Amazing Z-Image Workflow

A workflow developed while experimenting with Z-Image-Turbo, incorporating additional features beyond the base ComfyUI workflow. Three versions utilizing different checkpoints optimized for varying GPU VRAM capacities are included.

Features

• Eight configurable image styles for testing and experimentation.
• Versions in both .safetensors and .gguf formats to support a range of GPUs.
• Custom sigma values adjusted to my preference (subjectively better prompt adherence).
• Generated images are stored within the "ZImage" folder, organized by date.
• Includes a trick to enable CivitAI automatic prompt detection.

Workflow Overview

The repository contains three workflow files, each optimized for different GPU VRAM capacities:

1. "amazing_zimage-GGUF.json" : Recommended for GPUs with 12GB or less VRAM.
2. "amazing_zimage-GGUFSMALL.json" : For GPUs with less than 8GB VRAM.
3. "amazing_zimage-SAFETENSORS.json": Based directly on the ComfyUI example.

Required Checkpoints Files

Note

The workflows may require some custom nodes; see below for more details.

"amazing_zimage-GGUF.json"

Works smoothly with 12GB of VRAM or less, it may handle around 8GB as well.
Checkpoints used:

• z_image_turbo-Q5_K_M.gguf _{(5.52 GB)}
  Local Directory: ComfyUI/models/diffusion_models/
• Qwen3-4B.i1-Q5_K_M.gguf _{(2.89 GB)}
  Local Directory: ComfyUI/models/text_encoders/
• ae.safetensors _{(335 MB)}
  Local Directory: ComfyUI/models/vae/

"amazing_zimage-GGUFSMALL.json"

Optimized for GPUs with limited VRAM (less than 8GB), though prompt accuracy might be affected.
Checkpoints used:

• z_image_turbo-Q3_K_M.gguf _{(4.12 GB)}
  Local Directory: ComfyUI/models/diffusion_models/
• Qwen3-4B.i1-Q2_K.gguf _{(1.67 GB)}
  Local Directory: ComfyUI/models/text_encoders/
• ae.safetensors _{(335 MB)}
  Local Directory: ComfyUI/models/vae/

"amazing_zimage-SAFETENSORS.json"

Based directly on the official ComfyUI example, suitable for GPUs with around 12GB of VRAM or more.
Checkpoints used:

• z_image_turbo_bf16.safetensors _{(12.3 GB)}
  Local Directory: ComfyUI/models/diffusion_models/
• qwen_3_4b.safetensors _{(8.04 GB)}
  Local Directory: ComfyUI/models/text_encoders/
• ae.safetensors _{(335 MB)}
  Local Directory: ComfyUI/models/vae/

Required Custom Nodes

The workflows require the following custom nodes:
(which can be installed via ComfyUI-Manager or downloaded from their respective repositories)

• rgthree : https://github.com/rgthree/rgthree-comfy
• ComfyUI-GGUF: https://github.com/city96/ComfyUI-GGUF

License

This project is licensed under the Unlicense license.
See the "LICENSE" file for details.