GOOGLE KILLS IMAGEN, LAUNCHES VIDEO AI THAT GENERATES AND EDITS FOOTAGE IN A SINGLE CONVERSATION Google shipped two new generative AI models on June 30 and simultaneously deprecated the entire Imagen product line, forcing an aggressive consolidation that moves everything visual under the Gemini API. The two new arrivals are Nano Banana 2 Lite for images and Gemini Omni Flash for video. Nano Banana 2 Lite is positioned as Google’s fastest and cheapest image model, generating 1K-resolution images in roughly four seconds at $0.034 per image. The model retains strong prompt adherence and character consistency according to Google’s internal benchmarks. Gemini Omni Flash is the more significant product. Available in public preview, it does not simply generate video from text prompts. It also conversationally edits video, allowing users to refine and revise footage through dialogue rather than starting over each time. That capability represents a meaningful gap over most competitors in the current market. The deprecation of all Imagen models is effective as early as June 30, with Google directing developers to migrate to the Nano Banana and Gemini Image families immediately. Existing Gemini preview image models are also being shut down. Development teams still running Imagen-dependent workflows were given almost no runway. The move is aggressive and signals that Google considers the consolidation non-negotiable. The Gemini Omni Flash model is available in Google AI Studio, the Gemini API, and the Gemini Enterprise Agent Platform beginning immediately. Keywords: Google Gemini Omni Flash, Nano Banana 2 Lite, Imagen deprecated, Google AI video generation