Z Image Turbo – A local Model 4 Megapixel Output

Z Image Turbo is bringing something back which was getting lost with some of the newer models like Flux.2 Dev, Wan 2.2, Qwen Image…the ability to run it locally! Of course some of these other models have quantized version available but that’s so far distilled that your results won’t be as good as the original but it will satisfy your need to run it locally.

However, Z Image Turbo model fits and runs very nicely on my RTX4080 with only 16GB VRAM. Its a distilled version of Z-Image which is a powerful and efficient image generation model with 6B parameters, developed by Alibaba’s Tongyi Lab. It should run in most consumer grade GPU under 16GB as well. The model was posted online on Hugging Face website.

Since then the model has been ported over to ComfyUI and the files available online for direct download.

Model Download and Storage Location

• Diffusion Model – z_image_turbo_bf16.safetensors
• Text Encoder – qwen_3_4b.safetensors
• VAE – ae.safetensors

📂 ComfyUI/
├── 📂 models/
│   ├── 📂 diffusion_models/
│   │   └── z_image_turbo_bf16.safetensors
│   ├── 📂 vae/
│   │   └── ae.safetensors
│   └── 📂 text_encoders/
│       ├── qwen_3_4b.safetensors

Once you have models organised you can download the workflow which I created using the Sub-graphs which simplifies the workflow a lot. I cover how to create and use sub-graphs in ComfyUI in my YouTube video.

The workflow will get you started with using Z Image Turbo.

If you'd like to support our site please consider buying us a Ko-fi, grab a product or subscribe. Need a faster GPU, get access to fastest GPUs for less than $1 per hour with RunPod.io