See what 5 builders are making with Gemini Omni

Google published a Keyword Blog post on Aug 07, 2026, by Lindsey Lanquist, showcasing five builders and companies using Gemini Omni. The piece frames Gemini Omni as Google’s new AI model for video creation and editing, positioned as conversational rather than timeline-heavy. The post ties the model to demos and messaging from this year’s I/O.

Gemini Omni is built for natural-language video work. Users can change camera angles, swap objects, and turn simple sketches into realistic animations through conversation-style prompts. Google says the model understands how the world works, so motion and edits stay natural and smooth instead of looking pasted or jumpy. Creation and editing sit in the same loop: describe the change, get a new cut.

For engineers and builders, that lowers the cost of iteration. A sketch or rough clip can become a passable animation without a full VFX stack, and object swaps or angle changes can be tried in chat instead of re-rendering a scene graph by hand. Product and content teams can prototype narrative, UI motion, and demo footage in the same session as the idea. The hard part shifts from tool fluency to prompt clarity and review of what the model actually produced.

Google is using the five-builder showcase to show real product and company use, not only lab demos. Conversational video sits next to other multimodal tools that already handle text, image, and code; Omni’s pitch is that video editing joins that same chat surface. Builders comparing stacks will weigh how far “change the angle / swap the object / animate the sketch” goes before they still need traditional editors, stock, or custom pipelines.

Watch how far those five builds go past demos: production pipelines, review workflows, and limits when the world-model assumption fails. Track whether conversational edits hold up for longer sequences and multi-shot continuity, and whether Google keeps the I/O-era capabilities in a product surface builders can ship against.