MANILA, Philippines – Google unveiled the Gemini Omni–a new model that includes an enhanced AI video-editing feature–at the Google I/O event on May 20, 2026.
Since Google’s introduction of their own native AI model, Gemini, in 2024, the multi-tech giant has continuously evolved its product and capabilities.
RELATED STORY: Google announces AirDrop compatibility for other Android brands
The new video editing model, Gemini Omni, brings in a conversational and transformational way to edit videos.
Video transformation through AI
Gemini Omni takes images, texts, audio and video from the user to generate videos according to prompts. It can also create its own imagery based on the instruction.
Prompts given to the model can either involve changing specific items on the video or every detail.
Users can replace actions, add people or objects, and manipulate angles. They can create new worlds or reimagine existing landmarks.
RELATED STORY: Googlebook unveiled at Google I/O 2026
Gemini’s knowledge translates the needed physics, gravity, and shadow to the videos.
“A simple circle turns into a black hole or an evening stroll comes to live,” said Demis Hassabis, CEO of DeepMind Technology. “Anything becomes a canvas for creating entire new realities.”
Digital avatar
With Gemini Omni, it is also possible for users to create their own digital avatar. The AI takes the user’s voice to insert it on an avatar prompted by the user.
Videos generated with Omni will include a watermark as part of Google’s policy for transparency. Users can use SynthID to verify if videos or images were made with AI.
The introduction of the Gemini Omni Flash is one of the many innovations in the event. It is the first within the Gemini Omni family
Gemini Omni can be used in the Gemini App and Google Flow subscribers. The feature is free for use in YouTube Shorts and YouTube Create App.