Some months and... years ago Support for Forge main Branch and Automatic 1111 stop and community advanced on branching and giving support to those new branches.
Forge Neo is the current branch with that support and I'm writing here a guide to install it without modifying your system main installs, so you will be able to render Z-Image Turbo models.
So, straight and forward ->
Installing Forge Neo
1 ) First we need to install uv
pip install uv
2 ) Then Download Python 3.11 Standalone ( Not the installer )
https://www.python.org/ftp/python/3.11.9/python-3.11.9-embed-amd64.zip
Extract the zip file on a secure location, example on my machine: [ C:\Dev\python\python-3.11.9-embed ]
3 ) Clone Forge Neo branch
git clone https://github.com/Haoming02/sd-webui-forge-classic sd-webui-forge-neo --branch neo
4 ) Go to the sd-webui-forge-neo folder then create the venv with this ( the path to python is used on this example ):
uv venv venv --python C:\Dev\python\python-3.11.9-embed --seed
5 ) Edit the webui-user.bat file
Add to the set COMMANDLINE_ARGS the next flag: --uv
set COMMANDLINE_ARGS=--uv
6 ) Then activate venv
venv\Scripts\activate
7) Once venv is activated run the webui-user.bat
If everything goes as expected all the dependencies will be downloaded and installed and the browser at the end will jump with Forge Neo running.
Rendering the Models
Those 4 Folders are the most needed, lets work with my model [ Catwoman_AnneHathaway_ZImage_smf_v01.safetensors ] as example:
1) Lora Folder is where Catwoman_AnneHathaway_ZImage_smf_v01.safetensors goes.
This model is exclusive for Subscribed members :)
2) Stable-Diffusion Folder is where z_image_turbo_bf16.safetensors checkpoint goes.
Download it from here: https://huggingface.co/SmoonHacker/checkpoints/tree/main
3) text_encoder Folder is where qwen_3_4b.safetensors goes.
Download it from here: https://huggingface.co/SmoonHacker/TextEncoders/tree/main
4) VAE Folder is where ae.safetensors goes.
Download it from here: https://huggingface.co/SmoonHacker/VAE/tree/main
Guide you with this image to configure the checkpoint, the VAE/Text Encoder and the other Generation setting to render it.
The most easy way to load that config is , Go to the PNG Info tab and load the image added on this post attachments:
Then Hit the Send to txt2img Button
Advanced Config on the webui-user.bat File ( LORAs / Checkpoints / Text Embeddings ) custom Folders.
Some of us we have already an old install and we don't want to move every LORAs , Checkpoint or Text Embeddings from there, so... we can configure the webui-user.bat to point to our old or any specific folders. I attached on this post the modified webui-user.bat and find it as example for your purpose.
Also I included on the attachments the bat file runForgeNeo.bat, this one goes on the sd-webui-forge-neo folder, this will run from the activated venv the webui-user.bat on a double click.
If you have any issues, please post it on the comments and Ill try to help.
Cheers!