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

【Concept】Dogeza 概念-土下座

50
This model generates images of characters in the "dogeza" (kneeling apology) pose, now with improved clarity and the ability to render faces, multiple characters, and rear angles. Use lower weights for faces (0.6-1.0) and higher weights for multiple characters or rear views, though image consistency and compatibility with other LoRAs may be unstable.animeconcept

LoRA · SD

Asian girls face

50
This model generates images of Asian girls. It is recommended to use specific settings like VAE, Sampler, Steps and Lora weights for optimal results. The model's creator is not the original author, but is sharing it due to user demand.asianfacegirlsphotorealisticstylewoman

LoRA · SD

KoreanDollLikeness (v2.0)

130
KoreanDollLikeness (v2.0) is a LORA model offering a variety of faces for generating doll-like characters. It is recommended to use specific settings and samplers for optimal quality and can be mixed with other DollLikeness LORAs for interesting effects. The model is open access under the CreativeML OpenRAIL-M license, with restrictions against harmful outputs and a call for responsible use.asiankoreanphotorealisticstylewomen

LoRA · SD

Gacha splash LORA

130
This LORA creates images in a gacha splash style, trained on a dataset generated by NovelAI. It works best with prompts including isometric or hexagon perspectives and may require negative prompts to refine details and avoid unwanted elements like wood or stone. Two versions are available: one suited for abstract backgrounds and another optimized for architectural structures.animebackground

LoRA · SDXL

Harrlogos XL - Finally, custom text generation in SD!

20
Harrlogos XL is an SDXL LoRA that enables custom text generation without requiring control nets or image-to-image techniques. It allows users to generate text logos with various styles, colors, and modifiers by specifying terms in a specific order within the prompt. The model works well with other checkpoints and LoRAs.band logoemogothharrowedlogometalstyletext

LoRA · Wan Video 2.2 I2V-A14B

Wan Multiscene Photoshoot: Softcore Edition

5540
This Lora creates longform videos in the style of a fast-cut photoshoot with multiple angles. It requires a specific workflow for optimal results, including generating an initial image and processing in two phases: low-resolution for seed finding, followed by high-resolution denoising. The Lora is trained at 2x speed, requiring interpolation, and the quality depends on image inputs.concept

LoRA · Wan Video 2.2 I2V-A14B

Pornmaster_wan 2.2_14b_I2V_ double penetration

10
下載前請仔細閱讀使用說明，否則將無法正確生成影片。 Please read the instructions carefully before downloading, otherwise the video will not be generated correctly. 通義萬相AI生影片—使用指南-中文 培訓使用的字幕下載. Download the subtitles used in the training. -- PornMaster-wan 2.2_14b_I2V_69 position blowjob -- 使用最新的加速器，可以讓生成的影片更穩定。 Using the latest accelerator can make the generated video more stable. wan2.2_t2v_A14b_high_noise_lora_rank64_lightx2v_4step_1217 wan2.2_t2v_A14b_low_noise_lora_rank64_lightx2v_4step_1217 -- 與” WAN DR34ML4Y - All-In-One NSFW ”一起使用時解剖學可以獲得更加穩定的效果。 如果人物的臉被改變時，就降低lora權重。 When used with " WAN DR34ML4Y - All-In-One NSFW ", the anatomy can achieve a more stable effect. If a character's face is altered, reduce the lora weight. -- bukkake + double penetration: -- double penetration: -- 低噪聲和高躁聲lora必須一起使用。 範例影片使用wan 2.2官方標準工作流生成。 我的影片使用 wan2.2_i2v_14b_fp8_scaled 生成。 主要用於臉部特寫照片，也可以使用已經進行雙插做愛的姿勢的全身照片。 Lora權重使用0.9~1。 需簡單的描述做愛場所和背景，例如:他們在草地上性交，背景是森林。 -- Low-noise and high-noise LoRa must be used together. The example video was generated using the official WAN 2.2 standard workflow. My video was generated using "wan2.2_i2v_14b_fp8_scaled". Primarily used for close-up facial photos, but full-body photos of double penetration positions can also be used. Lora uses a weight of 0.9~1. Please provide a brief description of the location and background of the sexual activity, for example: they had sex on a grassy field with a forest in the background. -- 觸發詞(Trigger words) 從左邊進行做愛(Having sex from the left side): a woman, the camera quickly pulls back to reveal her vagina, her breasts are bouncing and her ass are bouncing, she was sitting and leaning back against a man, the man beneath her thrusting his penis in and out of her anus, another man on the left was thrusting his penis into and out of her pussy, their testicles continued to sway, her legs were spread apart, and her breasts could be seen spread apart, the man's buttocks were visible beneath her, -- 從右邊進行做愛(Having sex from the right side): a woman, the camera quickly pulls back to reveal her vagina, her breasts are bouncing and her ass are bouncing, she was sitting and leaning back against a man, the man beneath her thrusting his penis in and out of her anus, another man on the rigth was thrusting his penis into and out of her pussy, their testicles continued to sway, her legs were spread apart, and her breasts could be seen spread apart, the man's buttocks were visible beneath her, -- 我的其他檢查站和Lora。 My other checkpoints and Lora. -- 客製化lora:我的Telegram ID:@iamddtla0620 我的微信ID:iamddtla 我的line ID:iamddtla 時間寶貴，請直接說明來意，例如:訂製lora. Customized lora: My Telegram ID: @iamddtla0620 My WeChat ID: iamddtla My Line ID: iamddtla Time is precious, please state your intention directly, for example: customized LORA. 請勿將此lora用於違法用途或是真實存在的人物。 Please do not use this lora for illegal purposes or for real people.

