Porn Factory AILogin

realistic models for adult image and video generation

Compare 137 public adult AI checkpoints and LoRAs tagged realistic. Review each model's versions, base-model family, trigger words, previews, and validated image or video workflows before you create.

137 models available

LoRA · SDXL

beautiful girl for SDXL

081
This LoRA generates images of beautiful girls, best used with a strength less than 1. The trigger word is "majicmix style," and the model is recommended for use with the realisticStockPhoto_v10 checkpoint. A V2 model based on feedback includes smiling images.girlsrealisticstylewoman

LoRA · Flux.1 D

Amateur Photography

280
This model provides settings for amateur photography image generation, specifically optimized for the v6 Flux Dev version. Key settings include a distilled CFG scale of 3.5, DEIS sampling with DDIM schedule, 20 steps, and a resolution of 896x1152, including a hires fix model with specific parameters. The recommended Lora weight is 0.8, but experimentation is advised.amateur photographyphotorealismphotorealisticrealisticstyle

LoRA · SDXL

Realistic Insta Influencers By Stable Yogi

078
This model creates realistic influencer images using SDXL+Pony-based LoRAs, trained with one face and one body image. It generates fictional characters through AI, and any resemblance to real persons is coincidental.charactergirlsinstainstagram babeinstagram starloraphotorealisticrealisticrealisticphotosexywaifuwoman

LoRA · SD

Droptop

273
Droptop is a LoRA model designed to create images of people with their tops dropped. It gives realistic results, but can also work with anime style models. Recommended weight range is 0.4-0.7.breastsflashingphotographyphotorealisticpornposesrealisticsexywoman

LoRA · SD

Yuri kiss

172
not convinced with controlnet for kissing, so I'm testing with loraskissnsfwrealistic

LoRA · IL

0__11Xx | Shiiro's Styles | Niji

265
The model, 0__11Xx | Shiiro's Styles | Niji, generates images with surreal and glitch themes, using the trigger word "0__11Xx_illu". Effective sampling can be achieved with DPM++ 2M Karras (20-40 steps) or Euler a (25-40 steps), CFG Scale 1-7, and Clip Skip 2. Usable weight range is from 0.1 to 1.abstract paintinganimeartart styleartistartstylebackground platebackground ponybackgroundscartooncolordetailflowersgame charactergeometrygirlsloramalemanmangamanga artistnsfwpaintphotorealisticrealisticshapesshiirostylestyleswoman

LoRA · SD

【Realistic & Anime】Genshin Impact AI Cosplayers Collection 原神 AI Coser合集

365
This model generates realistic and anime-style images of Genshin Impact cosplayers. It is recommended to use adetailer for face fixing. The model creator also suggests a hand-drawn LORA for anime checkpoint users.animebabaraclothingeulafaruzanfischlgenshingenshin impactgenshinimpactgirlshutaojeankamisato ayakakeqingkirara (genshin impact)luminelynettenilouningguangningguang (genshin impact)photorealisticraiden shogunrealisticsangonomiya kokomishenheyae mikoyoimiya

LoRA · Pony

BDSM Cocksleeve - Undercarriage - Monsters (Pony & Illustrious)

062
Ties character to another character to be used as cocksleeve/onahole. For characters tied to centaurs or other 4 legged monsters, Landscape aspect ratio tends to work much better. Check examples for prompts.animebdsmbondageconceptrealisticwoman

LoRA · SD

[Free] Breast Helper 1.2

558
This model helps generate images of breasts with specific attributes. It allows users to control breast size, shape, nipple color and style, and areola size using trigger words. The model includes fixes for face issues and improved stability for animation generation.big breastsbimbobreastsconcepthelpermedium breastsrealisticsmall breasttits

LoRA · IL

Niji goth style | Goofy Ai

057
This model generates images in a Niji goth style. It is recommended to use a LoRA at 0.7-1 and ADetailer for faces, as well as img2img upscale with 4x-ultra sharp.animeconceptgothmechanijiniji journeyrealisticsemi-realisticwoman

LoRA · SD

Underboob CropTop by Stable Yogi

052
The model generates images of characters wearing various underboob-revealing tops. It is triggered using keywords such as "Underboob CropTop," "Underboob t shirt," "Underboob bra," and "Underboob bikini," with a recommended weight of 0.8 to 1.0.actressanimeclothingfemalegirlslorapersonphotorealisticrealisticsexywoman

LoRA · IL

厚塗り / ATNR STYLE

146
The 厚塗り / ATNR STYLE model is for online image generation. It is recommended to use the DPM++ 2M SDE sampling method with SGM Uniform schedule type and a CFG Scale of 6-4, with a weight of 0.9-0.8.animefemalegirlsrealisticstyle

LoRA · Pony

Transparent Wedding Dress / Kinky Wedding Dress +18 - Concept / Clothes - | Illustrious XL | PONY XL | - COMMISSION - by YeiyeiArt

242
This model generates images of transparent or kinky wedding dresses for Illustrious XL and PONY XL. The trigger word is "WeddingDressXL", and adding terms like "bridal veil, see-through, naked body" can enhance the output. It is designed for creating explicit content.animeart stylescartoonclothesclothingdressfantasyhentailoramangarealismrealisticsexystyles

