← sources

A quick Minimax H3 news round-up - 6th September 2026

source: reddit (reddit:comfyui)published 2026-09-06 16:33ingested 2026-09-07 07:20

retained text

Another quick Minimax H3 news and goodies round-up, for those who may have missed some items. 

-> ComfyUI-H3-PowerLoraStack. "A stacked multi-LoRA loader for MiniMax H3, in the spirit of rgthree's Power Lora Loader, but built around the three things that actually break H3 LoRAs."

https://github.com/cicalooo/ComfyUI-H3-PowerLoraStack

-> 'Skynet Dark - MiniMax H3 Style LoRA'. A style LoRA to enhance a... "dark, rain-soaked cinematic look ... warm tungsten practicals, cold rain-window rim light, deep shadows, wet reflective streets". Trigger: *skynetstyle* placed first in the prompt. Works best with... "a pencil-sketch storyboard plus character references", which gives the LoRA some room to enhance the scene.

https://civitai.com/models/2918509/skynet-dark-minimax-h3-style-lora

-> The new ComfyUI-Viggle-Animate-H3 fine-tune model is designed for full character replacement. "ComfyUI nodes for Viggle-Animate, a 33.1 B full finetune [21Gb] of MiniMax-H3's ref2va transformer for character replacement in video. It takes a driving video and a reference still, and re-renders the performer(s) in the clip as the character in the still. Motion, camera, timing, background and lighting come from the video; identity comes from the image." 22Gb appears to be the smallest H3 Viggle combination (Viggle model + matching DMD2 LoRA).

https://github.com/Saganaki22/ComfyUI-Viggle-Animate-H3

-> And talking of character inputs with motion, note that a full Poser 12 Pro licence is just $20 via Graphixly. Poser 12 is mature 'pose 3D human figures and render' software from 2022, and its animation tools are fully bundled. Potentially useful for quickly producing clay-shaded 3D figure/expression animations, for use as H3 reference inputs (Poser's 3D figures can take any .BVH motion-capture file). Poser also has a real-time comic-book inking mode built in, giving you quick and pleasing line-art from 3D figures. Clay-shaded or line-art renders will prevent reference character-traits from leaking into your Minimax characters, and also give H3 more room to be a little creative. May also be useful for making references for closeups of hands and feet.

https://graphixly.com/products/poser-pro

https://www.youtube.com/watch?v=dkN8dN-FtGI (tutorial on the real-time comic-book inking)

-> Comfyui-h3-anime-batch-runner. "This is a Windows-based local controller that uses AI to create prompts for MiniMax H3 ref2v/ref2va based on a rough story or concept and one to nine reference images, and then continuously inputs them into ComfyUI clip by clip." In Japanese, and assumes you want "natural Japanese dialogue" in your outputs.

https://github.com/oodate-lab/comfyui-h3-anime-batch-runner

-> From HardGravy, Minimax_H3_FLF_HardGravy_6_Step_Turbo_Merge is new single-file 6-step "merge of fast LoRAs", along with one from JonXL for photorealism. Is it cold hard gravy or hot yummy sauce? I guess we test it and see.

https://huggingface.co/HardGravy2/Minimax_H3_FLF_HardGravy_6_Step_Turbo_Merge

-> And finally, Comfy Cloud has just... "added beta nodes that run curated [...] MiniMax H3, and Music 3 workflows on Comfy Cloud GPUs". Cloud also now has various H3 Max options.

https://docs.comfy.org/changelog

~ OLD POSTS ~

https://old.reddit.com/r/comfyui/comments/1w85caz/a_quick_minimax_h3_news_roundup_4th_september_2026/

https://old.reddit.com/r/comfyui/comments/1w74jy4/a_quick_minimax_h3_news_roundup_4th_september_2026/

https://old.reddit.com/r/comfyui/comments/1w6cozj/a_quick_minimax_h3_news_roundup_3rd_september_2026/

https://old.reddit.com/r/comfyui/comments/1w5i9iq/a_quick_minimax_h3_news_roundup_2nd_september_2026/ (See 2nd September post, for links to even older posts)

document fetched 2026-09-07 00:37 · canonical https://reddit.com/r/comfyui/comments/1w90vxd/a_quick_minimax_h3_news_roundup_6th_september_2026