MiniMax: MiniMax H3 Regeneration API

The new generation of open and universal multimodal video models is designed for multimodal contexts composed of text, images, video, and sound, enabling a unified understanding of creative intent and achieving more natural and coherent generation and expression.