>>76284489just going to clear up some terminology usage, just based on some things you've said.
model: large set of data that tells the ai how to create an image.
lora: small set of data that you tack onto a model to fill in some gap of knowledge, e.g. a character or artist style.
vae: some kind of something that "finishes" the image. not really sure what it is exactly, i think there's only like one or two that are commonly used?
NAI: novelai, a website and paid image generation service that has their own secret model. An old version of it was leaked and formed the basis for many other models. "NAI speedrun" refers to that leaked model I believe. It's like over a year old?
SD: stable diffusion, the basis for ai generation. was 1.5, recently overtaken by
XL: stable diffusion xl, just... higher res pictures or something. the new hotness or something. i dunno man I pay for my NAI and have only dabbled in the local generation stuff. they told me XL was what to go for.
pony:
https://civitai.com/models/257749/pony-diffusion-v6-xlanimagine:
https://civitai.com/models/260267/animagine-xl-v31artiwaifu:
https://civitai.com/models/435207/artiwaifu-diffusionautism mix is also popular i hear:
https://civitai.com/models/288584?modelVersionId=324524 