From Anime to 2.5D: Introducing 6 Captivating Stable-Diffusion Models

This site uses affiliate links.

Introducing Six Latest Models

Hello, I’m kojirom.
Here are the six latest models I’ll introduce this time.

DuboisC50.3D – v5.0

helloL3 – helloL3_v1.0h

NaGi Mix – v1.2

RaemuMix – v7.1

RealCartoon-Anime – V8

TruePettanMix – v1.0

Three Challenges to Test the Models’ Descriptive Powers

I gave each model three prompts to compete on their descriptive abilities.

  1. a girl in a leotard,
  2. a woman in a swimsuit,
  3. a 70-year-old Japanese woman in a high-end one-piece dress,

Points of Particular Attention

The prompt for “a 70-year-old woman” is a bit mischievous, but I included it on purpose because I think there are people who use AI images in manga and such. The reason for specifying “in a high-end one-piece dress” is because if you include the prompt “Japanese,” the rate of generating images of women in traditional Japanese attire becomes high. Please pay attention to whether the models follow the instructions and depict a one-piece dress properly.

The Importance of Distinguishing Character Age Groups

If a model can differentiate between women in their teens, 20s, and 70s, it’s probably safe to say it can generate most character types. However, just because a model can’t differentiate between various characters doesn’t mean it’s inferior. Essentially, if the model can generate images suitable for company or school events… in other words, the ultimate ethically and socially acceptable images we all seek, that’s more than enough. I hope for your understanding in this matter.

Introducing Images Available for Commercial Use

On this blog, we only introduce models that are clearly marked as available for commercial use. All 306 illustrations posted here can be used freely within the bounds of good sense and common decency.

Illustration Gallery (306 images)

Models are listed in alphabetical order


helloL3 – helloL3

NaGi Mix




  • The illustrations on this site are available for commercial use.
  • Some illustrations may have imperfections in details such as hands and fingers. Please excuse these minor issues.
  • When using these images, you are encouraged to crop and adjust them as you see fit.
  • Please make the most of your creativity and use the illustrations freely and imaginatively.


を よろしく

にほんブログ村 イラストブログ CG・デジタルイラストへ

Recommended Article

Stable-Diffusion Configuration Parameters

Positive prompt: a girl in a leotard,
Negative prompt: easynegative,
Steps: 16, Sampler: DPM++ SDE Karras, CFG scale: 5.5, Seed:
3370008135, Size: 512×512, Model hash: 812e47e175, Model:
truepettanmix_v10, VAE hash: 40da092002, VAE:, Denoising strength: 0.65, Clip skip: 2,
Hires upscale: 1.5, Hires steps: 60, Hires upscaler: Latent (bicubic
antialiased), TI hashes: “easynegative: c74b4e810b03”, Eta: 0.667,
Version: v1.6.1