See what 5 builders are making with Gemini Omni

Google's newest AI model, Gemini Omni, simplifies video editing and production, likening the process to engaging in a conversation. This innovative tool can modify camera angles, replace objects, and transform simple sketches into lifelike animations. It possesses a deep understanding of the physical world, enabling it to produce natural and seamless videos.

At the recent I/O event, Google unveiled Gemini Omni Flash, the initial model in the new Omni series. Omni simplifies both creating and editing videos, allowing users to produce high-quality videos from text, image, video, or audio references. It can also edit existing videos. Leveraging both intuitive physics understanding and Gemini's real-world knowledge, Omni generates more realistic outputs.

Early adopters of Omni have already utilized it for various projects. Here are some notable examples.

One builder, Leon Lin (@LexnLin on X), capitalized on Omni's ability to easily edit videos. He captured a woman in the city from 20 different perspectives, showcasing her from close-ups to far away, head-on and profile views, above and below. The background changes accordingly, with her seen on sidewalks, crosswalks, city streets, with pedestrians, cars, trams, and diverse building architecture. Omni allows altering a video's content and generating scenes unattainable through traditional filming.

Another builder, Carlos Santana (@DotCSV on X), modified an outdoor setting using only voice commands, adjusting lighting from day to night, introducing cloudy skies and rainfall, transforming leaves to orange, and covering the ground with snow.

A sketch-to-video feature in Google Flow enables turning doodles into realistic videos. Builder Pan (@sebatheepan on X) utilized this feature to whimsically transform common objects. A lemon morphs into a submarine, a cup of coffee into a hot air balloon, and a pair of hot peppers transform into a sleeping dragon breathing fire upon being startled. Scissors metamorphose into a menacing shark.

Omni's versatility extends to changing video aesthetics. Builder Jerrod Lew (@jerrod_lew on X) tested this by rendering a woman walking down the street in four distinct animation styles. The live-action video seamlessly transitions from anime to claymation, maintaining the woman's natural forward motion. The Hyperagent team (@hyperagentapp on X) employed Omni to visualize three concepts as video, incorporating landscaping into an empty park, personifying data with an animated professor explaining business dashboards, and gamifying a to-do list with a character clearing tasks.

Omni combines knowledge with creativity, allowing the creation of scenes adhering to real-world logic. Experience Omni in various platforms, including the Gemini app, Google Flow, Google AI Studio, the Gemini API, or the Gemini Enterprise Agent Platform.