MiniMax H3 brings text-to-video, image-to-video, and reference-to-video into one browser workspace. Create 4–15-second native 2K videos from prompts plus image, video, and audio references, with optional synchronized stereo sound. Multimodal references help guide identity, composition, motion, camera direction, timing, and style. The same Studio also includes text-to-image, image-to-image, and AI image editing with supported outputs up to 4K.
