Gemini Omni multimodal generative AI platform announced at Google I/O

Google is rolling many of its existing generative AI models into what should act as a single all-encompassing service in Gemini Omni.

At Google I/O, the company showed off a new service which is meant to combine many of its generative AI models into one overall service. The new service is Gemini Omni, which is a multimodal generative AI platform said to be able to “create anything from any input, starting with video.”

Google showed off Gemini Omni during its Google I/O 2026 presentation this week. Omni promises to be a more comprehensive offering of Gemini Gen AI services than its previous products, which is said utilize several of Google’s AI models in one to offer better image and video generation than ever before. The new platform will begin in video with Gemini Omni Flash, and will be available in the Gemini app, the Flow by Google AI creative studio, and in the YouTube shorts suite.

The suggestion that Gemini Omni is “starting with video,” suggests that this could possibly move toward other forms of media, which may include music, photography, or even interactive media. In a following demonstration, Google devs even demonstrated how it could drum up a somewhat bizarre clone of Doom (though they started by saying it could run Doom).

Gemini Omni is going to be the next chapter of the company’s stretch into generative AI technology, so get used to hearing about it in variety of forms on the Google Gemini topic.

Shacknews staff does not use generative artificial intelligence (AI) in their content. Shacknews strictly prohibits the use of its content for AI training or to generate text, including text in the style or format used for this publication. Shacknews reserves all rights to this work.