Google's Gemini Omni 1.1 Flash, a new iteration of its multimodal AI model, provides developers with enhanced capabilities for video generation and editing. The model supports the creation of video clips up to 15 seconds in length, an increase from previous versions, and includes native synchronized audio for each output. This update also brings improvements to visual consistency and narrative flow within generated videos.The Gemini Omni 1.1 Flash model processes text, images, audio, and video inputs simultaneously, aiming for more cohesive and controllable outputs. It features conversational editing, allowing users to refine and modify videos through natural language interactions. This conversational approach enables iterative adjustments, such as swapping characters, altering lighting, or changing camera angles, while maintaining the integrity of other video elements.Developers can access Gemini Omni 1.1 Flash through Google AI Studio and the Gemini API. The pricing for video output has been adjusted, with some reports indicating a 30% reduction in cost for video output tokens, bringing the rate to approximately $0.071 per second for 720p video. Previously, Gemini Omni Flash video output was billed at roughly $0.10 per second for 720p video.The model supports various resolutions, including 360p for draft previews, and can upscale final projects to 4K resolution. Users can also specify first and last frames for interpolation, which facilitates smoother camera movements and transitions between scenes. The Gemini Omni family of models is designed to integrate Gemini's reasoning capabilities with generative media technology, aiming to bridge the gap between photorealism and narrative storytelling. This model is positioned to replace the Veo 3.1 model in the Gemini app, though Veo 3.1 models remain available for specific use cases in Google Flow.