CivArchive
    NSFW Female Nudity - NaturalBeauty FLUX.2 [klein] - v2.0 9B
    NSFW
    Preview 131653273
    Preview 131653198
    Preview 131653910
    Preview 131653113
    Preview 131653464
    Preview 131653727
    Preview 131653777
    Preview 131653845
    Preview 131653884
    Preview 131653955
    Preview 131702160
    Preview 131702378
    Preview 131702462
    Preview 131703836
    Preview 131704285
    Preview 131704381
    Preview 131704502
    Preview 131704650
    Preview 131704787
    Preview 131705181

    NaturalBeauty FLUX.2 Nudity is a photorealistic FLUX.2 [klein] 9B LoRA for creating nude and topless images of beautiful women.

    • To trigger the LoRA, add naked or topless to your prompt. (bottomlessalso works!)

    • You can also add breasts visible, nipples visible, and / or pussy visible to accentuate things further.

    NaturalBeauty enables FLUX.2 [klein] 9B to reliably generate naked and semi-naked photos with correct anatomy, and with a more photorealistic look than the default model alone.

    This LoRA is the result of a two-year project to tag 29k+ images from a diverse dataset. No watermarks, no tattoos; just beautiful images of beautiful women.

    Tagging

    NaturalBeauty FLUX.2 Nudity can do more than just nudity! In addition to naked and topless, try adding any of the following extra details for more control:

    {a caucasian|a black|an asian|a latina|an indian} woman,
    {standing, wide stance|standing, feet together|kneeling|squatting},
    {naked|topless|bottomless|breasts visible|nipples visible},
    {bare feet|high heels},
    {short|medium length|long|very long} {braided|braids|crimped|curly|ponytail|straight|wavy}
    {black|dark brown|medium brown|light brown|chestnut brown|auburn|orange|red|strawberry blond|light blond|medium blond|bleached blond|platinum blond} hair,
    {neutral expression|slight smile|full smile},
    {closed|slightly open|open} mouth,
    {closed|open} eyes,
    {looking at camera|looking away},
    {facing camera|facing away|facing left|facing right}
    {tiny|small|medium|large|huge|gigantic} {perfect|pendulous} {fake|natural} breasts,
    {shaved pussy|landing strip|trimmed pussy|hairy pussy},
    {arms by side|hands behind head|hands in hair|hands cupping breasts|hands on thighs|hands on hips}

    For example:

    a caucasian woman, standing, wide stance, naked, long wavy medium blond hair, slight smile, closed mouth, open eyes, looking at camera, facing camera, huge perfect natural breasts, shaved pussy, hands behind head

    NaturalBeauty FLUX.2 Nudity is optimized to produce good results without needing a negative prompt.

    NaturalBeauty FLUX.2 Nudity is trained to generate portrait images with a size of 1024x1536 pixels. Other 2:3 aspect ratio sizes should also work.

    All of the example images were generated in ComfyUI with the following settings:

    • Sampler: Heun

    • Steps: 4

    • CFG: 1.0

    • LoRA model strength: 1.0

    All of the example images for this model have a ComfyUI workflow and instructions embedded inside them. Just drag one of the sample PNG images into ComfyUI and generate!

    Training settings

    • 29,014 hand-tagged images (naked and clothed, no tattoos or watermarks), plus flips

    • Extensive tagging for ethnicity, breasts, nudity, pubic hair, clothing, leg pose, arm pose, hair, eyes, mouth, jewelry, piercing

    • 1x H200

    • Constant learning rate of 1e-5 over 18 epochs at full 1024x1536 size with batch size 1

    • 64-dim bf16 LoRA export from fine-tuned base model

    • Trained with musubi-trainer

    Notes

    This LoRA is a follow-on to a FLUX.1 LoRA I released back in December 2024, named NaturalBeauty FLUX Nudity. There's a Qwen-Image-2512 LoRA too!

    Description

    • Much-improved breast, nipple, and genital realism over v1

    • Now supports prompting for pubic hair (shaved pussy, landing strip, trimmed pussy, hairy pussy)

    • Now supports body direction prompts (facing camera, facing away, facing left, facing right)

    • Larger and more varied training data set

    • Much longer training time than v1 for improved prompt following

    FAQ