Veda-Sparse/Minimax-H3-T2VA-Veda-8NFE-600Step-Preview

Veda-MiniMax-H3 (Preview)

Paper · Project Page· Comfy Node; · Code · Turbo LoRA · Deployment Guide

🎉Veda now comes withComfyUI support. 🎉

Available as a custom node on the Comfy Registry; see ComfyUI below.

📌 Modality- and step-agnostic 📌

This checkpoint applies to all generation modalities supported by MiniMax-H3, including T2VA, FL2VA, and R2VA, as well as to arbitrary numbers of diffusion steps. The T2VA and 8NFE tags in the repository name only denote the training configuration and do not restrict where the predictor can be used. A dedicated R2VA fine-tuned checkpoint with further improved quality will be released in the near future.

Introduction

Veda is a learned sparse-attention method for video diffusion models. Attention dominates the inference cost of video diffusion, yet only a small fraction of it contributes meaningfully to the output. Veda trains a lightweight predictor, distilled from the full model, to identify the most important 10% of attention and skip the rest.

This repository provides the Veda predictor for MiniMax-H3 text-to-audio-video generation.

Highlights

• Faster inference. Up to 6.8× attention speedup and 3.1× end-to-end speedup, with larger gains on longer videos.
• Preserved quality. Visual and audio quality on par with full attention while skipping 90% of it.
• Plug-and-play. Veda is not a LoRA: it leaves the backbone weights and style untouched and only changes how attention is computed. It is compatible with any LoRA and with fine-tuned MiniMax-H3 variants.

This checkpoint is trained and evaluated with MiniMax-H3-Turbo-Lora for 8-step generation.

Samples

Comparison between full attention (left) and Veda (right) under the same prompt, seed and Turbo LoRA. All videos are 16:9, 14.4 s, generated on a single RTX PRO 6000 Blackwell GPU. Generation time is shown in each title bar.

Raw files are available under media/.

Performance

Speedup per denoising step relative to full attention (MiniMax-H3 + Turbo LoRA, 10% attention kept).

GPU	Clip	Attention speedup	End-to-end speedup
RTX PRO 6000 Blackwell	16:9 · 14.4 s	6.79×	3.12×
RTX 4090	16:9 · 5.17 s	4.75×	1.77×
RTX 4090	16:9 · 10.1 s	6.22×	2.42×
RTX 4090	16:9 · 14.4 s	6.31×	2.82×

Usage

ComfyUI

Install the node via the Comfy CLI:

comfy node install veda-sparse-attention

Then download the predictor checkpoint into ComfyUI/models/veda/ through hf or direct download:

hf download Veda-Sparse/Minimax-H3-T2VA-Veda-8NFE-600Step-Preview \
  minimax_h3_t2va_veda_8nfe_600step_preview_fp8.safetensors \
  --local-dir ComfyUI/models/veda

Miowtion

Requires the Miowtion codebase. See AGENTS.md for hardware requirements and per-architecture notes.

git clone https://github.com/veda-sparse/Miowtion.git && cd Miowtion
pip install -e '.[gpu,encode]'

hf download MiniMaxAI/MiniMax-H3 --local-dir weights/MiniMax-H3
hf download Veda-Sparse/Minimax-H3-T2VA-Veda-8NFE-600Step-Preview \
  --local-dir weights/veda/h3-t2va-8nfe-600

Encode the prompt, then generate:

echo '{"id": "demo", "task": "t2va", "prompt": "<structured T2VA prompt>"}' > prompts.jsonl
python scripts/encode_samples.py --root weights/MiniMax-H3 \
  --manifest prompts.jsonl --out artifacts/samples/demo

python scripts/generate.py \
  --root weights/MiniMax-H3 --variant FL2VA \
  --schedule turbo --num-steps 8 \
  --adapter weights/turbo_lora/<8-step-lora>.safetensors \
  --sample-cache artifacts/samples/demo --sample-id demo \
  --geometry 16:9@37 --attention veda \
  --predictor weights/veda/h3-t2va-8nfe-600/minimax_h3_t2va_veda_8nfe_600step_preview_fp8.safetensors \
  --out-dir artifacts/generate/demo

Use --attention dense veda to render both modes with a side-by-side video. On 24 GB GPUs, add --offload-blocks 50 --mlp-chunk-rows 4096.

Option	Description
--geometry	<aspect>@<latent_t>, one of the 12 packed plans
--keep-ratio	Override the default 0.1 (trained at 0.1)
--dense-steps	Denoising steps to keep dense
--offload-blocks	Number of transformer blocks streamed from host memory
--mlp-chunk-rows	MLP chunk size; reduce if out of memory

Loading the predictor directly:

from miowtion.veda import bundle

loaded = bundle.load(
    'minimax_h3_t2va_veda_8nfe_600step_preview_fp8.safetensors', device='cuda')
loaded.predictor          # TileScorePredictor
loaded.plans.select(geo)  # tile plan for a geometry
loaded.keep_ratio         # 0.1

License

This model inherits the MiniMax H3 Community License from its base model.

Citation

@inproceedings{han2026veda,
  title={Veda: Scalable Video Diffusion via Distilled Sparse Attention},
  author={Han, Shihao and Yang, Hao and Hu, Xinting and Mei, Xiaofeng
          and Jiang, Yi and Qi, Xiaojuan},
  booktitle={International Conference on Machine Learning (ICML)},
  year={2026}
}