The 'old/bad' version is the original upload. It's decent but often has weird image artifacting on due to one lora i used it the merge. That lora got retrained and now made into the 'v1' version.
Aetherveil: Where softness meets the edge of something eternal.
A versatile anime-style checkpoint focused on clean rendering, cinematic composition, and strong visual storytelling.
This model blends multiple aesthetic domains into a cohesive style:
soft, expressive anime portraits
detailed anthro / furry characters
fantasy and sci-fi hero compositions
modern cyberpunk portrait lighting
occasional high-contrast ink / graphic styles
It excels at producing images with:
vibrant yet balanced color palettes
soft natural lighting and atmospheric skies
dynamic wind, particles, and environmental motion
strong composition (low-angle shots, centered subjects, scenic framing)
clean linework with smooth shading and high detail
Characters tend to have:
expressive eyes and soft facial rendering
detailed hair with natural flow
well-defined clothing folds and materials
consistent anatomy with a slightly stylized anime proportion
Strengths:
Outdoor scenes (fields, skies, mountains, seasonal settings)
Anthro / furry characters with clean fur detail
Elegant outfits & dresses (flowy fabric, soft lighting)
Cinematic hero shots (low angle, dramatic framing)
Soft, cozy moods and intense high-contrast pieces
Works well across both fantasy and modern settings
Recommended Usage
Sampler: Euler a
Steps: 25–30
CFG: 4
Prompt tips:
Use lighting descriptors (soft light, rim light, sunset, bloom)
Add environmental motion (wind, petals, leaves, particles)
Specify camera angle (low angle, portrait, cinematic framing)
Works well with both simple prompts and rich descriptive prompts
Notes:
This model is intentionally flexible and can shift between:
soft, wholesome aesthetics
dramatic, high-contrast stylization
detailed character-focused renders
For best results, guide the style using:
color mood (warm autumn, cool blue, neon, etc.)
lighting type
environment