Installing Stable Diffusion SDXL

Hello Algorithm Architect.

Thank you for being a subscriber 💛

If you like what I do, consider checking out my guides. Or sharing the newsletter with your friends. There is a referral link at the bottom of this post.

If you really like what I do and want to support me in my efforts, consider subscribing.

Let’s get to today’s topic - How to use SDXL model with Stable Diffusion.

This new model is HUGE.

Think Midjourney quality with more control over your image.

And it’s not as difficult to operate as you might think.

Ok, it can be difficult, if you want to dive really deep. But there is also a basic image workflow I want to show you today.

First, we need to download the interface called ComfyUI. It’s a node-based interface. If you’re familiar with those, you’re gonna have a lot of fun.

There is an installation section on the GitHub page with a Direct link download. It’s a one-click installer. Crazy! 🤯

What a dream.

Once you’ve extracted the files, download the SDXL model + refiner and put them into the models folder inside ComfyUI folder you extracted.

After you have the models downloaded, you can click on the run_nvidia_gpu file.

This will launch the terminal and open an empty UI inside your browser.

Now comes the really good part about this.

You can download any image generated with the SDXL model and drop it directly into this interface. It will load the workflow used to create that image automatically.

For example, I used this image, which I downloaded here.

And it loaded this workflow inside the ComfyUI.

It’s a great way to learn as well, as you can investigate each step on how the nodes are connected and try modifying it for your own purpose.

This specific image also loads descriptions for each of the nodes, so you can easier understand what is going on.

At the basic level, you can rewrite the prompt and generate new images with the same workflow.

I know this might feel intimidating. Using Midjourney still feels easier and more comfortable.

The control this setup offers is incredible though. You can change, modify and add to any step of the image generation process. And the fact that you can load other people’s workflows makes it very very easy to learn and experiment.

I still believe Stable Diffusion and it’s various Interface options are the place to be for creative professionals. We want autonomy and control over our creative process and other platforms have a lot of limitations when it comes to that.

I hope this post will inspire you to give ComfyUI + SDXL a try. If only to play around and have fun.

You can stop by my website or social media and as always,

Keep creating and don't forget to have fun. ☀