Google Revolutionizes Video Creation with Gemini Omni and 3.5

At the Google I/O 2026 conference, Google unveiled its latest artificial intelligence models, Gemini Omni and the Gemini 3.5 series, which promise to transform the way users interact with multimedia content and digital tools.

Gemini Omni stands out for its ability to generate high-quality videos from various inputs such as images, audio, video, and text. These videos are grounded in Gemini's real-world knowledge, providing a solid foundation for creation. One of its key features is video editing through natural language conversations, allowing for precise modifications while maintaining the coherence of scenes and characters. For example, it is possible to transform a scene by changing the lighting or adding complex visual elements.

Additionally, Omni allows users to reimagine the action in a video, such as adding new characters or objects, or transforming a moment into something unexpected. Users can refine their videos through multiple iterations, changing the environment, angle, style, or specific details without losing the thread of the original scene.

On the other hand, Gemini 3.5 Flash focuses on the rapid execution of complex tasks. Powered by Antigravity, it can execute multi-step workflows for tasks like renaming and automatic categorization of unstructured assets. It also enables the creation of interactive web interfaces and sophisticated mathematical visualizations. Gemini 3.5 Flash is the default model for the Gemini application and the AI mode in global search.

Gemini 3.5 Flash can also develop multiple creative concepts in parallel and code faster through iterative loops. For instance, it can generate different UX approaches for a payment flow in just 60 seconds.

The capabilities of Gemini 3.5 Flash extend to personal AI agents, which are integrated into Google tools such as Gmail and Docs. These agents operate in the background to provide relevant information and take actions based on user needs. For example, an agent can keep a user informed about the latest sneaker collaborations. Information agents will be launched first for Google AI Pro and Ultra subscribers this summer.

Gemini Spark, a personal AI agent operating on Gemini 3.5, uses the Antigravity harness to help users navigate their digital lives. It is deeply integrated with the Workspace tools you use daily, such as Gmail, Docs, and Slides. Gemini Spark is now available to all Google AI Ultra subscribers in the United States.

Gemini 3.5 Flash is available through Google Antigravity, the Gemini API in Google AI Studio and Android Studio, thus strengthening Google's AI tools ecosystem.