Google’s new Gemini Omni AI can turn almost anything into video

Google announces Gemini Omni Flash
It aims to make video creation easier by allowing users to adjust their projects naturally, rather than using editing software.
Emphasizing transparency and security through AI watermarking and privacy protection

Google’s next big AI move is aimed squarely at creativity. The company announced Gemini Omni at Google I/O 2026 as part of Gemini’s vast new features.

Omni is supposed to combine the reasoning power of Gemini with media creation tools that can generate and edit content in a variety of formats.

Our first release, Gemini Omni Flash, focused on video and was launched with very ambitious goals. Google wants to enable users to create content from almost any type of input, including text, images, audio, and existing videos.

conversational editing

One of Gemini Omni’s biggest ideas is to take the complexity out of editing. Google says users can make changes to videos through natural language, with consistency between changes.

The characters remain recognizable. Scenes maintain continuity. The motion remains consistent instead of resetting every time the prompt changes. The system is also designed to provide a deeper understanding of how objects behave in the physical world and incorporates improved handling of motion, gravity, and kinetic dynamics.

This is how the mirror above ripples like liquid when someone touches it, or how sculptures are created with bubbles. Google is trying to position Gemini Omni as something bigger than a video generator.

This puts Google directly into the rapidly escalating competition for AI media tools. But it’s a race to see who can make AI video tools as intuitive as any other tool that the average person will actually want to use. Google’s answer seems to take a conversational route.

Ultimately, Google said Gemini Omni goes beyond video. Future versions will support combining photos, prompts, music, and reference footage into a single project.

Trust AI creations

Powerful and creative AI creates trust challenges, something Google acknowledges. The company wants to highlight how videos created with Gemini Omni incorporate SynthID watermarking technology aimed at identifying AI-generated media. The company also says its verification tools will work across Gemini, Chrome, and Search as part of its broader transparency efforts.

Users will initially be able to create a video avatar based on themselves that includes their own voice. But while Google is working on safety considerations, more advanced features involving voice modification are still being evaluated.

This cautious approach reflects the increasingly difficult balancing act facing all major AI companies. Building a more capable system does not mean that trust in that system will be built at the same time.

Google logo on a black background next to the text