z-image-turbo最近挻火，试了一下速度确实挻快，生成质量也不错，就做了个混合工作流。
工作流展示如下：
工作流所需资源
Z Image Turbo Mix Workflow, includes features: Direct/Image2Image/Controlnet.
Requirements:
Models:
Main:
– Z Image Turbo, Put in: ComfyUI\models\diffusion_models; vRAM 16G can use Q8_0, vRAM 8G can use Q5_K_M.
https://huggingface.co/Comfy-Org/z_image_turbo/resolve/main/split_files/diffusion_models/z_image_turbo_bf16.safetensors
– qwen_3_4b; (CLIP) Put in: ComfyUI\models\text_encoders
https://huggingface.co/Comfy-Org/z_image_turbo/resolve/main/split_files/text_encoders/qwen_3_4b.safetensors
Optional abliterated version
https://huggingface.co/mradermacher/Huihui-Qwen3-4B-abliterated-v2-GGUF/resolve/main/Huihui-Qwen3-4B-abliterated-v2.Q8_0.gguf
– ae.safetensors (VAE) Put in: ComfyUI\models\vae\Img
https://huggingface.co/Comfy-Org/z_image_turbo/resolve/main/split_files/vae/ae.safetensors
ControlNet:
– Controlnet-Union-2.0; Put in: ComfyUI\models\model_patches
https://huggingface.co/alibaba-pai/Z-Image-Turbo-Fun-Controlnet-Union-2.0/resolve/main/Z-Image-Turbo-Fun-Controlnet-Union-2.0.safetensors
ComfyUI Nodes:
– rgthree-comfy
– ComfyUI-KJNodes
– ComfyUI-GGUF
– comfyui-impact-pack
– comfyui_controlnet_aux
– comfyui-custom-scripts