Porn Factory AILogin

AI porn models for adult image and video generation

Browse the building blocks behind the factory. Filter checkpoints and LoRAs by image or video support, base model, and tags, then open any version directly in the unified generator.

1487 models available

LoRA · SD

anxiang | 暗香

40
This model generates images of classical Chinese characters in Hanfu with plum blossom and landscape backgrounds, aiming for simple elegance. It is designed to work with various mainstream models and benefits from using controlnet for pose and background variations. It also utilizes the sd-dynamic-thresholding plugin for detail enhancement.chinese clotheshanfustyle古风

LoRA · Other

Perky breasts

20
meaning tag >>TAG Perky breasts<< There's no need to tag perky breasts as LoRA will change the look of your character breasts. Not sure if it fits realistic or 3D models, it's train on anime modelsanimeconceptloransfwperky breasts

LoRA · SD 1.4

Breast curtains

30
>>Donate Coffee for Gtonero<< This LoRA has been retrained from 4chananimebreast curtainsconcept

LoRA · SD

concept Nipple tweak

40
This model generates images of breasts, often nude and larger in size. It has two modes triggered by "grabbing from behind" and "pov hands." Negative prompts and weight adjustments can be used to generate clothed figures or modify breast size.actionanimenipples

LoRA · SD

BreakRealize LoRA

10
BreakRealize LoRA adjusts images to a realistic representation. It can also be used in "minus" to create illustrations.realstyle

LoRA · Hunyuan Video

HunyuanVideo POV Missionary

10
HunyuanVideo POV Missionary is a model trained to generate videos of missionary-position intercourse from a point-of-view perspective. Later versions blur faces in the generated content and aim to improve compatibility with character LORAs. The model uses training data, configuration files, and captions to facilitate user training, with example prompts and workflows provided.conceptmissionarysex

LoRA · SD

tutu's HiSilk(8D Black Glossy Over-the-Knee Socks)/ 图图的嗨丝（8D黑色油亮长筒袜）/チュチュのハイシー（8D黒色のオイル光沢長い膝上靴下）

00
This LORA enhances the texture of black, glossy, 8D over-the-knee socks in images. It's made with real-shot materials and differential extraction to minimize impact on other image content, and its weight should be set relatively high (0.5-1.5) for optimal results.clothesclothingfemalegirlspantyhosesexy

LoRA · Wan Video 14B t2v

Holding Deepthroat POV - WAN Video

10
This LoRA model makes subjects look at the camera during deepthroat scenes. It is trained on WAN datasets and can be combined with other blowjob models for enhanced motion. Different noise levels are designed for text-to-video (T2V) and image-to-video (I2V) applications.blowjobconceptdeepthroatfacefuckgagholding

LoRA · SD

AsianMale

20
AsianMale is an initial model designed for Asian male subjects. It exhibits issues with clothing textures, potentially resulting in wrinkles and colored noise at higher Lora weights. The model is currently undergoing retraining to address these problems.asian mancharactermalemanphotorealistic男人

LoRA · Wan Video 14B i2v 480p

The Walk - Wan 2.2/2.1 - I2V & T2V

00
The Walk - Wan 2.2/2.1 I2V & T2V is a LoRa model designed to enhance walking posture in generated videos, making movement smoother and more natural. It offers different versions (Lite, Slow, Normal) with varying degrees of walking quality and potential artifacts, and i2v versions update and improve walking stability and add hand movements. The model aims to address issues like slow motion and unnatural walking, including walking from behind.concept

LoRA · SD

Figma Anime Figures

50
This model generates images of anime figures, specifically in the style of Figma products. It was trained on a dataset of GoodSmile's Figma catalogue images and works best with specific settings like AbyssOrangeMix2 model and certain positive/negative prompts for realistic results. The model offers variations that excel at generalization and capturing the figure-with-transparent-stand aesthetic.action figureanimeconceptfigmafigurines

LoRA · Wan Video

Pixel Art Animation (Birdman Style) I Wan 2.1 I2V LoRa

00
This LoRa model animates pixel art in the style of Birdman, converting images to video. It is trained for consistent animation and can also be used for pixel art stabilization. It's designed for use with pixel art outputs and works best with lower frame rates to maintain the pixel art aesthetic.animationcharacter_conceptconsistentidle animationmemepixel artspritesstyle_

