MiniMax Officially Releases H3 Video Generation Model with Audio Support

MiniMaxAI · hf · 2026-08-03

MiniMax has released its latest MiniMax-H3 model, now trending on Hugging Face. The model focuses on multimodal video generation, supporting text-to-video, image-to-video, and video-to-video workflows, alongside native audio generation (text/image-to-audio-video). Weights are available in diffusers and safetensors formats, with ComfyUI integration already supported.

Related event: MiniMax Releases Open-Source Omni-modal Model H3 with Native 2K Stereo Video(13 posts)→

Original post →

More from Models

Models channel →