CivArchive
    Vete ┃ Trash Angel ILL - v.2
    NSFW
    Preview 111268328
    Preview 111297365
    Preview 111268325
    Preview 111268322
    Preview 111268323
    Preview 111268330
    Preview 111268326
    Preview 111268324
    Preview 111268327

    Vete ┃ Trash Angel ILL is a merged and finetuned high-end model designed for creators who want clean anime rendering with premium lighting, controlled gloss, and consistently accurate output. Its style blends soft anime beauty with sharp clarity, delivering polished characters with stable anatomy, expressive eyes, and smooth tonal transitions.

    This model specializes in crisp linework, clean facial structure, high-accuracy shading, and balanced lighting control—handling warm daylight, neon tones, soft interiors, and dramatic contrast without muddy colors or artifacts. Highlights render naturally, skin finishes stay smooth and even, and details like hair strands, accessories, and fabrics maintain precise definition across all prompts.

    Trash Angel ILL aims for professional-grade output: minimal distortions, reliable consistency, and a cohesive aesthetic that remains stable even in complex scenes or dynamic poses. Every generation feels intentional, refined, and visually clean.

    ___

    Recommended Danbooru Tags (Style-Focused):

    For Clean & Soft Output:

    (clean_anime), (soft_lighting), detailed_hair, smooth_shading, pastel_colors, soft_blush, warm_tone, high_accuracy, fine_lineart

    ___

    For Glossy & High-Contrast Output:

    (glossy_skin), dramatic_shading, rim_light, cinematic_lighting, vibrant_colors, precise_highlights, contrast_control, sharp_shadows

    ___

    Creator:

    Vetehine

    Discord: Vetehine

    Linktree: https://linktr.ee/Vetehine
    ____
    “Adetailer ensures cleaner, more accurate faces and eyes.”

    ________________________

    Recommended Settings -

    Clip Skip - 2

    DPM++ 2M Karras

    Sampling Steps - 25-50 (Recommended: 35-42 Steps)

    Cfg - 6-7

    ________________________

    Recommended Settings -

    Clip Skip - 2

    Euler Ancestral

    Sampling Steps - (Recommended: 25-50 Steps)

    Cfg - 6-7

    ________________________

    Recommended Settings -

    Clip Skip - 2

    DPM++ SDE Karras

    Sampling Steps - 25-50 (Recommended: 35-42 Steps)

    Cfg - 6-7

    ___

    Recommended Settings -

    Clip Skip - 2

    Restart

    Sampling Steps - Recommended: 66 Steps

    Cfg - 6-7

    ________________________

    Hires-

    R-ESRGAN 4x+ Anime6B, 4x-AnimeSharp

    Steps 10 - 20

    Denoising 0.2 - 0.3

    ___________

    Model: https://moescape.ai/models/2390f43d-d617-469d-b463-3b88466578ea?index=2&modelversion=e82eeecd-47b7-4b18-beea-140cb67712be

    Description

    V2 brings a big jump in overall quality cleaner renders, better fingers, stronger pose stability, and sharper details. Lighting is smoother, shading is more natural, and the style stays consistent across all prompts. This version simply outputs more reliable, polished results than V1.

    FAQ

    Checkpoint
    Illustrious

    Details

    Downloads
    402
    Platform
    CivitAI
    Platform Status
    Available
    Created
    11/23/2025
    Updated
    4/28/2026
    Deleted
    -

    Files

    veteTrashAngelILL_v2.safetensors

    Mirrors

    Available On (1 platform)

    Same model published on other platforms. May have additional downloads or version variants.