MiniMax H3 Video Model Natively Integrated into ComfyUI
bdsqlsz · x · 2026-08-03
ComfyUI announced day-zero native support for the MiniMax H3 video generation model weights.
The model consolidates multiple video generation tasks into one, with key capabilities including:
- Text-to-video: Generate from prompts only.
- Image-to-video: Supports first-frame, last-frame, or both for control.
- Reference-to-video: Carry a subject, motion, or voice through the clip.
- In-place editing: Modify existing shots without regenerating.
Its core strength lies in robust multimodal context understanding, processing images, audio, and video together while resolving their relationships based on the prompt, collapsing five tasks into a single model.
Related event: MiniMax open-sources 33B multimodal video model H3(35 posts)→
More from Multimodal
- Tessellation-free geometric queries on displaced surfaces cut caustics variance 10x — ssh4net · 2026-10-16
- Complex multi-link CAD mechanism finally animates, dev celebrates the breakthrough — zetalyrae · 2026-10-16
- OpenMOSS releases MOSS-Image-SFT: Apache 2.0 text-to-image model on Hugging Face — bdsqlsz · 2026-10-16
- Hyperflow open-sourced: 49 sampling steps down to 8, ~3x faster video generation — mhdfaran · 2026-10-16
- Cinematic 3D fantasy character prompt template with salvaged-robot armor — azed_ai · 2026-10-16
- Midjourney vintage woodcut style: --sref 2446102396 beaver with a cleaver — michaelrabone · 2026-10-16