MiniMax H3 SH Reference to Video
Open-weights video generation with native audio: text, first/last frames, and multimodal references. 3-15 seconds at 480p, 540p, 768p or 1080p.
Open-weights video generation with native audio: text, first/last frames, and multimodal references. 3-15 seconds at 480p, 540p, 768p or 1080p.