At this year’s I/O, Google introduced Gemini Omni Flash, its first model in the new Omni family, which allows users to generate and edit videos from text, image, video, or audio references. The model combines an understanding of physics with Gemini’s real-world knowledge for more realistic outputs. Developers recently gained access to Omni, and builders have used it for various projects. Builder Leon Lin captured a woman standing in a city from about 20 different perspectives, changing camera angles and backgrounds without losing the original scene. Builder Carlos Santana used his voice to alter an outdoor scene, switching lighting from day to night, adding cloudy skies and rainy sounds, turning leaves orange, and covering the ground in snow. Builder Pan used doodles in Google Flow to animate everyday objects, transforming a lemon into a submarine, an espresso cup into a hot air balloon, hot peppers into a sleeping dragon that breathes fire, a match into a rocket ship, and scissors into a shark. Omni also allows users to change styles and apply visual effects to match their vision.