CivArchive
    Nya-Iris - Anima - Base1 - v1.0
    NSFW
    Preview 132833332
    Preview 132153736
    Preview 132773509
    Preview 132667481
    Preview 132152170
    Preview 132800695
    Preview 132795167
    Preview 132795310
    Preview 132667422
    Preview 132795481

    Christmas in May — Anima 1.0 is out! c:

    TAGS
    [artist] [style] [color palette] [1girl/1boy/etc] [character] [series] [others]
    Quality scores: No more positive quality tags..
    Negative: score_1, score_2, score_3, worst quality, low quality, blurry, lowres, extra fingers, bad feets, bad eyes, logo, noisy, watermark, signature
    The watermark consists of a pair of trigger words and that display (~in the case of this model) a "text" in the middle of the image; this lets me know which model is merged with this one, no matter how many times it's mixed.
    ✦•··· Recommended Settings ···•✦

    Steps: 12 ~ 40 (Civitai generator 20+)
    CFG scale: 5     ~ 6
    Sampler: EulerA / dpmpp_2m_sde_gpu - SGM_Uniform / Simple/ Beta

    Hires-fix: Same settings you used to generate the image..

    Description

    Second Iteration 𐔌՞. .՞𐦯

    • Anima Base-v1.0

    • This version took on a style that I'll build upon.
      Fewer random artifacts, ~ imo...

    FAQ

    Comments (26)

    SkyLaptorJun 5, 2026· 3 reactions
    CivitAI

    Using the built-in CLIP results in an error, so I'm using the standard CLIP (qwen3.06b) only in that case.

    Karely_ai
    Author
    Jun 5, 2026

    It doesn't have a clip or vae built-in, but thanks for reminding me to upload that

    SkyLaptorJun 5, 2026· 1 reaction

    I see, I thought everything would be included since it's a checkpoint. Thanks for the reply.

    Lizardon1025Jun 5, 2026· 8 reactions
    CivitAI

    I have tested every Anima model so far and Nya-Iris Anima is one of my favorites so far, thank you so much for creating this! <3
    However when I compared your Base v1.0 with the previous Preview3 v0.1, I still prefer the preview version as the base v1.0 version is a bit brighter and has less contrast than the previous one. Maybe you can fix that in upcoming versions? :)

    PamelaEunomieJun 9, 2026· 2 reactions

    Thanks for your article/review, currently looking at anima models, and you definitely helped speed up the process!

    Lizardon1025Jun 9, 2026

    @PamelaEunomie I'm glad that I could help you ^^

    LazyMofoJun 11, 2026

    feels the same~

    xiaobai_solonJun 6, 2026· 5 reactions
    CivitAI

    Feedback from trial: This model has great aesthetics, delivering clean, crisp colors with a hint of eroticism. The contrast is sharp and clear. The character composition and sizing resemble the Illustrious model, making it ideal for illustrations. The images tend to be slightly bright and don't respond well to 'dark' prompts, so it's perfect for creating bright, highly-detailed scenes.

    KwiwiJun 6, 2026· 5 reactions
    CivitAI
    the images come out surprisingly detailed. it understands the positive prompt well but is not very flexible on styles and certain tags

    atisomaia3Jun 14, 2026· 1 reaction
    CivitAI

    tested this model out with 8 steps and 0 negative prompt came out really good!

    ognomekop97647Jun 15, 2026· 1 reaction
    CivitAI

    why this model size (4.1G) are larger than original base model(3.9G) ?

    Karely_ai
    Author
    Jun 15, 2026

    My honest reaction to that question: ┐(￣ヮ￣)┌

    Seriously though, apparently Anima models have between 560 and 685 tensor layers (Anima-base-v1.0 has 685), but these models have 879... (⊙_⊙)

    ognomekop97647Jun 16, 2026

    @Karely_ai Is these improve for better image compre to base?🤔🤔🤔

    InemanJun 16, 2026
    CivitAI

    Very good model. But it make naked ass too big. And can not make full nelson with from side.

    zzozzJun 17, 2026· 2 reactions
    CivitAI

    If hand gestures are not specified, there is a very strong tendency for it to assume the Salute pose.

    There are also symptoms of awkward finger movements, but I am not sure if this is a limitation of Anima or FineTune.

    zzozzJun 17, 2026

    nevermind. (anime shading:4) was the problem.

    InemanJun 17, 2026

    I think same. It used to make hand up over face.

    Aemeath1Jun 17, 2026· 5 reactions
    CivitAI

    Feels like this is still the best anima model out there...

    Dante3rdJun 19, 2026· 2 reactions
    CivitAI

    beautiful checkpoint, but upscaler recommended? remacri maybe?

    Karely_ai
    Author
    Jun 19, 2026

    I use 4x_foolhardy_Remacri or 2x-AnimeSharpV4_RCAN for hi-res fix, but for tiled upscaling I use realESRGAN_x4plus.

    Dante3rdJun 19, 2026

    @Karely_ai oh thx so much

    and u use comfyui or other for image generation?

    Karely_ai
    Author
    Jun 19, 2026

    @Dante3rd ComfyUI bcuz there's no other way on macOS..

    Dante3rdJun 19, 2026

    @Karely_ai Oh, I see, although it's strange because the workflows aren't showing up in my ComfyUI xd

    Anyway, great checkpoint, bro n.n

    kuzunoha54936Jun 22, 2026· 2 reactions
    CivitAI

    The only thing I don't like is that the gold trim appears frequently without my intention.

    frigidblast381Jun 23, 2026· 6 reactions
    CivitAI

    wow, very good model. one of my favorite!!

    Karely_ai
    Author
    Jun 28, 2026· 11 reactions
    CivitAI

    ꒰ᐢ. ̫ .ᐢ꒱ v2.0 ready by Thursday or Friday..

    Checkpoint
    Anima
    by Karely_ai
    Download (Beta) View on CivitAI
    illustration
    anime
    anima
    base model
    digital art
    colorful
    cartoon
    fantasy
    2.5d
    2d

    Details

    Downloads
    2,285
    Platform
    CivitAI
    Platform Status
    Available
    Created
    6/5/2026
    Updated
    8/13/2026
    Deleted
    -

    Files

    qwen_image_vae.safetensors

    Size:
    242.05 MB
    SHA256:
    a70580f0213e67967ee9c95f05bb400e8fb08307e017a924bf3441223e023d1f

    Mirrors

    HuggingFace (220 mirrors)
    CivitAI (166 mirrors)
    TensorFiles (1 mirrors)
    ModelScope (3 mirrors)
    ModelScope CN (5 mirrors)

    nyaIrisAnima_base1V10_txt.safetensors

    Size:
    1.11 GB
    SHA256:
    cd2a512003e2f9f3cd3c32a9c3573f820bb28c940f73c57b1ddaa983d9223eba

    Mirrors

    HuggingFace (103 mirrors)
    CivitAI (151 mirrors)
    ModelScope (1 mirrors)
    ModelScope CN (2 mirrors)

    nyaIrisAnima_base1V10.safetensors

    Size:
    4.13 GB
    SHA256:
    b55f11f44bd0cb972ba17ebb2c28983a1443b039f72a94f1a8eabdc7deebfd1b

    Mirrors

    HuggingFace (1 mirrors)
    CivitAI (1 mirrors)