Ok, so I have been playing around with mage.space for a bit but they now have it as a pay for play service if you wanna do anything nsfw. I’m worried my computer is just not powerful enough for the AI software I have seen used here. Does anyone have alternatives?

I've been using this google colab notebook https://colab.research.google.com/github...1111.ipynb. Now I can generate AI images on my phone if I want to. It takes some learning but its totally worth it if you want to fall down a real rabbit hole of creativity.

Seconding that recommendation. The arbitrary time limitations of using the free version of Colab are more than made up for the extra control you have over the options compared to any app or website. If you can run "Hello World" in python you can probably get it working; I found it far easier to set up the linked Colab notebook than it was to get the same webUI running locally.

A few tips:

Create and use a dedicated google account that is only for running the colab and using google drive in conjunction with that. It's a security vulnerability to use your main. Also installation is about 9-10gb and the AI models can easily push the total over the maximum of 15gb for the free tier of Google Drive. It's tight but it's enough to work with and even dabble in blending models. 

Download the models you want to use prior to loading the colab. There's many official models available on huggingface.co, and all sorts of stuff on civitai.com.

The first time it runs it takes a long time to set itself up, subsequent uses will be much quicker. 

You can use the webui's merge checkpoint feature to convert 32-bit models to 16-bit which halves their size on disk, important as free google drive only gives you 15GB, helps them to load quicker, and seems to improve performance as well on the colab with no noticeable drop in output quality. Just select one as checkpoint 'A', set the slider to 0, and select "save as float16". You can then delete the original. If a model is ~2gb it's already 16bit;

If you are confident to actually read a bit of code, it's worth running the original, non-webUI Stable Diffusion colab to get a better idea of what's happening behind the scenes. Understanding the process and what the parameters actually mean will help you to get the best results possible.

Lastly, as far as paid AI art services go mage.space is very competitively priced. There may be other good ones but what I've seen is overpriced considering it's free to run it yourself. Otherwise there's plenty of guides and videos on how to run Stable Diffusion on a rented GPU.