Google didn't announce Gemini Omni. Users found it on their own. Reddit users posted screenshots of a revised Gemini interface exposing a new model card that read: "Create with Gemini Omni: meet our new video model, remix your videos, edit directly in chat, try templates, and more." That's a public-facing UI string, not a dev flag buried in an APK.
A Gemini Omni model has been spotted inside Google's own Gemini video generation interface, fueling speculation that Google is about to launch a single unified model capable of handling text, images, and video in one system. The leak comes just days before Google I/O 2026, scheduled for May 19–20, where the company is expected to make major AI announcements.
As someone who covers this beat closely, this one is worth paying attention to. Not because of the leak itself, but because of what the outputs are already showing.
What Is Gemini Omni
While we've known about Google's Veo model for a while, according to a report from 9to5Google, a brand-new iteration called Gemini Omni is already starting to show up for some users in the wild. At least one Gemini user was prompted to "Create with Gemini Omni," with Google describing the new video generation model as a way to remix your videos, edit directly in chat, and try out pre-made templates.
Toucan is Google's internal codename for the current Veo-3.1-powered video generation pathway inside Gemini. The Omni UI string appeared next to Toucan references, suggesting it may be a replacement or successor.
The name itself carries architectural weight. The most ambitious read is a single Gemini model that handles image generation, video generation, and possibly audio in the same system, the way GPT-4o is positioned for text-image-audio. If true, Gemini would be the first top-tier omni-model with video output.
How the Leak Surfaced
On May 2, 2026, an X user named @Thomas16937378 discovered a UI string inside Google's Gemini video generation tab that read: "Start with an idea or try a template. Powered by Omni." TestingCatalog, a reliable tracker of Google AI leaks, quickly picked up the finding and published a report that spread across the AI community within hours.
A freshly created profile in Gemini's video tab surfaced the "Powered by Omni" line, suggesting the feature is in late-stage testing. This is not a developer build or an APK teardown — it appeared in the live interface.
Two details make it more than noise: the string is visible to users, not just buried in source code or feature flags. UI copy that mentions a brand name typically reaches that state only when the team is preparing for a public release.
Early Output Quality
The early demos are genuinely impressive for a pre-announcement build. Early feedback suggests Gemini Omni may already outperform Veo in several areas. One user praised the model's prompt adherence, smoother camera angle transitions, stronger scene coherence, and significantly improved voice generation quality.
