Gemini Omni is revolutionizing video creation by making it as simple as having a conversation. Here are some exciting projects that individuals and companies are developing with this innovative tool.
Google’s latest AI model, Gemini Omni, simplifies the process of video editing and creation. With this tool, users can adjust camera angles, replace objects, or transform basic sketches into detailed animations. It offers an understanding of the physical world, ensuring videos appear natural and seamless. Gemini Omni empowers creators to bring their most imaginative ideas to life.
At this year’s I/O event, Google introduced Gemini Omni Flash, the first model in the Omni series. This tool allows for the creation and editing of high-quality videos using text, images, videos, or audio as references. Builders are already exploring its capabilities through various projects. Here are some noteworthy demonstrations.
One standout feature of Omni is its ability to make video editing remarkably effortless. Users can alter camera angles, modify settings, and apply cinematic zooms without disrupting the original scene’s continuity.
For example, builder Leon Lin (@LexnLin on X) demonstrated this by capturing a woman in a city from approximately 20 diverse perspectives. The video shows her from numerous angles, including close-up, distant, frontal, and profile views, from both high and low positions. The background changes dynamically, depicting her on different urban settings like sidewalks and crosswalks, amidst pedestrians, cars, trams, and varied architecture.
Omni allows creators to modify video content in ways that would be difficult to achieve through traditional filming. Users can change actions or exchange objects while maintaining a coherent scene.
In one demonstration, builder Carlos Santana (@DotCSV on X) altered an outdoor scene with voice commands. He changed the lighting from day to night, added cloudy skies and rainy sounds, turned leaves orange, and covered the ground with snow.
Using Omni in Google Flow, creators can turn sketches into realistic videos by utilizing doodles to dictate how elements move. They can choose whether these hand-drawn components are included in the final output.
Builder Pan (@sebatheepan on X) used doodles to whimsically transform everyday items. A lemon morphs into a submarine in the ocean, a cup of espresso becomes a hot air balloon, hot peppers transform into a sleeping dragon, a matchstick is reimagined as a rocket, and scissors become a menacing shark.
Omni also provides the ability to change the visual style of a video according to the creator’s vision. Users can input references or describe what they envision in natural language, and Omni will integrate these elements to produce a cohesive clip.
Jerrod Lew (@jerrod_lew on X) tested this feature by using Omni in Google Flow to render a video of a woman walking down the street in four distinct animation styles. The clip transitions smoothly from live-action to anime, to claymation, and more, without disrupting the woman’s movement.
The team at Hyperagent (@hyperagentapp on X) used Omni to illustrate three different concepts with video. They enhanced a video of an empty park with landscaping to create a design proposal. They represented data with an animated professor explaining business dashboards. Additionally, they gamified a to-do list with a video of a character completing tasks.
Gemini Omni combines knowledge with creativity, enabling the creation of scenes that adhere to real-world logic. It is available in the Gemini app, Google Flow, Google AI Studio, the Gemini API, or the Gemini Enterprise Agent Platform.
Sign up for newsletters to receive product updates, event information, special offers, and more. Check your inbox to confirm your subscription, or subscribe using a different email address. Your information will be used in compliance with Google’s privacy policy, and you may opt-out at any time.