CivArchive

    This LoRA is designed to generate vibrant, clean anime artwork featuring bold, flat colors with minimal shading and depth—perfect for a striking, 2D-inspired aesthetic reminiscent of classic cel-shaded animation. It's trained on high-quality datasets to capture smooth gradients and vivid palettes while keeping things simple and eye-catching. Ideal for character designs, scenes, or illustrations that pop without complex details.

    Description

    this a new vesion of my flat anime lora from image to video, you can also do text to video its will came out good as long the character is not naked,

    FAQ

    Comments (7)

    hboxgames132Mar 29, 2026
    CivitAI

    I wish LTX could speak japanese lol

    tosermeplsMar 29, 2026

    It can.

    hboxgames132Mar 29, 2026

    @tosermepls  ? trying right away

    HinoriMar 29, 2026
    CivitAI

    Can you train this LTX2.3 lora for I2V as well? Seems to have better results on T2V then I2V

    Agino
    Author
    Mar 29, 2026

    i dont think you can train only one just image to video. my guess the ltx-2.3 is better with anime image that 2.2

    GlowingGuardianGirlApr 11, 2026
    CivitAI

    Hello 🙌 can you give us some insights about your training method for LTX2.3? Thank you

    cooperdkApr 25, 2026
    CivitAI

    Nice!

    But this is NOT a flat style.