---
title: FastVideo-FastH3
canonical_url: "https://www.modelscope.cn/models/Comfy-Org/FastVideo-FastH3"
md_url: "https://www.modelscope.cn/models/Comfy-Org/FastVideo-FastH3.md"
repository: Comfy-Org/FastVideo-FastH3
last_updated: 2026-09-16
license: other
base_model:
  - MiniMaxAI/MiniMax-H3
base_model_relation: finetune
parameters: 116.0B
tensor_type:
  - F32
  - F8_E4M3
  - I8
  - BF16
  - F16
  - U8
library_name:
  - safetensors
  - pytorch
frameworks:
  - pytorch
downloads: 588
stars: 6
tags:
  - comfyui
  - diffusion-single-file
---

# FastVideo-FastH3

> FastVideo-FastH3 - Comfy-Org 在 ModelScope 开源的模型。Repackaged model files for ComfyUI.

Comfy-Org/FastVideo-FastH3 是 ModelScope 魔搭社区上的 116.0B 参数机器学习模型，采用 other 许可，基于 MiniMaxAI/MiniMax-H3 构建。

- **Repository**: Comfy-Org/FastVideo-FastH3
- **License**: other
- **Parameters**: 116.0B
- **Base model**: MiniMaxAI/MiniMax-H3
- **Tags**: comfyui, diffusion-single-file
- **Downloads**: 588
- **Stars**: 6
- **Last updated**: 2026-09-16

Source: https://www.modelscope.cn/models/Comfy-Org/FastVideo-FastH3

---

# FastVideo FastH3

Repackaged model files for ComfyUI.

Original model repository: https://huggingface.co/FastVideo/FastVideo-FastH3-8-Step-V2

Place the files in the following folders:

```
📂 ComfyUI/
└── 📂 models/
    ├── 📂 diffusion_models/
    │   ├── fastvideo_fasth3_8step_v2_pruned_bf16.safetensors
    │   └── fastvideo_fasth3_8step_v2_pruned_int8_convrot.safetensors
    ├── 📂 text_encoders/
    │   ├── qwen3vl_32b_minimax_h3_bf16.safetensors
    │   ├── qwen3vl_32b_minimax_h3_int8_convrot.safetensors
    │   └── qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors
    ├── 📂 vae/
    │   ├── minimax_h3_audio_vae_fp32.safetensors
    │   ├── minimax_h3_video_vae_fp16.safetensors
    │   └── minimax_h3_video_vae_int8_convrot.safetensors
```