Google has introduced Gemini Omni Flash, a new multimodal AI model that can generate and edit videos from text, images, audio, and video input. The model is now rolling out to the Gemini app and Google Flow for subscribers of AI Plus, Pro, or Ultra plans. Free users can access it for remixing YouTube Shorts.
Multimodal model accepts images, audio, video, and text.
Gemini Omni Flash accepts images, audio, video, and text as input and initially outputs video. Google plans to add support for image and audio output in the future. The model can edit videos by changing the environment, angle, style, or specific details while maintaining consistency of characters and scenes. All videos created with Omni include a SynthID watermark.
Google claims Omni can "create anything from any input — starting with video" and has an "intuitive understanding of physics." The model is designed to handle complex edits through natural language prompts, allowing users to build on previous changes without losing the original scene's thread.
Gemini Omni Flash is available globally starting May 2026. It is included in the Gemini app and Google Flow for users with a Google AI Plus, Pro, or Ultra subscription. YouTube Shorts remixing is free for all users. Google has not confirmed whether a free tier for full video generation will be offered.
Discussion
0 comments
Log in to join the thread with a thoughtful take, question, or correction.