CivArchive
    BEST Froot model. bsapricot, apricot the lich, VSHOJO - v2.0
    NSFW
    Preview 32774202
    Preview 32774211
    Preview 32774234
    Preview 32774268
    Preview 32774274
    Preview 32774276
    Preview 32774314
    Preview 32774335
    Preview 32774345
    Preview 32774393
    Preview 32774399
    Preview 32774400
    Preview 32774403
    Preview 32774408
    Preview 32774445
    Preview 32774464
    Preview 32774906
    Preview 32775102
    Preview 32775257
    Preview 32775452

    Includes Froots (at the time of writing) newest model, the one made by Nanoless.

    Support both of them.

    # About this model

    This is THE best Froot model. This model is quite adaptive to different outfits, and the images of Froot that it generates have a nice color, likely an influence from Nanoless' artstyle.

    RECOMMENDATION, for specific Froot outfits, add weight: (frootvampidol:1.2)

    The main tags for the models are:

    froot, frootrace, froothoodie, frootvampidol, frootelegant, frootmodern, frootcheerleader

    There are some other less used tags. If you are in the superior webui, ComfyUI, you can right click the Lora, and see 'View Lora info...'

    # What has changed?

    The model includes the newest Froot outfit and has nice coloring. The labels of the training images were changed, these new labels don't work too well with all of Froots outfits, especially the hoodie one and the midriff one don't do too well in this model, but maybe you can combine this model with the previous version. This may sound bad, but really, this new version is better. Also, this does better with the vampidol outift.

    # Why sd1.5 when everyone else uses Pony/XL?

    I'd like to know that too. While XL allows for more details in the prompt, sd is so much faster, and pony has a tendency to just generate trash, and you need to deal with the weird rating system. In comparison, while sd doesn't allow as detailed prompts, all most all the images i generate with the Hassaku sd1.5 model looks better than what i generate with pony models. Hassaku is to me the best model for generating characters, and while it struggles with intercourse, everything else, it does better than any pony model.



    Froot's Twitch:
    https://m.twitch.tv/apricot/home
    Nanoless's Twitch:
    https://m.twitch.tv/nanoless/home


    This model was made using Holostrawberrys lora trainer, but the edited by uYouUs:
    https://colab.research.google.com/github/uYouUs/Hollowstrawberry-kohya-colab/blob/Experiments/Lora_Trainer.ipynb

    Adding Froot's other names to make it easier to search (dunno if that is true).
    Bsapricot, Apricot the lich.

    Description

    Includes Froots (at the time of writing) newest model, the one made by Nanoless.

    Support both of them.

    # About this model

    This is THE best Froot model. This model is quite adaptive to different outfits, and the images of Froot that it generates have a nice color, likely an influence from Nanoless' artstyle.

    RECOMMENDATION, for specific Froot outfits, add weight: (frootvampidol:1.2)

    The main tags for the models are:

    froot, frootrace, froothoodie, frootvampidol, frootelegant, frootmodern, frootcheerleader

    There are some other less used tags. If you are in the superior webui, ComfyUI, you can right click the Lora, and see 'View Lora info...'

    # What has changed?

    The model includes the newest Froot outfit and has nice coloring. The labels of the training images were changed, these new labels don't work too well with all of Froots outfits, especially the hoodie one and the midriff one don't do too well in this model, but maybe you can combine this model with the previous version. This may sound bad, but really, this new version is better. Also, this does better with the vampidol outift.

    # Why sd1.5 when everyone else uses Pony/XL?

    I'd like to know that too. While XL allows for more details in the prompt, sd is so much faster, and pony has a tendency to just generate trash, and you need to deal with the weird rating system. In comparison, while sd doesn't allow as detailed prompts, all most all the images i generate with the Hassaku sd1.5 model looks better than what i generate with pony models. Hassaku is to me the best model for generating characters, and while it struggles with intercourse, everything else, it does better than any pony model.



    Froot's Twitch:
    https://m.twitch.tv/apricot/home
    Nanoless's Twitch:
    https://m.twitch.tv/nanoless/home


    This model was made using Holostrawberrys lora trainer, but the edited by uYouUs:
    https://colab.research.google.com/github/uYouUs/Hollowstrawberry-kohya-colab/blob/Experiments/Lora_Trainer.ipynb

    Adding Froot's other names to make it easier to search (dunno if that is true).
    Bsapricot, Apricot the lich.

    FAQ

    LORA
    SD 1.5

    Details

    Downloads
    113
    Platform
    CivitAI
    Platform Status
    Available
    Created
    10/4/2024
    Updated
    5/13/2026
    Deleted
    -
    Trigger Words:
    Froot
    frootrace
    froothoodie
    frootvampidol
    frootelegant
    frootmodern
    frootcheerleader

    Files

    Frootv2-10.safetensors

    Mirrors

    CivitAI (1 mirrors)

    Available On (1 platform)

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