大模型
huggingface.co ·
Qwen-Image-2.1 非审查版 GGUF 量化模型发布，附 ComfyUI 使用说明
Hugging Face 用户 abenzerps 发布了 Qwen-Image-2.1 的 GGUF 量化版本，包含 Q4_0、Q4_K_M、Q5_K_M、Q6_K、Q8_0 五档文件（4.05GB–7.59GB），并附带 qwen3vl_8b 文本编码器（BF16/Int8）与 qwen_image_2.1_vae。仓库说明 Q8_0 在 ComfyUI 采样时会因 1D RMSNorm scale 量化出现张量形状不匹配错误（[136] vs [128]），建议使用 Q4_K_M、Q5_K_M 或 Q6_K，并称修复版 Q8_0 稍后上传。使用方式为配合 ComfyUI 与 leejet/ComfyUI-GGUF 节点，将扩散模型放入 diffusion_models、文本编码器放入 text_encoders、VAE 放入 vae 目录，并给出了节点配置与显存优化建议（扩散模型驻留 GPU、文本编码器放系统内存）。
AI推荐理由 这是第三方在 Hugging Face 上发布的 Qwen-Image-2.1 GGUF 量化仓库，提供了 Q4_0 至 Q8_0 多档量化文件、配套 text encoder 与 VAE，以及 ComfyUI 的目录放置和节点配置说明，对本地部署图像生成模型的开发者有直接参考价值。正文还明确记录了 Q8_0 在 ComfyUI 采样时因 1D RMSNorm scale 量化触发张量形状不匹配（[136] vs [128]）的问题，并建议改用 Q4_K_M/Q5_K_M/Q6_K，属于可复现的具体技术信息。局限在于：仓库自称使用上游基础权重，但未提供量化前后质量对比或基准数据；'Uncensored' 的表述缺乏说明；发布者身份与官方关系不明，不能视为官方发布。
62
正文
Notice regarding Q8_0:
Running qwen-image-2.1-Q8_0.gguf in ComfyUI currently triggers a tensor shape mismatch error ([136] vs [128]) during sampling due to 1D RMSNorm scale quantization.
Please use Q4_K_M, Q5_K_M, or Q6_K instead, which are verified to work completely fine out of the box. A patched Q8_0 file will be uploaded soon.
Qwen-Image-2.1 Uncensored GGUF
GGUF quantizations of Qwen/Qwen-Image-2.1 for local image generation using the original upstream base weights.
Benchmark
GGUF files
Quantization
File
Size
Q8_0
qwen-image-2.1-Q8_0.gguf
7.59 GB
Q6_K
qwen-image-2.1-Q6_K.gguf
5.88 GB
Q5_K_M
qwen-image-2.1-Q5_K_M.gguf
5.22 GB
Q4_K_M
qwen-image-2.1-Q4_K_M.gguf
4.60 GB
Q4_0
qwen-image-2.1-Q4_0.gguf
4.05 GB
Q4_K_M is recommended for the best balance of size and quality.
Text Encoders & VAE
Companion model files packaged for ComfyUI:
Type
File
Precision
Size
Text Encoder
text_encoders/qwen3vl_8b_bf16.safetensors
BF16
17.53 GB
Text Encoder
text_encoders/qwen3vl_8b_int8_convrot.safetensors
Int8
9.35 GB
VAE
vae/qwen_image_2.1_vae_bf16.safetensors
BF16
676 MB
Usage
Use the model with ComfyUI and ComfyUI-GGUF.
All required companion files (GGUF transformer, text encoder, and VAE) are hosted directly in this repository.
1. Download & File Placement
Download the files and place them in their respective ComfyUI directories:
ComfyUI/
└── models/
├── diffusion_models/
│ └── qwen-image-2.1-Q4_K_M.gguf # Choose one GGUF quantization (Q4_K_M recommended)
├── text_encoders/
│ └── qwen3vl_8b_bf16.safetensors # Or qwen3vl_8b_int8_convrot.safetensors (recommended for lower memory)
└── vae/
└── qwen_image_2.1_vae_bf16.safetensors
2. ComfyUI Setup
Install ComfyUI-GGUF: Use the maintained fork with native Qwen-Image 2.1 support by cloning leejet/ComfyUI-GGUF into your custom nodes:cd ComfyUI/custom_nodes
git clone https://github.com/leejet/ComfyUI-GGUF
(Note: If you have the older city96/ComfyUI-GGUF installed and encounter an Unknown model architecture! error, update to the leejet fork above or add ModelQwenImage to tools/convert.py).
Node Configuration:
Diffusion Model: Add the Unet Loader (GGUF) node and select your downloaded .gguf file.
Text Encoder: Add the standard CLIPLoader node, select qwen3vl_8b_bf16.safetensors (or int8), and set type to qwen_image.
VAE: Add the standard VAELoader node and select qwen_image_2.1_vae_bf16.safetensors.
Official Workflows:
You can use the official Comfy-Org workflow templates: Text-to-Image or Image Edit.
In the workflow, simply replace the default UNETLoader node with Unet Loader (GGUF).
Memory & Performance Notes
Optimal Setup (GPU + RAM): Keep the GGUF diffusion model in GPU VRAM (where speed is crucial during sampling) and let the text encoder run in / offload to System RAM (CPU). Because text encoding only runs once per prompt, this saves 9–17 GB of VRAM with virtually zero impact on generation speed.
Recommended Configuration:
Diffusion: qwen-image-2.1-Q4_K_M.gguf (~4.6 GB in VRAM)
Text Encoder: qwen3vl_8b_int8_convrot.safetensors (~9.35 GB in RAM)
Low VRAM Mode: If you experience VRAM out-of-memory errors, start ComfyUI with the --lowvram argument.
Uncensored
This GGUF release has no built-in safety checker or content filter. It generates adult, NSFW, and sensitive imagery directly without prompt refusals or blacked-out images. Output behavior depends solely on the input prompts and the environment in which the model is executed.
Source and build
Source model: Qwen/Qwen-Image-2.1
Text encoder & VAE source: Comfy-Org/Qwen-Image-2.1
Source revision: b3179ad355be050328e483a9dfdd9e60cd62adfa
Conversion: stable-diffusion.cpp commit 1330cebae8f2ba99249df846cc0c9444fcbd4308
License: Qwen Research License
Checksums: SHA256SUMS
Downloads last month -
GGUF
Model size
7B params
Architecture
Hardware compatibility
Log In to add your hardware
4-bit
Q4_0
4.05 GB Q4_K_M
4.6 GB
5-bit
Q5_K_M
5.22 GB
6-bit
Q6_K
5.88 GB
8-bit
Q8_0
7.59 GB
Model tree for abenzerps/Qwen-Image-2.1-Uncensored-GGUF
Base model
Qwen/Qwen-Image-2.1
Quantized
(18)
this model