Gemini Omni Flash claims top spot in Video Arena rankings

Google’s Gemini Omni Flash debuted at Google I/O on May 19, 2026, as the first entry in Google DeepMind’s new Gemini Omni family. The model combines advanced reasoning capabilities with generative media tools, including Google’s Veo video generation system, to create what amounts to a conversational video editor.

The model accepts text prompts, reference images, audio files, and existing video clips as creative starting points. It was designed specifically to maintain character consistency across multiple clips — one of the persistent challenges in AI video generation, where a character generated in one clip might look subtly different in the next.

ByteDance’s Seedance 2.0 currently leads independent evaluation platforms in Elo scores, with benchmarks placing Seedance 2.0 at 1,269 for text-to-video and 1,351 for image-to-video. As of early June 2026, Google has not submitted Omni Flash for official evaluations on these leaderboards, and independent reports indicate that Omni Flash’s raw visual fidelity currently lags behind competitors.

Google initially rolled out Omni Flash to Google AI Plus, Pro, and Ultra tier subscribers through the Gemini app and Google Flow. The company subsequently expanded free access to the model through YouTube Shorts and the YouTube Create app.

Worth noting: despite the broader trend of AI and crypto convergence in areas like decentralized compute and on-chain AI agents, there are no blockchain or token components attached to Gemini Omni Flash.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.