Z-Image-Turbo in Stable Forge UI Neo - One Click Windows Installer

Z-Image-Turbo is now supported in Stable Diffusion Forge UI Neo, and this guide shows you how to get it running on Windows with a one-click installer. Z-Image-Turbo is a 6B-parameter, few-step text-to-image model focused on fast, photorealistic generation that still fits on consumer GPUs with around 6 GB of VRAM using lightweight quantization methods.
This post walks through using the one-click Windows installer that sets up Stable Forge UI Neo with the full Z-Image-Turbo stack: Forge, PyTorch/CUDA, Sage, Flash Attention 2, Triton, the Z-Image-Turbo GGUF model, Qwen3 4B text encoder, and the VAE, all dropped into the right folders automatically.
GitHub Repository: https://github.com/Haoming02/sd-webui-forge-classic/tree/neo
Hugging Face: https://huggingface.co/Tongyi-MAI/Z-Image-Turbo
 
Included in the Package
The installer automatically sets up:
Sage
Flash Attention 2
Triton for Windows
PyTorch: 2.8.0+cu128
 
Preloaded Models
Z Image Turbo Q5_K_M GGUF model (z_image_turbo-Q5_K_M.gguf) — Hugging Face
Qwen3 4B Q5_K_M GGUF model (Qwen3-4B-Q5_K_M.gguf) — Hugging Face
Z Image Turbo VAE model (ae.safetensors) — Hugging Face
 
Speed
Generate 1024 x 1024 images in less than 1 minute (9 steps) on an RTX 4050 6 GB VRAM GPU using the Q5_K_M GGUF model.
 
System Requirements
Nvidia RTX 4090, 5090 series GPU or better
CUDA-compatible GPU with a minimum of 24 GB VRAM
Windows OS
At least 25 GB free storage
 
Usage Notes
Download and place the installer files into their own folder.
Double-click to install the project locally — no additional setup needed.
Ensure FFmpeg is installed: https://www.ffmpeg.org/download.html
Once the web interface is open, follow the configuration settings shown in the images.
Add the LoRA model under the LoRA tab.