* This blog post is a summary of this video.

Install Stable Diffusion XL on Your Local Windows Machine

Author: MattVidPro AITime: 2024-03-23 06:25:00

Table of Contents

Check If Your Computer Can Run Stable Diffusion XL

The first step is to check if your computer meets the system requirements to run Stable Diffusion XL smoothly. You'll need a decent GPU with at least 8GB of VRAM for reasonable performance. To check your GPU and VRAM, open Task Manager in Windows and look under the Performance tab.

If you don't have 8GB of VRAM, you may still be able to run Stable Diffusion XL using your CPU instead of GPU. However, generation will be much slower - up to 30 minutes per image. Running on CPU is only recommended for testing purposes if you lack the GPU requirements.

Use Task Manager to Check GPU and VRAM

Hold down the Windows key and press X to open the power user menu. Click on Task Manager. Go to the Performance tab. Look for your GPU and check the dedicated GPU memory, which shows your VRAM. You need at least 8GB for a good experience.

Minimum 8GB VRAM Recommended

For smooth performance with Stable Diffusion XL, your GPU should have at least 8GB of VRAM based on community feedback. More VRAM allows you to generate higher resolution images faster.

CPU Mode Alternative for Low-End Systems

If your system doesn't have a capable GPU, Stable Diffusion XL can run in CPU mode. This uses your CPU for generation instead of GPU. However, it's much slower - taking up to 30 minutes per image. CPU mode allows low-end machines to test and use the AI.

Download Stable Diffusion XL and Comfy UI

The next step is to download the Stable Diffusion XL model file and Comfy UI. Comfy UI is an open source app that provides an interface for generating images with Stable Diffusion. Links to download both files are in the video description.

The Stable Diffusion XL file is around 6GB and Comfy UI is 1.4GB, so make sure you have decent internet speeds to download these large files reasonably quickly. They are hosted on reputable sites for open source AI content.

Install 7-Zip to Extract Comfy UI

Comfy UI downloads as a zip file, so you need to extract it before use. A free app called 7-Zip allows you to easily extract zip files. Download and install 7-Zip, then right click the Comfy UI zip file, choose 7-Zip file manager, and extract it to a location like your desktop for easy access.

Locate Checkpoints Folder in Comfy UI

Inside the extracted Comfy UI folder is a checkpoints folder. This is where you will later move your Stable Diffusion XL model file so Comfy UI can access it.

Move Stable Diffusion Checkpoint File

Now navigate back to your downloads folder and locate the Stable Diffusion XL file you downloaded earlier. Move this file into the checkpoints folder inside Comfy UI. This transfers the model file to where Comfy UI expects it.

Launch Comfy UI and Generate Images

Understand Comfy UI Interface

Double click the 'run Nvidia gpu' file inside Comfy UI to launch it. This opens the web app interface. The main sections are prompt fields to describe the image to generate on the left, advanced settings in the middle, and the generation button on the right.

Customize Settings as Needed

You can tweak settings like image dimensions in the middle column. Default settings work well for starting out. The Stable Diffusion XL model you downloaded earlier should show as the loaded checkpoint.

Hit Generate and View Outputs

Type a text prompt describing the image you want to create, then press Generate. Once it finishes, find the output images in Comfy UI's output folder to view and save creations. Enjoy exploring what Stable Diffusion XL can design!

Get Help from Video Author and Community

If you have any issues installing or running Stable Diffusion XL with Comfy UI, please post in the comments or reach out on Discord. The open source creative AI community is very helpful and can likely assist with getting your system fully up and running smoothly.


Q: What are the system requirements to run Stable Diffusion XL?
A: You need at least an Nvidia GPU with 8GB or more VRAM. 16GB VRAM is recommended for best performance. Some high-end CPUs can run it without GPU but very slowly.

Q: Is there an AMD GPU version?
A: Currently Stable Diffusion XL works best on Nvidia GPUs. You can try to run it in slow CPU mode on AMD systems.

Q: Where do I get help if I have issues installing or running it?
A: You can ask the video creator or join their Discord community for help in the AI Help section.

Q: Can I generate uncensored images?
A: Yes, using the open source Stable Diffusion XL allows generating uncensored imagery on your local machine.