The next evolution of SteinXL, once again bringing together various aspects of my recent, disparate merges, while tying then back into their roots as a more cohesive model. Improvements to anatomy, style, and prompt adherence, though it leans heavily into digital/anime/cartoon styles, rather than realistic generation. VAE is baked in, but if for some reason it's not working, this model uses sdxl.vae.
The next evolution of SteinXL, once again bringing together various aspects of my recent, disparate merges, while tying then back into their roots as a more cohesive model. Improvements to anatomy, style, and prompt adherence, though it leans heavily into digital/anime/cartoon styles, rather than realistic generation. VAE is baked in, but if for some reason it's not working, this model uses sdxl.vae.