LoRA · Wan Video 14B i2v 720p

Breast Squeeze and Lactation (Milk spray, milking) WAN 2.1 i2v/t2v model

20
This model generates images and videos focused on breast squeezing and lactation. It produces outputs depicting milk spray and the milking process, using keywords such as "squeeze," "nipple," and "jet of milk." The model is designed for NSFW content creation.conceptnsfwnudity

LoRA · SD

Hu Tao | Genshin Impact

30
This is a LoRA model for the character Hu Tao from Genshin Impact. It was trained primarily on the NovelAI model, but can also work well with Anything-V4 or AOM2, and typically requires a weight of 0.6-0.8. The tag "boo tao" can be used to generate images with her ghost.animecartooncharacterfantasyfemalegame charactergenshin impactgirlsillustrationpersonportraitsvideo gamewoman

LoRA · SD

武侠 Wuxia V2

20
Wuxia V2 is a LORA model that generates martial arts style images with simple and neat compositions. This version features improved details, backgrounds, and more Chinese-style clothing, and is triggered by the words "weapon" and "sword".style

LoRA · SD

Better eyes+face+skin | 更好的眼睛+脸+皮肤

30
This model enhances facial features, particularly eye detail across the entire face, and adds skin detail. It can be mixed with other LoRAs for improved facial results; use a flexible weight. Be aware of potential facial collapse at high magnifications (v2).beautifulbeautiful detailed eyesconceptdetaildetailed facedetailerenhancereyeeyesfacehandsomeinpaintinpaintingphotorealisticrealistic

LoRA · Wan Video 14B i2v 480p

M4CROM4STI4 - Huge Natural Breasts Physics - Hun | Wan2.1 Lora

10
This LoRA model, M4CROM4STI4, is designed to create images and videos with large, natural-looking breasts. It was trained on a dataset of video clips and is intended to be used with specific workflows for image and video generation. The model is shared with the condition that it is used responsibly and lawfully.style

LoRA · Wan Video 2.2 I2V-A14B

Breast Insertion [WAN 2.2]

1.5K0
This LoRA model inserts a man's penis between a woman's breasts. It was trained on video clips where the man is not initially visible. The trigger phrase is "A man appears and inserts his penis between her breasts."actionbreastbreastfuckbreastfuckinginsertionpaizurititfucktitfuckingtittyfucktittyfucking

