Back to LoRA Models — Free AI Art Models for Flux, Illustrious & More

Reclining Nude, After Sex [Flux, Illu, Qwen, Z-Image]

Reclining Nude, After Sex [Flux, Illu, Qwen, Z-Image]

## Description

v2 beta 1 out now. Please try it out and post your images in the gallery, thoughts in the comments.

This is a Flux LoRA for detailed female anatomy in a variety of styles, including realistic photography, digital art, anime, and cartoon. Breasts and vaginas can be shown partially covered or fully exposed, and with or without cum and sweat. It can do close-up details of the vagina as well. Making them completely dry may be difficult, but that is not really the point.

It was originally focused on reclining and standing female nudes, but can be used for many different poses and amounts of clothing. Figures can be shown reclining on their backs or standing. Standing figures can be shown from the front or from behind. Side boobs and nipples usually work right.

This LoRA is much more detailed and more flexible compared to the older https://civitai.com/models/660022/fluxxxible-sexual-nudes-for-flux, with a lot more detail around the vagina and anus.

### v2 changes

  • better hands
  • better prompt adherence
  • support for a greater variety of styles
  • better compatibility with other LoRAs

### Poses

  • reclining, on back
  • standing, from front
  • standing, from behind
  • vagina, close up

### Concepts

  • Cum, sweat, and dripping fluids can be controlled
  • Breast size and shape can be controlled
  • Nipples can be modified, although not always reliable
  • Vaginal opening and spreading can be controlled

## Download

### Reclining Nude v2 Beta 1

Download reclining-nude-v2-beta-1.safetensors (302.38 MB)

SHA256: b27518020ffe0f084c57a028822affc213b56a4582b2307130f2472dfcc0a719

Model
Base ModelFlux1Training Precision
LicenseNon-Commercial LicenseBase LicenseFlux1.Dev License
Network Type
NetworkLoraNetwork Precisionfp16
Training Parameters
Mergedtrue
Dimension8Alpha8

### Reclining Nude v2 Alpha 1

Download reclining-nude-v2-alpha-1.safetensors (584.36 MB)

SHA256: ee50e785707b5526a02315279fba8047bb8c6ed30190473e53b6c23df8e2eb22

Model
Base ModelFlux1Training Precision
LicenseNon-Commercial LicenseBase LicenseFlux1.Dev License
Network Type
NetworkLoraNetwork Precisionfloat
Training Parameters
Mergedtrue
Dimension8Alpha8

### Reclining Nude v1

Download reclining-nude-v1.safetensors (73.18 MB)

SHA256: 7cf5626bae523655086f30a437861b3f24221c156f80cf55134b461a8f9d7b5a

Model
Base ModelFlux1Training Precisionfp8
LicenseNon-Commercial LicenseBase LicenseFlux1.Dev License
Network Type
NetworkLoraNetwork Precisionbf16
Training Parameters
OptimizerProdigySteps2100
D Coef2Training Images251
Dimension8Alpha8
Min SNR GammaNoneWeight Decay0.1

### Illustrious v01

Download reclining-nude-v01-illu.safetensors (94.15 MB)

SHA256: a3fcdc723b54a6ba76ed215eb4677e896d9969430b74d18cbce420d622820203

Model
Base ModelIllustriousTraining Precisionbf16
LicenseNon-Commercial License
Network Type
NetworkLoraNetwork Precisionbf16
Training Parameters
Steps2000
DimensionAlpha
Min SNR GammaNoneWeight DecayNone

### Qwen v01

Download reclining-nude-v01-qwen.safetensors (281.47 MB)

SHA256: ad4999fcb1f574806cc39329771eb224eaa7ba08abcb435d3f92af50691189fa

Model
Base ModelQwenTraining Precisionbf16
LicenseNon-Commercial License
Network Type
NetworkLoraNetwork Precisionbf16
Training Parameters
Steps3500
DimensionAlpha
Min SNR GammaNoneWeight DecayNone

### Flux v02

Download reclining-nude-v02-flux.safetensors (302.38 MB)

SHA256: b27518020ffe0f084c57a028822affc213b56a4582b2307130f2472dfcc0a719

Model
Base ModelFlux1Training Precisionfp8
LicenseNon-Commercial LicenseBase LicenseFlux1.Dev License
Network Type
NetworkLoraNetwork Precisionfp16
Training Parameters
Dimension8Alpha8
Min SNR GammaNoneWeight DecayNone

### Flux v01

Download reclining-nude-v01-flux.safetensors (73.76 MB)

SHA256: 75de3fbf98544e89f13e6ded55b03e8d9e2f44ac22990d5ba380df5d25e2c5d2

Model
Base ModelFlux1Training Precisionfp8
LicenseNon-Commercial LicenseBase LicenseFlux1.Dev License
Network Type
NetworkLoraNetwork Precisionbf16
Training Parameters
OptimizerProdigySteps1470
D Coef2Training Images294
Dimension8Alpha8
Min SNR GammaNoneWeight Decay0.1
  • Model and CLIP strength from X - Y, depending on how much of an effect you want
  • Guidance from X - Y

## Example Prompts

The general format is picture of [pose] woman. [breasts], [vagina] [cum], [background].

Some keywords were translated into l33t versions, the same as my other Flux LoRAs. I have not been able to prove if this is necessary or not, but it has been working well for me.

  • picture of reclining sulty woman. medium br34sts in black corset, pretty pink v4g1n4 and 4nus thick black cum
  • picture of reclining beautiful princess. large br34sts, fingering her spread v4g1n4 with open v4g1n4l c4n4l and tight 4nus
  • picture of standing s3xy witch. small perky br34sts, tight pink v4g1n4 dripping cum
  • picture of standing redhead barbarian queen. small perky br34sts, pretty pink v4g1n4 dripping cum
  • picture of brunette barmaid’s 4ss, from behind, 4ss shot. medium br34sts, thin thong panties
  • picture of raven-haired beauty’s 4ss, from behind, 4ss shot. large br34sts with perky n1ppl3s, barely visible v4g1n4
  • a close-up picture of pink v4g1n4 in crotchless panties soaked in white cum. faded background
  • a close-up picture of gaping pink v4g1n4 slightly moist. faded background

Some keywords were translated into l33t versions, the same as my other Flux LoRAs. Those words include:

  • penis: p3n1s
  • vagina: v4g1n4
  • vaginal sex: v4g1n4l s3x
  • nipple: n1ppl3
  • sex toy: d1ld0
  • sex: s3x
  • oral: 0r4l
  • oral sex:0r4l s3x
  • breast: br34st
  • clitoral: cl1t0r4l
  • clitoris: cl1t0r1s
  • anus: 4nus
  • anal: 4n4l
  • anal sex: 4n4l s3x
  • ass: 4ss
  • buttocks: 4ss
  • butt: 4ss
  • dildo: d1ld0
  • naked: n4k3d
  • nude: nud3

## Dataset

Trained on about 250 images that I generated in the desired poses.

Images with bad hands, multiple belly buttons, or watermarks were removed.

v1 captions were written by a human, which does not seem to be as effective as VLM captioning (I recommend Joy or Florence 2).

v2 captions used Joy captioning.