Multi-Artist Anime Illustration Style LoRA (v1.0)

An experimental LoRA that integrates four distinct commercial anime illustration style profiles into a single model.
Each style is activated via a dedicated trigger token and designed to function independently, with optional controlled blending.

Included Style Profiles

• so-bin inspired
  Heavy decorative line presence, dramatic layered shading, strong contrast
  (Overlord-style visual direction)

• shinotsuki shinobu inspired
  Sharp, clean structural lines with precise detail control
  (Youjo Senki-style illustration tone)

• mishima kurone inspired
  Simplified outlines, bright colors, expressive and cartoon-like facial design
  (KonoSuba-style aesthetic)

• shinichirou otsuka inspired
  Soft integrated linework, delicate coloring, intricate hair rendering, gentle lighting
  (Re:Zero-style illustration finish)

Usage Notes

• Use one style trigger for the clearest and most faithful result

• Multiple triggers can be combined for experimental hybrid styles, though some feature blending may occur

• Recommended LoRA weight: 0.6 – 1.0

• Captions are intentionally kept minimal to preserve stylistic clarity

Current Status

This is v1.0 of an ongoing multi-style experiment.
While style separation is generally stable, minor cross-style influence may still appear in complex scenes.
Further refinement and dataset tuning are planned for future updates.

Thank you for your support and feedback.