LoRA · SD

Sex Machine

20
This model enhances sex machine imagery with improved detail and performance. Key features focus on mechanical details and depiction of intimate scenes. Avoid using "sex machine" and "vaginal object insertion" prompts if you do not want explicit content.animeconcept

LoRA · SD

Ahegao, rolling eyes(realistic)

110
Ahegao in common sd models tend to be Anime style (and lacks of expression like rolling eyes), so I trained this lora on 20 photos (http://ahegao.casply.com/) as title. (Nice website to recommend. Plz! Keep in low profile.) Suggested weighted 0.5~0.95(it depends on your model & feel free to try). This can go with my merged checkpoint.ahegaorealisticsexual activitysexy

LoRA · SD

TaiwanDollLikeness (v2.0)

30
TaiwanDollLikeness (v2.0) is a model featuring a new variety of faces for image generation, built with a combination of samples from previous versions and a new pool of training images. It is designed to be used with specific settings like VAE, samplers, and LORA weights to achieve optimal results, and can be mixed with other DollLikeness LORAs. The model is open access with a CreativeML OpenRAIL-M license, requiring users to avoid harmful outputs and take responsibility for their creations.asianphotorealisticstylewomen

LoRA · SD

Hairstyles Collection

50
femalegirlsphotorealisticrealisticsexystylewoman

LoRA · SD

Oversized Sweater/Hoodie

20
This LoRA model generates images featuring oversized sweaters and hoodies. It's trained on a small dataset and is easy to use with various colors and checkpoints. Recommended prompt keywords are "oversized_hoodie" or "oversized_sweater" with a LoRA weight of 0.2-0.7.clothingfemalegirlsphotorealisticportraitsrealisticsexywoman

LoRA · SD

breast size slider

50
breast size slider trained on no data using https://github.com/ostris/ai-toolkit (https://github.com/ostris/ai-toolkit)breastsnsfwslidertool

LoRA · SD

墨心 MoXin

230
MoXin is a Lora model trained on traditional Chinese paintings from masters of the Ming and Qing dynasties, designed to generate images in Xieyi (sketching thoughts) and Gongbi (meticulous) styles. It can be used with a companion Lora called "Shukezouma" to create compositions with large areas of negative space, typical of traditional Chinese painting.artstylechinesechinese dresshanfupaintingsstyle

LoRA · SD

娜乌斯嘉nwsj_realistic

60
The 娜乌斯嘉nwsj_realistic model is a virtual human Lora based on the creator's facial features, designed for generating realistic images. It has evolved through multiple versions focusing on improved face clarity, hairstyle flexibility, and works best with Majic series models, with recommended weights varying across versions. Social media handles are provided for interaction.characterphotorealisticwoman

LoRA · Wan Video 14B t2v

Wan 2.1 Facials - Cum on face, Bukkake, Lora, T2V, I2V

20
This model generates videos focused on facial scenarios. It's designed to depict cum on face and bukkake, incorporating T2V and I2V techniques, but results may vary. The model is combined with other Loras, and optimal settings may require experimentation.bukkakeconceptcumcumshotfacialnsfw

LoRA · SD

Shenhe - LoRA Collection of Trauter's

40
This is a collection of anime LoRAs featuring characters from various series. It provides LoRAs for personal use with sample prompts and metadata, focusing on the creator's specific use cases. It contains guides and instructions on how to use LoRAs.2d3danimecharactergenshin impact

LoRA · SD

POV Doggy Anal + Creampie LoRA

50
This LoRA model generates images focused on POV doggy-style anal and creampie scenes. Version 4 includes dataset and tagging improvements for better image quality and prompt responsiveness. Recommended strength is 0.7-0.8, and it is based on UberRealisticPornMerge.actionanalasscreampiecumhardcorephotorealisticpornsexsexual activitywoman

LoRA · SD

concept Headback(arched back)

40
This model generates images with an arched back pose, also known as "headback." It primarily functions from a front view perspective, as side views can distort facial features. It's designed to be used with other models.animeposessexy
Load more models

Showing 24 of 1487

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.