Made from the perfect blend of three NSFW models, Trifecta XL is capable of producing a much wider range of realistic NSFW images than any of its constituent parts. It scored higher on my prompt test than any other SDXL model and showed a very high level of understanding of positioning, posing, and the most common sexual acts. Trifecta was created from the painstaking evaluation of multiple rounds of merging and analysis of the individual u-net blocks from all three models until the perfect weight ratio and merge method was achieved.
The Trifecta model was created to be the base of a new generation of mixes, but it seems to work just fine on its own. However, its creativity is somewhat limited, and it tends to oversexualize everything. Which I guess is the point if you are going to use it straight up to generate images.
A note on watermarks: Yep, it's got them. However, they appear less often, are smaller, and are often positioned off the image. I managed to "push" many off or nearly off the generated image. Check below for usage tips for more information on how to limit them.
Usage Tips:
This is a replacement for using Pyro’s XL model as the base of some of my own future merges. It was optimized to produce ****. It doesn’t know many art styles, and it has a very narrow understanding of the world outside of ****. If you use it for something else, you are doing it wrong.
Noting the above, Trifecta does make an excellent base model to train LoRAs and use in your own NSFW merges. It has a very good understanding of anatomical structures, including hands and feet. Having training data that is a bunch of naked people doing all kinds of stuff with their hands and bodies will do that. And if you do use Trifecta to make or make a LoRA, let me know, and I’ll link to it.
Trifecta XL is capable of generating a wide range of soft and ******** images, but its conceptional understanding is also pretty vanilla – and that’s intentional. If you want something to cater to your darkest and most perverse desires, you’ll have to find a LoRA for it. Again, if you find or make LoRAs that work well, let me know, and I’ll link them to them.
The model knows what a penis is. It understands all the things that can be done with a penis. It does not always understand that a penis needs to be connected to a man. Perhaps the model is being progressive in its understanding of gender fluidity and that disembodied penises sometimes just float around giving and receiving pleasure. AI’s can be weird. If it happens, try another seed and make sure your prompt includes a male participant.
Keep your prompts simple when prompting NSFW images. The model gets confused easily if it thinks elements of the prompt conflict with each other, particularly when it comes to framing and points of view.
This model has much less of a watermark issue than other NSFW models – as long as you don’t mention it. Funny story: I tested this model out with 50 NSFW prompts to measure its abilities and comprehension. Out of 50 hard-core images, less than 1 in 5 were watermarked. Then I started generating images for the showcase, and EVERY image had a watermark. The only difference? I didn’t use a negative prompt or any quality keywords in my test. My showcase images all used “high resolution, best quality, 4k” in the positive prompt and a bunch of watermark terms and keywords in the negative prompt.
Prompting:
Natural language works best. Describe your subjects and what they are doing. As noted above, if you don't want watermarks, avoid using quality keywords in the positive prompt and mentioning watermark-type things (websites, logos, etc.) in the negative prompt. You'll still get some, but most are tiny or almost pushed off the image.
This model is not intended and does not tend to produce illegal acts and images. However, to be 100% safe, add "children, loli, child, kids, teens" to your negative prompt. The model understands "girl" = "woman," but it's safer to use the latter.
Because it's a merge of several models and LoRAs and was optimized for certain acts, most keywords, phrases, and concepts from any of the merged models still work. The examples below can be used but don’t need to be used in their entirety; a few of the phrases or keywords below out of the list usually work.