LoRA · SDXL

SDXL YAMER'S STYLE 😍 Princeps Omnia LoRA

041
SDXL YAMER'S STYLE 😍 Princeps Omnia LoRA is a model designed to reinforce the creator's artistic style or inject it into other checkpoints. It excels at generating colorful, chaotic, and aesthetically unique images, especially with abstract works and singular character images. The LoRA was trained using images in various styles and resolutions, but 576x1024 is recommended.2d3dactressalphaanimeanime irlartbeautifulbeautiful girlboyscutedarkfantasyfantasy artfemalegamegame charactergirlpackgirlsillustrationillustrationskawaiiloramanmemeneworiginalitypersonphotorealisticportraitsquality uprealisticrealisticlorasdxlsdxl 1,0sdxl animesdxl style lorasfwsharpstylestyle lorastyle mixstylemixstylesterrorvideo gamewomanxlxl1.0yamer

LoRA · SD

Crop Shirt + Underboob (anime+realistic) | Goofy Ai

340
This model generates images of characters, both anime and realistic, wearing crop shirts with underboob exposure. It works well with a Lora weight of 0.7-1 and benefits from a highres fix.animeclothclothingfemalegirlsoutfitrealisticsexywoman

LoRA · Pony

Retro Game (CPS II) - Pixel Art - STYLE - | Illustrious XL | Pony XL | - By YeiyeiArt

140
This model generates images in the style of retro CPS II video game pixel art. It offers both a standard and a Pony XL version, with the latter providing improved prompt understanding and adjustable intensity. Key prompts include "Csp2Pixel" or "CspPixelArt_Style" and phrases like "fighting stance" or "dynamic pose" for creating Street Fighter-like images.animecartoonfantasygirlslorapixelartrealismrealisticretrostylevideo gamevideo game characterwoman

LoRA · IL

Ski Slope Breasts | Flux + Pony + ILL + SDXL

039
This model generates images of breasts with varying slopes and nipple puffiness, aiming to recreate the "Swooping Breasts" aesthetic. It offers multiple versions, including mixed, steep, and shallow slopes, as well as an XXL version for larger breasts. The SDXL version is best used with lower strength settings to avoid cropping the head.breastsconceptrealisticsexytitswoman

LoRA · IL

Ka Marukogedago | Style LoRa

038
Ka Marukogedago is a style LoRa model that generates images with abstract, watercolor, or oil painting styles. The trigger word is k4_illu, and useful tags include abstract, watercolor, traditional media, scan, and blending. It is recommended to use a weight between 0.1 and 1, with samplers DPM++ 2M Karras or Euler a, CFG Scale between 1 and 7, and Clip Skip of 2.abstract paintinganimeartart styleartistartstylebackground platebackground ponybackgroundscartooncolordetailflowersgame charactergeometrygirlsloramalemanmangamanga artistnsfwpaintphotorealisticrealisticshapesshiirostylestyleswoman

LoRA · IL

Hooters Uniform - Concept / Clothes - | Illustrious XL | - Waifus At Work! by YeiyeiArt

037
This model generates images of characters wearing a Hooters uniform. Use the trigger word "HootersUniform" along with prompts like "employee uniform, short shorts, orange shorts, tank top, white shirt, cleavage" to achieve the desired look.animeart stylescartoonclothesclothingfantasyloramangarealismrealisticsexystylesuniform

LoRA · SD

Cyberpunk Style

037
Cyberpunk Style is a model trained using the same data as SDXL. It was trained on the CivitAI trainer.animatedanimecyberpunkfemalefuturisticgame charactergirlsloramanpersonphotorealisticportraitspost apocalypticrealisticsexysexy attiresexy menstylestylesvideo gamewoman

LoRA · SD

RealBodiez

030
This LoRa is trained for imperfect mom bods, with muffintops, bulges, saggy boobs and stretchmarks. You know...real bodies.areaolafemalemuffintoprealisticsaggystretchmarksstyle

LoRA · Flux.1 D

Realistic/Aesthetic Poses FluxDev

112
This LoRA model generates realistic and aesthetic poses, suitable for vintage or social media styles. It was trained with a limited number of steps and aims to provide poses not readily available elsewhere.aestheticaesthetic posedramatic lightdynamic poseflux.1flux1.dnightphotorealisticpinterestposesrealistic

LoRA · Flux.1 D

Realistic Amplifier for UltraReal Fine-Tune

011
Not much to say about this LoRA - at strength 0.9-1.0, it gives an effect similar to old Nokia phone photos. At strength 0.5-0.7, it simply amplifies the realism of my checkpoint. I didn’t plan to upload this for a while (thought no one would need it), but after getting some PMs asking about it, I decided to share. I use this exclusively with my own checkpoint. For flux.dev it just serves as a minor realism enhancement.realisticstyle

LoRA · SD

Elegant Dresses Collection

29
This model generates images of elegant dresses. It is a Lycoris model, best used with specific sampler and CFG scale settings, and a Lycoris weight of 0.5 for optimal results.clothingfemalegirlsphotorealisticrealisticsexywoman
Load more models

Showing 24 of 137

How checkpoints and LoRAs fit together

Start with a checkpoint

A checkpoint controls the base rendering system. Its supported modes determine whether it can create images, edit an input, or produce video.

Match the base model

A LoRA must target the same base-model family as the checkpoint. Similar names do not guarantee compatible weights.

Check the validated workflows

A catalog version can be documented before it is generator-ready. Only versions with validated broker and artifact mappings can spend credits.