CivArchive
    VibrantAnime_FlatColor-moss image - v1.0
    NSFW
    Preview 148428168
    Preview 148428190

    Introduction
    This LoRA aims to replicate a highly visually striking retro-pop anime style. Its core features are extremely thick and sharp black outlines, hard-edged cel shading, and highly saturated colors. The artwork presents a fascinating aesthetic that blends elements of early retro animation, American comics, and modern pop culture, making it perfect for generating character illustrations with a strong personality and an edgy or "little devil" vibe.

    Core Style Features

    • Thick Lineart: The outer contours and structural lines are extremely bold and crisp, with no messy extra lines, delivering a strong comic/manga feel.

    • Crisp Cel Shading: Shadow edges are sharply defined with no soft transitions. Highlights (such as white blocks on hair and skin) are highly prominent and concretely shaped.

    • Vibrant Colors: The colors are extremely vivid and bright. Solid colors like red, blue, and pink are used extensively, creating a powerful visual contrast.

    • Detailed Facial Features: The eye design retains the delicate touch of Japanese anime, featuring rich, multi-colored pupils (e.g., gradients, radial patterns). This creates a strong contrast with the rough outer outlines.

    Recommended Generation Parameters

    • Base Model: Trained on the official MOSS-Image-SFT model. It is recommended to use this base model or other fine-tuned anime models.

    • LoRA Weight: 1

    • Sampler: Euler a or DPM++ 2M Karras

    • Trigger Word: dedeen_style



      本 LoRA 旨在复刻一种极具视觉冲击力的复古潮酷二次元风格。其核心特征为极粗且锐利的黑色轮廓线、硬边缘平涂(Cel Shading)以及极高饱和度的色彩。画面呈现出类似早期复古动画、美式漫画与现代社会流行元素碰撞的奇妙质感,非常适合用来生成具有个性、带点“小恶魔”或“潮酷”属性的角色立绘。

      画风核心特点

      粗犷描边 (Thick Lineart):外轮廓和结构线极其粗重、利落,没有多余的杂线,带来强烈的漫画感。

      硬核平涂 (Crisp Cel Shading):阴影边缘一刀切,没有柔和的过渡,高光(如头发、皮肤上的白色色块)非常醒目且形状具象。

      高饱和色彩 (Vibrant Colors):色彩极其鲜艳明快,红、蓝、粉等纯色大量使用,视觉对比度极强。

      精致五官:眼睛画法保留了日系的精致感,瞳孔内部色彩丰富(如渐变、放射状),与粗犷的外轮廓形成强烈的反差萌。

      推荐使用参数

      底模 (Base Model):训练使用了官方模型MOSS-Image-SFT,推荐使用本模型或其余二次元微调模型

      LoRA 权重 (Weight):1

      采样器 (Sampler):Euler a 或 DPM++ 2M Karras

      触发词:dedeen_style

    Description

    FAQ

    LORA
    MOSS Image

    Details

    Downloads
    109
    Platform
    CivitAI
    Platform Status
    Available
    Created
    11/9/2026
    Updated
    11/10/2026
    Deleted
    -

    Files

    VibrantAnime_FlatColor-moss image-v1.safetensors