At Google I/O 2026 on May 19, Google announced Gemini Omni, a new family of multimodal generation models built from the ground up as a single system rather than a collection of separate tools stitched together . Google described Omni as "our first step towards a model that can create anything from anything — starting with video"
.
The first model in the family, Gemini Omni Flash, launched on May 20, 2026 . It accepts text, images, audio, and video as input and produces editable video as output, supporting conversational editing where each prompt carries forward the prior context
.
Omni Flash generates 1080p video clips up to 10 seconds long with native audio, according to early reports . Because the model incorporates an understanding of physics and real-world object behavior, it aims for visual consistency across shots—characters keep the same appearance, lighting stays matched, and camera angles can be adjusted through conversation alone
.
Every output includes a SynthID watermark . An avatar feature also allows users to insert their own face and voice into generated clips
.
When you're texting friends about where to eat, Magic Cue now suggests using Gemini to find restaurant options, displaying results in a floating window so you never have to leave the conversation . Gemini can also handle routine tasks like ordering groceries or booking a ride while keeping the user in control, a capability rolling out as a beta feature in the Gemini app
.
The broader June 2026 Android Drop also includes Google Photos "Wardrobe," which begins rolling out the second week of June 2026 in the US, India, and Brazil, and Gemini integration in Chrome for Android arriving in late June .
The June 2026 Pixel Drop started arriving on supported Pixel devices around June 2-3, with features continuing to deploy in phases over the following weeks, depending on the carrier and specific device . Screen Reactions is expected on Pixels this summer
, while Gemini Omni Flash is already live for subscribers and YouTube Shorts users
.