Newer version is an alternative option, not an upgrade. Both have their own art styles.

Anime model based on several models available here and elsewhere.

I used CLIP SKIP 1, HIRES. FIX and ADETAILER on all the sample images.

They are all simple generations, just to show off what this model can do in its basic state. So no LORA was applied as well.

For hires. fix, I used 4x-UltraSharp model.

Quantization in K samplers is enabled in my settings.

VAE is baked in. (kl-f8-anime2)

I tend to use "ornate & intricate details" a lot in my generations to crank up the details, which are already enhanced by Easynegative & Badhandv4 textual inversions.

I'd like to thank Skumerz, whose model I used as a starting point to create is merge.

If you are interested, please check out my semi-realistic anime model : Mavka