flash distinguishes itself by leveraging existing video content as a foundational input for generating new videos. Furthermore, Koray Kavukcuoglu, CTO of Google DeepMind and Chief AI Architect at Google, highlighted that Omni Flash possesses "a lot" more comprehensive world knowledge compared to Veo, a benefit derived from Gemini's extensive training data.
Gemini Omni Flash is slated for release starting Tuesday, and will be accessible through the Gemini app, Google Flow, and YouTube Shorts.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.