a woman in a reverse gang bang, sex, threesome
a woman titty fucking, penis, sex, titty ****
a woman is sucking a man's cock, a woman, a woman, Blowjob, Sucking a penis, a woman giving a blowjob to a man, a woman sucking a man's penis
a woman sitting on a man, penis penetrating *****, a woman having riding cowgirl sex, sex, riding cowgirl, a woman sitting on a man's lap, a woman's ***** is getting fucked by a huge penis, a woman is riding a penis
a woman, sex from behind, penis penetrating *****, a woman having doggy style sex, doggystyle, from behind, on all fours
a woman licking a penis, a woman licking cock
a woman is riding a penis, sex, viewed from the back, a woman sitting on man, penis penetrating *****, from behind, sex, reverse cowgirl
a woman bending over, ass grab
a woman, cum in *****, creampie
a woman with her fingers in her vagina, a woman fingering her *****, female masturbation
a woman getting a cumshot on her breasts, sex
a woman getting a cumshot on her face, sex
a woman giving a handjob
a woman having sex in missionary position, a woman on her back, spreading legs, penis penetrating *****, sex, a woman's ***** is getting fucked by a huge penis, vaginal, missionary sex, laying
Model Credits
The following models have been or are currently used in a version of Trifecta XL. Credit goes to the model creators (and, of course, the creators of the models that were used in their merges/training as well).
OmnigenXL (NSFW & SFW) https://civitai.com/models/203014/omnigenxl-nsfw-and-sfw
Aesthetic Mix XL https://civitai.com/models/262187/aesthetic-mix-xl
Pyro's NSFW SDXL https://civitai.com/models/133808?modelVersionId=191882
NSFW POV All In One SDXL https://civitai.com/models/144203/nsfw-pov-all-in-one-sdxl-realisticanimewd14-74mb-version-available
FappXL https://civitai.com/models/248856/fappxl-multi-nsfw-concept
Legal Stuff
AI model output may sometimes be unpredictable and is not guaranteed to match the input prompt nor to be of any given nature or quality. Do not attempt to generate content that is illegal in your legal jurisdiction or within your country or district of residence. By using this model, you agree that the creator of this model shall in no event be responsible or liable to you or to any third party, whether under contract, warranty, tort (including negligence), strict liability, indemnity or other theory, for any direct, indirect, special, incidental, consequential, exemplary, liquidated or punitive damages or any other damages, including but not limited to loss of profit, revenue or business, cost of substitute procurement, arising in whole or in part from your use of (or inability to use) the model.
Model Restrictions:
To be clear, this model is a derivative of the SDXL 1.0 model and the following restrictions are part of the 2023 Stability AI CreativeML Open RAIL++-M License dated July 26, 2023, with applies to this model and all models that use any component or derivatives of this model. You agree not to use the Model or Derivatives of the Model:
In any way that violates any applicable national, federal, state, local or international law or regulation;
For the purpose of exploiting, harming or attempting to exploit or harm minors in any way;
To generate or disseminate verifiably false information and/or content with the purpose of harming others;
To generate or disseminate personal identifiable information that can be used to harm an individual;
To defame, disparage or otherwise harass others;
For fully automated decision making that adversely impacts an individual’s legal rights or otherwise creates or modifies a binding, enforceable obligation;
For any use intended to or which has the effect of discriminating against or harming individuals or groups based on online or offline social behavior or known or predicted personal or personality characteristics;
To exploit any of the vulnerabilities of a specific group of persons based on their age, social, physical or mental characteristics, in order to materially distort the behavior of a person pertaining to that group in a manner that causes or is likely to cause that person or another person physical or psychological harm; For any use intended to or which has the effect of discriminating against individuals or groups based on legally protected characteristics or categories;
To provide medical advice and medical results interpretation;
To generate or disseminate information for the purpose to be used for administration of justice, law enforcement, immigration or asylum processes, such as predicting an individual will commit fraud/crime commitment (e.g. by text profiling, drawing causal relationships between assertions made in documents, indiscriminate and arbitrarily-targeted use).
Clarifications on the permissible use of this model:
This model is a merge. Therefore it inherits the permissions and restrictions of all models used in its creation.
You may not make merges, extractions, or derivatives of this model and use a less restrictive license than the 2023 Stability AI CreativeML Open RAIL++-M License.
This model may be run and hosted for download on Civitai.com. At this time, no other organization or individual is authorized to host or distribute this model.
You may not sell this model or any of this model’s merges or derivatives.
You must credit this model’s creator if you use this model to create a new merged, mixed, trained, or extracted model.
You may sell images you generate with this model.
You may share merges using this model as long as any merge has the same or more restrictive permissions.