LoRA · SD

Anime Tarot Card Art Style LoRA (塔罗牌/タロットカード)

130
This LoRA generates anime tarot card art in the style of Phantom Star illustrations. The model is updated for improved consistency and supports varied backgrounds using a specific prompt format. It's recommended to use with Anything V4.5 and orange-mix VAE, at a strength of 0.8-1.animeartstylecardillustrationstyletarot

LoRA · SD

Animix - Anime Screenshot-like Style Mix LoRA (アニメスクショ風/动画截图风)

20
Animix is a LoRA designed to generate images resembling anime screenshots by blending multiple anime-style LoRAs. It combines styles like 2020s anime magazine illustration, anime lineart, screencaps, and others to create a unique aesthetic. Use with Anything V4.5 and adjust LoRA block weights when combining with character LoRAs to minimize style distortion.animeanime screencapanime styleart styleartstylebeautifuldigital illustrationfemaleillustrationmalemenscreenshotstylewoman

LoRA · SD

Hipoly 3D Model LoRA

150
This LoRA is trained on high-polygon 3D model images to generate clean, high-resolution skin, hair, and detailed clothing. It enhances 3D texture and depth without requiring a trigger word, and its style can be controlled using `3d` and `realistic` tags. The LoRA improves detail without eroding the base model's art style.3dfantasygame charactergirlsrealisticstylewoman

LoRA · SD

Princess Peach - Nintendo The Super Mario Bros. Movie (Illumination Entertainment Animation Style)

10
This LoRA model generates images of Princess Peach in the Illumination Entertainment animation style from "The Super Mario Bros. Movie," trained on screenshots and promo images. It supports various outfits like her pink dress, blue dress, racing jumpsuit, and wedding dress, and works with several base models. Key trigger words are PrincessPeach, Jumpsuit and WeddingDress.characterfemalegirlsillumination entertainmentnintendopeachprincess peachuniversal pictures

LoRA · SD

Panties pulled aside fuck | Sex Lora 011 | Perfect

50
This model generates images depicting sex with the concept "panties pulled aside," suitable for both anime and realistic styles. It is triggered by the phrase "panties pulled aside fuck" and works with various subjects and backgrounds. Recommended weights are 0.6~1.pornstyle

LoRA · Wan Video 2.2 I2V-A14B

69 Deepthroat Blowjob - I2V Wan 2.2 Video LoRa - K3NK

3110
This LoRa model generates videos, specifically focused on the 69 deepthroat position. It was trainedactionfemale ejaculationfemale orgasmnsfwsquirtvaginal fluid

LoRA · SD

dunhuang(敦煌)

30
The dunhuang model generates images of Chinese Dunhuang Flying Fairies. It's trained on 512x768 images using locon 128dim, featuring diverse clothing colors, but may produce unstable hands, which can be improved with Controlnet.clothingdunhuang_clothsdunhuang_style

LoRA · Other

real spread pussy(experimental)

70
This experimental model generates detailed 2.5D images of spread genitalia. Newer versions offer improved accuracy but may negatively impact stylistic consistency; use lower weights or revert to older versions to preserve style. Trigger words and weight recommendations are provided, and the model is compatible with specific 2D models.2danimehentainakednudespornsemi-realisticsexvagina

LoRA · SD

Panties to side LoRA

20
This LoRA model generates images of women with their panties pulled to the side. Trained with improved steps and tagging, it performs well from various angles. Use trigger word 'pnts' with recommended LoRA strength between 0.6-0.9.clothingnsfwpantiesphotorealisticwoman

LoRA · SD

Rage Unleashed

10
Rage Unleashed generates images of characters transforming into a "Super Saiyan" mode. The model is trained on anime and comic characters, using "r1ge" as the instance token. It supports customization of energy aura colors like yellow, violet, green, red, and blue.animeconceptfemalegame characterloramanpersonportraitsvideo gamewoman

LoRA · Other

Split

30
Split position, doesn't work well with realistic models. Used a mix of AbyssOrangeMix2 with Anything V3animeerotic poseposespositions
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.