diff --git a/docs.json b/docs.json index da76b1df..af3ebd9f 100644 --- a/docs.json +++ b/docs.json @@ -156,6 +156,7 @@ "group": "Qwen", "pages": [ "tutorials/image/qwen/qwen-image", + "tutorials/image/qwen/qwen-image-2512", "tutorials/image/qwen/qwen-image-edit", "tutorials/image/qwen/qwen-image-edit-2511", "tutorials/image/qwen/qwen-image-layered" @@ -825,6 +826,7 @@ "group": "Qwen", "pages": [ "zh-CN/tutorials/image/qwen/qwen-image", + "zh-CN/tutorials/image/qwen/qwen-image-2512", "zh-CN/tutorials/image/qwen/qwen-image-edit", "zh-CN/tutorials/image/qwen/qwen-image-edit-2511", "zh-CN/tutorials/image/qwen/qwen-image-layered" diff --git a/tutorials/image/qwen/qwen-image-2512.mdx b/tutorials/image/qwen/qwen-image-2512.mdx new file mode 100644 index 00000000..9d6db30e --- /dev/null +++ b/tutorials/image/qwen/qwen-image-2512.mdx @@ -0,0 +1,86 @@ +--- +title: "Qwen-Image-2512 ComfyUI Native Workflow Example" +description: "Qwen-Image-2512 is the December update of Qwen-Image's text-to-image foundational model, featuring enhanced human realism, finer natural detail, and improved text rendering." +sidebarTitle: "Qwen-Image-2512" +--- + +import UpdateReminder from '/snippets/tutorials/update-reminder.mdx' + +**Qwen-Image-2512** is the December update of Qwen-Image's text-to-image foundational model. Compared to the base Qwen-Image model released in August, Qwen-Image-2512 features significant improvements in image quality and realism. + +**Key Enhancements in Qwen-Image-2512**: +- **Enhanced Human Realism**: Significantly reduces the "AI-generated" look and substantially enhances overall image realism, especially for human subjects +- **Finer Natural Detail**: Delivers notably more detailed rendering of landscapes, animal fur, and other natural elements +- **Improved Text Rendering**: Improves the accuracy and quality of textual elements, achieving better layout and more faithful multimodal (text + image) composition + +**Official Links**: +- [GitHub Repository](https://github.com/QwenLM/Qwen-Image) +- [Hugging Face](https://huggingface.co/Qwen/Qwen-Image-2512) +- [ModelScope](https://modelscope.cn/models/Qwen/Qwen-Image-2512) +- [Tech Report](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/Qwen_Image.pdf) +- [Blog](https://qwen.ai/blog?id=qwen-image-2512) + +## Supported Aspect Ratios + +| Aspect Ratio | Resolution | +| ------------ | ---------- | +| 1:1 | 1328x1328 | +| 16:9 | 1664x928 | +| 9:16 | 928x1664 | +| 4:3 | 1472x1104 | +| 3:4 | 1104x1472 | +| 3:2 | 1584x1056 | +| 2:3 | 1056x1584 | + +## Qwen-Image-2512 ComfyUI Native Workflow Example + + + +### 1. Workflow file + +After updating ComfyUI, you can find the workflow file from the templates, or drag the workflow below into ComfyUI to load it. + +The workflow includes two subgraphs: +- **Text to Image (Qwen-Image 2512)**: Standard 50-step generation +- **Text to Image (Qwen-Image 2512 4steps)**: Accelerated 4-step generation using Lightning LoRA + + +

