Google Debuts Gemini Omni and 3.5 Flash to Power Next-Gen AI

Google has unveiled Gemini Omni and Gemini 3.5 Flash, marking a significant shift toward agentic AI that can perform complex tasks across video, audio, and code. Announced during the I/O 2026 keynote, these models power a new era of digital assistants capable of reasoning and executing actions in real-time. The company also introduced Antigravity 2.0, a desktop application designed to manage multiple AI agents simultaneously.

The new Gemini Omni model is a natively multimodal system that generates high-quality video from text, image, or audio inputs. It introduces conversational video editing, which allows users to change scenes, characters, or physical properties using natural language commands. To ensure transparency, Google is integrating SynthID digital watermarking into all generated content. This model is available for free on YouTube Shorts and the YouTube Create App, with full access provided to Google AI Ultra subscribers.

Advanced Capabilities of Gemini 3.5 Flash

For users requiring high-speed performance, Gemini 3.5 Flash offers a fourfold increase in speed compared to previous frontier models. This efficiency is central to the new Managed Agents feature in the Gemini API, which lets developers deploy AI that can reason and run code in secure, isolated environments. Google also launched Antigravity 2.0 to help users orchestrate these agents through a unified desktop interface, supporting parallel workflows across different tasks.

The company is also restructuring its premium offerings with the launch of the Google AI Ultra subscription. Priced at $100 per month, this tier provides the highest level of access to agentic tools and multimodal models. As an introductory incentive, current subscribers can receive $100 in Antigravity credits if they sign up before May 25. These updates represent the most substantial change to the search and assistant interface in over two decades, moving from simple queries to proactive AI agents.

Mobile developers and enthusiasts can now access an expanded Google AI Studio, which includes native Android support and a dedicated mobile app for rapid prototyping. This allows for the creation of AI-driven experiences directly on smartphones, utilizing the low-latency capabilities of the Flash model. With Gemini Omni understanding complex physics like fluid dynamics, the boundary between AI-generated and captured video continues to blur for everyday creators.

While we strive for accuracy, bytevyte can make mistakes. Users are advised to verify all information independently. We accept no liability for errors or omissions.

Sources

Building the agentic future: Developer highlights from I/O 2026

Introducing Gemini Omni

Photo by John Doe on Unsplash

✔Human Verified

Advanced Capabilities of Gemini 3.5 Flash

Sources

Related Articles