MiniMax H3 ComfyUI - Run MiniMax turbo lora H3 33B omni-modal AI model locally with ComfyUI workflow. Text-to-video, image-to-video, reference-to-video generation with native stereo audio. ComfyUI custom nodes, workflow templates (T2V, I2V, R2V), H3-VisualVAE and H3-AudioVAE decoders. ComfyUI v0.31.0 support. 4-15s video at 768p. github
-
Updated
Aug 8, 2026 - Python