Google DeepMind announced the ranking of its Gemini Omni Flash as the #2 model for video editing in the Video Arena, trailing only behind HappyHorse 1.0 by nearly 40 points. This ranking is part of the evaluation conducted on a limited set of models, showcasing the competition in AI video editing capabilities. Additionally, Google DeepMind unveiled two major model releases today: the fast and cost-effective Nano Banana 2 Lite for images, and the Gemini Omni Flash, which is accessible via the Gemini API and Google AI Studio to assist developers in generating and editing high-quality videos.

HappyHorse 1.0: HappyHorse 1.0 is an AI model participating in video editing performance evaluations. It ranked directly below Gemini Omni Flash in the Video Edit Arena. This placement underscores its role among competing systems in specialized video AI tasks.
Google DeepMind: Google DeepMind is an AI research organization focused on advancing multimodal models across language, vision, and video domains. It develops and releases models under the Gemini family for both research and practical applications. In the current news, the lab announced and shipped Gemini Omni Flash along with another model, highlighting its continued emphasis on video capabilities for developers.
Gemini Omni Flash: Gemini Omni Flash is a multimodal AI model specialized in high-quality video generation and editing. It integrates into the Gemini API ecosystem to support developer workflows. The model achieved a strong ranking in video editing evaluations shortly after its release announcement.
Nano Banana 2 Lite: Nano Banana 2 Lite is a streamlined Gemini image model designed for rapid and low-cost image generation. It forms part of recent updates expanding the Gemini lineup for accessible AI tools. The model was released concurrently with Gemini Omni Flash to broaden options for developers.

Model Release: Google DeepMind announced two major AI model updates focused on image and video capabilities for the Gemini platform.
Developer Access: The new models are now available through the Gemini API and Google AI Studio to support video and image workflows for developers.
Video AI Evaluation: Specialized benchmarks such as the Video Edit Arena assess AI models on video editing performance with a small number of participating systems.