CivArchive
    Preview 38685724
    Preview 38685743
    Preview 38685749
    Preview 38685759
    Preview 38685790
    Preview 38685845
    Preview 38685855
    Preview 38685866
    Preview 38685868
    Preview 38685874
    Preview 39067457
    Preview 39067528
    Preview 39067588
    Preview 39067614

    Naturally fat gals are XL! (including SDXL, PonyXL, IllustriousXL and NoobAI)

    LoRa Strength range: 0.8 - 1.1 (the more - the bigger)

    This is a model, which style was inspired by "Obese Girls | Concept" by LeonTalon.

    I really missed using that model while working with SDXL so I decided to remake it myself.

    As with the OG model this one has three tags to work with.

    WeightChubby

    For all your slightly chubby girls needs.

    Recommended prompts: plump, chubby, fat is the supposed ceiling of this trigger tag. You can also decrease the LoRa's strength for lesser weight of characters.

    WeightObese

    Entering the heavy weights category.

    Recommended prompts: fat, obese, bloated belly, big belly. From chubby gals up to sizable bellies. Recommended LoRa strength: 1.0. You can also raise the strength of the trigger tag and auxiliary prompts to 1.1 if you need to get a bigger size within this tag's range.

    WeightImmobile

    A whole room worth of snacks!

    Recommended prompts: hyper fat, hyper obese, huge bloated belly, immobile, ssbbw... you name it. Recommended LoRa strength: 1.0-1.1. You can also raise the strength of the trigger tag and auxiliary prompts to 1.2 if you need to get THE BIGGEST GALS THIS LORA CAN PROVIDE.

    Description

    50 likes on v1 special! Not specifically better, just different.

    More training data, higher quality training data and here we are!

    Wasted all my dearest free buzz on this train which is a bit disappointing, but I hope you like this version as much as the first one.

    Overall noticable changes: less gentle and realistich style push, eyes are now EDGY, bellies look less bloated depending on size, sadly less quality effect on hair than v1 as this version doesn't have any SD 1.5 output in its dataset at all

    FAQ

    LORA
    Pony

    Details

    Downloads
    554
    Platform
    CivitAI
    Platform Status
    Deleted
    Created
    11/6/2024
    Updated
    4/22/2026
    Deleted
    4/16/2026
    Trigger Words:
    WeightChubby
    WeightObese
    WeightImmobile

    Files

    ReFat v1.1.safetensors

    Mirrors

    Huggingface (1 mirrors)
    CivitAI (1 mirrors)
    Other Platforms (TensorArt, SeaArt, etc.) (1 mirrors)

    Available On (1 platform)

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