Download JSON Workflow

+ + +### 2. Model download + +**Text Encoders** + +- [qwen_2.5_vl_7b_fp8_scaled.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/text_encoders/qwen_2.5_vl_7b_fp8_scaled.safetensors) + +**LoRA (Optional - for 4-step Lightning acceleration)** + +- [Qwen-Image-Lightning-4steps-V1.0.safetensors](https://huggingface.co/lightx2v/Qwen-Image-Lightning/resolve/main/Qwen-Image-Lightning-4steps-V1.0.safetensors) + +**Diffusion Models** + +- [qwen_image_2512_fp8_e4m3fn.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/diffusion_models/qwen_image_2512_fp8_e4m3fn.safetensors) (Recommended for most users) +- [qwen_image_2512_bf16.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/diffusion_models/qwen_image_2512_bf16.safetensors) (If you have enough VRAM and want better quality) + +**VAE** + +- [qwen_image_vae.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/vae/qwen_image_vae.safetensors) + +**Model Storage Location** + +``` +📂 ComfyUI/ +├── 📂 models/ +│ ├── 📂 text_encoders/ +│ │ └── qwen_2.5_vl_7b_fp8_scaled.safetensors +│ ├── 📂 loras/ +│ │ └── Qwen-Image-Lightning-4steps-V1.0.safetensors +│ ├── 📂 diffusion_models/ +│ │ ├── qwen_image_2512_bf16.safetensors +│ │ └── qwen_image_2512_fp8_e4m3fn.safetensors +│ └── 📂 vae/ +│ └── qwen_image_vae.safetensors +``` + + diff --git a/zh-CN/tutorials/image/qwen/qwen-image-2512.mdx b/zh-CN/tutorials/image/qwen/qwen-image-2512.mdx new file mode 100644 index 00000000..60f83b98 --- /dev/null +++ b/zh-CN/tutorials/image/qwen/qwen-image-2512.mdx @@ -0,0 +1,84 @@ +--- +title: "Qwen-Image-2512 ComfyUI 原生工作流示例" +description: "Qwen-Image-2512 是 Qwen-Image 文生图基础模型的 12 月更新版本，具有增强的人物真实感、更精细的自然细节和改进的文字渲染能力。" +sidebarTitle: "Qwen-Image-2512" +--- + +import UpdateReminder from '/snippets/tutorials/update-reminder.mdx' + +**Qwen-Image-2512** 是 Qwen-Image 文生图基础模型的 12 月更新版本。与 8 月发布的基础 Qwen-Image 模型相比，Qwen-Image-2512 在图像质量和真实感方面有显著提升。 + +**Qwen-Image-2512 的主要增强**： +- **增强的人物真实感**：显著减少"AI生成"的感觉，大幅提升整体图像真实性，尤其是人物主体 +- **更精细的自然细节**：风景、动物毛发和其他自然元素的渲染更加细腻 +- **改进的文字渲染**：提高文字元素的准确性和质量，实现更好的排版和更忠实的多模态（文字+图像）组合 + +**官方链接**： +- [GitHub 仓库](https://github.com/QwenLM/Qwen-Image) +- [Hugging Face](https://huggingface.co/Qwen/Qwen-Image-2512) +- [ModelScope](https://modelscope.cn/models/Qwen/Qwen-Image-2512) +- [技术报告](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/Qwen_Image.pdf) +- [博客](https://qwen.ai/blog?id=qwen-image-2512) + +## 支持的宽高比 + +| 宽高比 | 分辨率 | +| ------ | ------ | +| 1:1 | 1328x1328 | +| 16:9 | 1664x928 | +| 9:16 | 928x1664 | +| 4:3 | 1472x1104 | +| 3:4 | 1104x1472 | +| 3:2 | 1584x1056 | +| 2:3 | 1056x1584 | + +## Qwen-Image-2512 ComfyUI 原生工作流示例 + + + +### 1. 工作流文件 + +更新 ComfyUI 后，您可以在模板中找到工作流文件，或将下面的工作流拖入 ComfyUI 加载。 + +工作流包含两个子图： +- **Text to Image (Qwen-Image 2512)**：标准 50 步生成 +- **Text to Image (Qwen-Image 2512 4steps)**：使用 Lightning LoRA 的 4 步加速生成 + + +

下载 JSON 工作流

+ + +### 2. 模型下载 + +**文本编码器** + +- [qwen_2.5_vl_7b_fp8_scaled.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/text_encoders/qwen_2.5_vl_7b_fp8_scaled.safetensors) + +**LoRA（可选 - 用于 4 步 Lightning 加速）** + +- [Qwen-Image-Lightning-4steps-V1.0.safetensors](https://huggingface.co/lightx2v/Qwen-Image-Lightning/resolve/main/Qwen-Image-Lightning-4steps-V1.0.safetensors) + +**扩散模型** + +- [qwen_image_2512_fp8_e4m3fn.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/diffusion_models/qwen_image_2512_fp8_e4m3fn.safetensors)（推荐大多数用户使用） +- [qwen_image_2512_bf16.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/diffusion_models/qwen_image_2512_bf16.safetensors)（如果您有足够的显存并想要更好的质量） + +**VAE** + +- [qwen_image_vae.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/vae/qwen_image_vae.safetensors) + +**模型存储位置** + +``` +📂 ComfyUI/ +├── 📂 models/ +│ ├── 📂 text_encoders/ +│ │ └── qwen_2.5_vl_7b_fp8_scaled.safetensors +│ ├── 📂 loras/ +│ │ └── Qwen-Image-Lightning-4steps-V1.0.safetensors +│ ├── 📂 diffusion_models/ +│ │ ├── qwen_image_2512_bf16.safetensors +│ │ └── qwen_image_2512_fp8_e4m3fn.safetensors +│ └── 📂 vae/ +│ └── qwen_image_vae.safetensors +```