The commit reorganizes the "/workflows" directory by dividing it into `/workflows/release` and `/workflows/experimental` folders. This structure helps in categorizing workflows, with the "release" folder containing tested and finalized workflows to be included in published versions, while "experimental" houses experimental workflows that are still under testing and development.
Z-Image Power Nodes: Example Workflows
Folders
/BF16: Contains workflows pre-configured for .safetensors BF16 checkpoints./FP8: Contains workflows pre-configured for .safetensors FP8 checkpoints./GGUF: Contains workflows pre-configured for GGUF format checkpoints.
Workflows
forging_custom_styles: Example of how to configure your own custom styles.injecting_style_into_string: Example of how to retrieve the complete string combining style and prompt.selecting_my_top_10_styles: Example of how create a list of your 10 most-used styles for quick selection.z-image-turbo__inpainting.json: Main reference workflow for inpainting tasks using power nodes.z-image-turbo__t2i_double_trouble.json: Text-to-image workflow combining two different styles.z-image-turbo__text2image.json: Main reference text-to-image generation workflow.
Requirements
To utilize any of these workflows, you need to have ComfyUI configured and the "Z-Image Power Nodes" installed. Additionally, ensure that the required checkpoints (in GGUF or Safetensors format) are placed in the appropriate directories within your ComfyUI setup.
If you choose to use GGUF-format checkpoints, it is necessary to install the "ComfyUI-GGUF" nodes as well, since ComfyUI does not natively support GGUF files. You can find information about these nodes at: https://github.com/city96/ComfyUI-GGUF
Below in this README, you will find detailed instructions on how to install the necessary nodes and files.
Please be aware that although I tested "Z-Image Power Nodes" using the recommended checkpoints below, it should also be compatible with other "Z-Image Turbo" checkpoints and LoRAs. However, I cannot guarantee full functionality for all alternative combinations; you may need to tweak the workflows to optimize them for your own setup.
Checkpoint Files
The following list includes the pre-configured checkpoints used in the workflows. I chose these specifically because they performed best during my testing phase. However, given the diversity of GPUs, VRAM, and ComfyUI versions, it is very difficult to determine which one will work best on most systems. Therefore, I recommend testing them all to find the one that works best for you.
GGUF (Q8/Q5)
-
"z_image_turbo-Q5_K_S.gguf" |5.19 GB| Download
Local Directory:ComfyUI/models/diffusion_models/ -
"Qwen3-4B-Q8_0.gguf" |4.28 GB| Download
Local Directory:ComfyUI/models/text_encoders/ -
"ae.safetensors" |335 MB| Download
Local Directory:ComfyUI/models/vae/
Safetensors (FP8)
-
"z-image-turbo_fp8_scaled_e4m3fn_KJ.safetensors" |6.16 GB| Download
Local Directory:ComfyUI/models/diffusion_models/ -
"qwen3_4b_fp8_scaled.safetensors" |4.41 GB| Download
Local Directory:ComfyUI/models/text_encoders/ -
"ae.safetensors" |335 MB| Download
Local Directory:ComfyUI/models/vae/
Safetensors (BF16)
-
"z_image_turbo_bf16.safetensors" |12.3 GB| Download
Local Directory:ComfyUI/models/diffusion_models/ -
"qwen_3_4b.safetensors" |8.04 GB| Download
Local Directory:ComfyUI/models/text_encoders/ -
"ae.safetensors" |335 MB| Download
Local Directory:ComfyUI/models/vae/
Z-Image Power Nodes
Z-Image Power Nodes can be installed via the ComfyUI Manager or downloaded from its respective repository.
Installation via ComfyUI Manager (Recommended):
- Open ComfyUI and click on the "Manager" button to launch the "ComfyUI Manager Menu".
- Within the ComfyUI Manager, locate and click on the "Custom Nodes Manager" button.
- In the search bar, type "Z-Image Power Nodes".
- Select the option from the search results and click the "Install" button.
- Restart ComfyUI to ensure the changes take effect.
Manual Installation:
For manual installation, follow the instructions provided in the GitHub repository of the project: