By gerogero
Updated: April 4, 2025
This is a realistic/semi-realistic NSFW model list + hentai/anime NSFW model list for Stable Diffusion
New and don’t know how to use models? Read the getting started guide.
For a weekly updated list of top models and prompt examples, check out the models page.
Base models are expensive and time consuming to create, so there are not too many of them.
The most popular base models are:
All the models you see in common use are fine-tuned from (trained on top of) a base model.
In this article, we’ll go through the best custom models for each base model category.
We make this distinction because each base model category has its own advantages and disadvantages, as well as optimal settings.
You can try these models on this site without needing to download them:
Let’s get into the models.
SD 1.5 models are the bread and butter of the image generators you see online, both free and paid. Despite being the first generation of models, they are still widely used, and popular models are continually updated by their creators.
SD 1.5 models are very versatile, good for realism and anime, and have the widest range of accompanying LORAs (add-on models for specific tasks, like art styles or characters).
One of the most popular NSFW models of all time.
Anonymous creator, one of the most popular NSFW models of all time.
Better for sexy or cute girls than sex acts.
Aims for the perfect balance between realism and anime. Flexible with many kinds of sex acts – much better at actual sex than chillout mix. Highly recommended as your first model, because easy to get good results.
Very versatile model, improving considerably through new versions. One of the best models right now.
Good at a certain aesthetic of girl.
A relatively newer model, very aesthetic artistic.
Good at a certain aesthetic of girl.
Note: write at the prompt (Realistic:0.1~1.4) or(realistic:0.1~1) in the negative prompt.
Another anime / realistic blend merge.
Model that better represents Western faces.
A popular mix of other anime models.
Anime model with a illustrator/painterly style. Very popular for Pixiv, you see this style a lot.
Model with good colors, lots of creative possibilites.
Hardcore version of AOM.
Softcore/SFW/Soft NSFW version of AOM.
Pony Diffusion started as a model that could draw my little pony characters (and porn of them).
The creators realized that adding other content, including furry art and anime made the model better. Because it was trained on so many niche fetishes and sex positions, it became better at hentai and combining fetishes than most otherexisting models.
Today it’s usage goes far beyond Pony porn; most people use it to generate anime and stylized porn of humans.
Pony models work best with Pony LORAs.
The model that started it all. Technically, Pony is a fine-tune of SDXL, but it’s legacy is so enduring, and so many models have been trained from it, that it is considered a base model in its own right.
Pony has a distinctive semirealistic look; ultimately it’s not a realistic model, so you’ll have to rely on a realistic fine-tune of Pony for that.
An anime fine-tune of Pony, that comes with much flatter shading. A very great model for dynamic sex scenes and NSFW in general.
An aesthetic semirealism model capable of very high quality results
What are LoRAs?
They’re files you add to existing models to get better results. They can be trained for anything – an art style, a position, a pose etc.
Longer list: List of NSFW LoRAs
You use them by downloading the lora file, placing them in the lora folder in your stable-diffusion-webui folder, and using a trigger word for the LoRA in the prompt.
Here are some pretty good LoRAs that you use in combination with the models above to enhance results.
Mix from Chinese Tiktok influencers, not any specific real person. Works well with Chilloutmix but works with other models as well, to generate natural, cute Chinese girls.
This is a LORA for better nude body types out of generated images: makes proportions sexier.
Make the vulva on the inside for better aesthetics.
Complicated desired outputs = Complex prompts with mix of natural language and tags Complex prompt structure and order: Simple Prompt Example: Resu...
This guide was created to bring inspiration to this visual vocabulary. There is a short description for each pose so that you can connect the word ...
GPT-4o, released on March 25, 2025 went viral soon after release, bolstered by the Studio Ghibli animation style trend. Most people are curious if ...
This guide is intended to get you generating quality NSFW images as quickly as possible with Automatic1111 Stable Diffusion WebUI. We’ll be u...