MiniMax H3 Day-0 Support in ComfyUI: Open Weights, Native Audio, and 2K Video

MiniMax H3, an open-weights omni-modal video model, launched today with native support in ComfyUI and can generate videos up to 2K resolution and 15 seconds long with real stereo audio from text, images, video, or audio inputs. The model excels at multimodal understanding, allowing users to combine multiple input types and control elements like opening/closing frames and motion transfer while keeping audio generation integrated rather than as a post-process. It can run locally on consumer hardware like an RTX 3060 and is optimized for ComfyUI workflows.