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

Kon-kit Style LoRA

115
The Kon-kit Style LoRA generates images in the style of the artist Kon-kit, known for drawing seductive girls. It is best used with a LoRA weight of 0.6-0.7 and the trigger word "kon-kit". The dataset contains mostly NSFW concepts, but the LoRA itself is a drawing style and not directly NSFW.art styleartstylebreastsdoujinfemalegirlskon-kitmangasexystylewoman

LoRA · IL

Spooning - Concept

015
The model generates images based on the concept of "spooning." It is triggered by keywords related to intimacy and sexual acts like cuddling, lying, and various sexual positions between a heterosexual couple. The images often include nudity, blushing, and motion effects, taking place in a bedroom setting.concept

LoRA · SD

Chain leash

215
This model generates images containing chain leashes, often with a viewer holding the leash and a collar present. Use trigger words like "chain leash," "viewer holding leash," and "collar," but be aware the results may require prompt tweaking due to potential inaccuracies.animebdsmboundchainchain leashclothingfetishhentaileashpet play

LoRA · IL

NTR happening behind the door/ドア越しに寝取られる/门后NTR

115
This model aims to generate images of NTR scenarios, specifically focusing on depicting what's happening behind a door. It attempts to render a man semi-transparently inside the room to visualize the scene, though achieving consistent and ideal results may require experimentation with prompts like "high contrast" or "door crack." The model may not work optimally with all checkpoints, requiring users to switch checkpoints if needed.behind the doorconceptdoorjapanesenetorarentrwoman

LoRA · IL

Doggy style Variantions: IL/NoobAI | Shrekman Hentai Loras

115
The model offers doggy style variations under the names IL/NoobAI and Shrekman Hentai Loras. It appears to be available on Tensor Art, CivitAI, and Seeart. The model creator requests support via Patreon and commissions.doggyhead pinhentaipornposesrough sexsex

LoRA · Wan Video 14B i2v 480p

Wan Thiccum

114
Wan Thiccum is a LoRA model designed to generate thicker and slimier cum in images, specifically optimized for close-up facial shots. It works best when the input image already displays these properties and it's trained on phrases related to cum shots.conceptcum

LoRA · SD

Millie (Helluva Boss)

214
Character for generating Millie from Helluva Boss animated show. Before training I set "Millie" as an activation word, but for some reason lora generates her only when I use "demon girl" in promptcharacterdemon girlhelluva bosslora

LoRA · Pony

Micro Panties | Goofy Ai

114
Micro Panties is a model hosted on Tensor.art. Prompts and usage details are available, including recommended Lora strength, ADetailer for faces, and upscaling techniques.animeclothinggirlsloramicro pantiespantieswoman

LoRA · Wan Video 2.2 I2V-A14B

WAN 2.2 I2V - POV Cowgirl

95814
WAN 2.2 I2V POV Cowgirl is a video lora model designed to generate POV cowgirl content. It aims to preserve movement and potentially benefit from a general NSFW lora for enhanced results. The model is undergoing iterative improvements to refine hip movement.actioncowgirlgirl on topstraddling

LoRA · SD

Tangbohu Line Tweaker LoRA (唐伯虎勾线调整LoRA)

114
This LoRA controls line markings in images while preserving the original composition. Positive weights increase line thickness, while negative weights decrease or remove lines. Experimentation with weights beyond the suggested range of -2 to 2 may be needed.stylestyle lora

LoRA · IL

Hug & Suck LoRa | Illustrious

114
This LoRa model generates anime-style images depicting fellatio/deepthroat scenes with hugging. It is trained for use with Illustrious models and activated using trigger words within a Stable Diffusion setup. The model is intended for NSFW content and should be used responsibly.conceptdeepthroatfellatiohugsuck

LoRA · SDXL

SDXL - Let It Flourish

014
Needed a little counterpart to my last model. I manly use my or other peoples LoRA's to enhance something. This one here is trained a bit differently and enhances details, colors and may introduce weird geometry and cluttered shapes. • Would suggest a low strength because not every model can handle it without making everything a bit messy. • Will bring a bit of hope and happiness to even dark themes.atmospherecolorfulflowerslightmoodysoothingstyle

LoRA · IL

Standing Breast Press Handjob - Concept

014
This model generates images of a standing figure performing a breast press handjob. Key features include depictions of sexual acts and suggestive details like cleavage, nipples, and blushing. It is triggered by specific keywords related to the pose and elements of the scene.concept

LoRA · Pony

Standing Groping Sex - Concept

014
The model generates images of standing sex, including groping and grabbing. Key features include triggering words related to sex, nudity, and specific body parts. It uses a LORA to influence the final image.concept

LoRA · Pony

(ilxl+Pony) Togawagatame / 戸川固め

014
This model generates images of a "togawagatame" pose. It depicts one character pinning another using their legs, specifically pressing their thigh against the opponent's to force their legs apart. Use the trigger word "togawagatame" with a weight of 1.conceptnsfw

LoRA · Wan Video 14B i2v 480p

Wan Full Nelson (I2V)

014
Wan Full Nelson (I2V) is a LoRA model trained to generate videos of the full nelson sex position. It supports both anal and vaginal penetration and works well with various camera angles. The model was trained on 28 videos using a 480p I2V model.conceptfull nelsonwan

LoRA · SD

Rosalina (Mario)

014
This model generates images of Rosalina from the Mario franchise, trained on the animefull model. It includes several of her alternate outfits, accessible via specific tags, and allows for modification of her dress color. The model works well with derivative merges from the animefull model.animecharactergame charactersuper mario

LoRA · IL

真下アングル/close-up from below

014
The recommended strength of lora is around 0.8. If you like, pleasehttps://ko-fi.com/bolero537poses

LoRA · Pony

restraind missionary sex (illustrious, pony)

114
lora for a missionary sex wi a girl restrained from behind.conceptgroup sexmissionaryrestrainedsex

LoRA · SD

Kidnapped / bdsm / willing partner

314
This LoRA model generates images of characters in kidnapping scenarios, specifically tied up and gagged. It can be used with various tags relating to bondage, gags, restraints, poses, and placement to create diverse images. The model does not use a trigger word for greater flexibility.animebdsmbondageboundclothingropeshibaritape gaggedtied up

Checkpoint · SD

ChilloutMix

8014
ChilloutMix is a merge model designed for generating 2.5D illustrations and realistic images, particularly of Asian subjects. It is recommended to use LORAs or Embeddings to enhance face variation. Commercial use is strictly prohibited and redistribution requires adhering to the modified Dreamlike License.base modeldreamlikeartphotorealisticpornsex

LoRA · SDXL

Polygon Style [LoRA 1.5+SDXL]

014
Trigger: ral-polygon ☕ Buy me a coffee: https://ko-fi.com/ralfingerai (https://ko-fi.com/ralfingerai) 🍺 Join my discord: https://discord.gg/g5Pb8qNUuP (https://discord.gg/g5Pb8qNUuP)polygonpolygonartpolygonsstylestylestechtrianglevector

LoRA · NAI

仰け反り/arched back orgasms

113
A strength of around 1.0 for Lora is recommended.concept

LoRA · Wan Video 2.2 I2V-A14B

Anal insertion (I2V)

67013
Helps to avoid melting effect and helps inserting it into backdoor. wan 2.2 test , trained on both realistic and animated clips about 60+40 % Prompt format : Anal insertion. The video depicts a man slowly inserting his dick into a woman's anus while she is lying on a couch Use last frame only with no start frame for different effect. Add other loras for more dynamic "transition".concept
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.