minimax/hailuo-3 is MiniMax's flagship omni-modal AI video generation model, built to create cinematic videos from text, images, video clips, and audio references within a single unified workflow. Unlike traditional video models, Hailuo 3 natively understands multiple input modalities together, enabling users to combine character identity, motion, voice, visual style, and creative direction into high-quality videos with synchronized stereo audio. It is designed for filmmakers, marketers, game developers, and content creators who need production-ready AI video generation with precise creative control. Supporting text-to-video, image-to-video, and reference-driven video generation, Hailuo 3 produces videos in up to 2K resolution, with realistic motion, consistent characters, and natural camera movement. Its unified multimodal architecture makes it ideal for storytelling, advertisements, virtual characters, product showcases, and professional creative workflows.