Skip to content
Other

Studio models: MiniMax H3

int8-сборка HyperFlow для ComfyUI

1 min read

Репозиторий nodevorg с переупакованными файлами MiniMax H3 для запуска в ComfyUI и uncomfy. Основа — HyperFlow H3 Flash от Video Rebirth и оригинальная MiniMax H3. Это не официальный продукт MiniMax или Video Rebirth.

Ключевые особенности

  • Две int8-модели: fl2va на 16,3 ГБ и ref2va на 16,8 ГБ.
  • fl2va — текст в видео и изображение в видео по первому кадру или первому и последнему кадрам через MiniMaxH3ImageToVideo.
  • ref2va — видео по референсным изображениям и аудио через MiniMaxH3ReferenceToVideo.
  • Сэмплирование за 6 или 8 шагов без PDD; int8-веса скопированы из файлов Video Rebirth без изменений.
  • Нужны HyperFlowInt8Loader из ComfyUI-Studio-HyperFlow и ComfyUI-HyperFlow-Flash для расписания. Штатный загрузчик H3 их не откроет.
  • Текст-энкодер и VAE — штатные из Comfy-Org/MiniMax-H3.

Пример

ComfyUI/
└── models/
    └── diffusion_models/
        ├── minimax_h3_fl2va_hyperflow_flash_int8_convrot_comfy.safetensors
        └── minimax_h3_ref2va_hyperflow_flash_int8_convrot_comfy.safetensors
ПодробнееПодробнее
Размеры файлов: 16,3 ГБ и 16,8 ГБ.
В fl2va удалены 3 блока и 102 linear-слоя переведены в low-rank; в ref2va — 1 блок и 114 linear-слоёв.
Действует MiniMax H3 Community License, включая территориальные ограничения и правила допустимого использования.
Discuss Telegram 0 154

Will it run?

Runs in INT8

Fits without quantisation.

  • INT8recommendedweights 21 GB · ready files23 GB
  • 4-bitweights 12 GB13 GB

Tip: go 8-bit.INT8 is within about a percent of full precision in quality, takes half the memory and runs faster on RTX 40/50 and H100. Full precision only pays off for fine-tuning.

What to get for this model

Massed ComputeOpen Massed Compute

Your own GPU: INT8 — from RTX 4090

Massed Compute hourly prices, November 2026. The discounted price applies with the coupon at checkout.

Estimate: weights plus working memory. Long videos, high resolutions and long contexts need more.

Run it on Massed Compute

The script downloads the model to the rented machine. How to run it is in the model card.

  1. 1. Rent a machine with a fitting GPU

    Open Massed Compute
  2. 2. Paste the script into the machine's terminal

    Connect over SSH (the Massed Compute panel shows the command) or open a terminal on the machine's desktop.

    HF_TOKEN / CIVITAI_TOKENThe script starts with token lines: paste yours between the quotes. The Hugging Face token is for gated models, the Civitai one for Civitai models.

    # nodevorg/studio-models-h3
    mkdir -p ~/models
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv tool install "huggingface_hub[cli]"
    hf download nodevorg/studio-models-h3 --local-dir ~/models/studio-models-h3
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to studio-models-h3
  3. 3. Use it

    The model is in ~/models. How to run it is in the model card (link above).

Stop the machine in the panel when you are done: it bills by the minute while it runs.

How to speed it up

Picked for this model type. Choose your hardware above to see what fits you exactly.

  • FlashAttention 44.0.0b40from November 18, 2026new release

    fastest exact attention on B200

    Kernels for Blackwell (beta). Exact attention without quantisation — fine for everything, LLMs included.

    How to enable: pip install flash-attn-4; vLLM and SGLang pick it up themselves.

    Earlier generations (2)
    • FlashAttention 3 — The Hopper (H100/H200) version with FP8.
    • FlashAttention 2 — The standard for RTX 30/40/50 and A100: exact attention without the memory cost of long contexts.
  • torch.compile2.15.0from October 28, 2026new release

    10–30% for free

    Compiles the model for your GPU. Slower first run, faster afterwards.

    How to enable: model = torch.compile(model); in ComfyUI the TorchCompileModel node.

MiniMax H3 prompts

All prompts

Ready prompts with the resulting video: copy and try.

Model family

  1. MiniMax-H3family root
  2. hyperflow-h3-flash
  3. studio-models-h3

New posts about versions of MiniMax-H3 (quants, finetunes, LoRAs) will show up under the bell.

Quantised versions 12All on Hugging Face
Fine-tunes 12All on Hugging Face
LoRAs and adapters 12All on Hugging Face
Merges 3All on Hugging Face

About the model

nodevorg/studio-models-h3
Author
nodevorg
Category
Other
Released
Posted

On NeyroLearn

Hugging Face

1540

Ask AI about this model

Answered by Gemini 4 · the chat is kept only in this tab

Comments