CivArchive
    Drag light to eat color - V1
    Preview 1
    Preview 2
    Preview 3
    Preview 4
    Preview 5
    Preview 6

    This model is specifically designed to generate high-quality, intricate modern anime and Japanese illustration styles. The core style characteristics lie in clear, precise lines and vibrant, rich color usage. Lighting and shadow effects are meticulously handled, adding depth and a sense of volume to both characters and backgrounds. The model excels at capturing character facial expressions and complex details, with detailed rendering of intricate clothing textures, diverse hairstyles, and environmental elements (such as flowers, plants, and interior decor). Compositions are flexible and varied. The overall aesthetic presents a clean, modern, and cinematic illustrative style, making it ideal for scenes requiring high levels of detail and artistic flair.

    Recommended settings:

    Steps: 15-30

    CFG scale: 5-7

    Sampler: Euler a

    The training set incorporates a large amount of exquisite line art content

    Description

    CHECKPOINT
    Illustrious

    Details

    Downloads
    53
    Platform
    TensorArt
    Platform Status
    Available
    Created
    4/15/2026
    Updated
    4/15/2026
    Deleted
    -

    Files

    拽光吃色V1.safetensors

    Mirrors

    TensorArt (1 mirrors)
    TensorHub (1 mirrors)