Easy prompts for beginners – choose the green easy mode to make this type

then choose a model like gaiaproject-flux (or) ag-flux

This exact example: the gaiaproject-flux model selected, enter this Prompt:

masterpiece, best quality, 8k, ultra-detailed, photorealistic, 1girl, (bedroom), (bed), (oversized sweater), (no pants), (bare legs), (reading book), (coffee mug), (cozy), (relaxed), (natural lighting)

acornspinning11-flux

masterpiece, best quality, 8k, ultra-detailed, photorealistic, 1girl, (18 years old), blonde, (nightclub), (neon lights), (dancing), (crowd), (short holographic dress), (platform shoes), (euphoric), (energetic), (blurred motion), (low angle shot), (fisheye lens)

acornspinning11-flux

masterpiece, best quality, 8k, ultra-detailed, photorealistic, 1girl, (18 years old), blonde, (abandoned building), (graffiti), (broken windows), (dust), (decay), (ripped fishnet stockings), (oversized hoodie), (platform boots), (smoldering look), (rebellious), (urban decay), (dramatic lighting), (low angle shot)

Change the prompt slightly with Age and Ethnicity to create your character

Give her a name and repeat the name towards the end of the prompt every time with her age, ethnicity, and other important features.  The word order matters, this is a key concept in maintaining AI consistency.

flux (schnell)

3 beautiful 18-year-old girls, at the beach, stylish swimwear, sunny day, natural lighting, detailed beach setting, smiling and enjoying, relaxed poses, detailed facial features, vibrant and lively atmosphere

unchained-q6-flux

2 girls are playfully lounging on a bed, the girls dressed in vibrant bikinis, adding a splash of color to the room. The sunlight filters in through the curtains, casting a warm glow over the scene. They're laughing and enjoying each other's company, their playful energy contagious. In the background, another girl, also wearing a bikini, watches from the doorway, her expression a mix of curiosity and amusement, secretly observing the playful interaction.

unchained-q6-flux 

1980s adult girl portrait

acornspinning11-flux

masterpiece, best quality, 8k, ultra-detailed, photorealistic, 1girl, (18 years old), redhead, (freckles), (long curly hair), (large breasts), (bathroom), (mirror selfie), (sports bra), (leggings), (workout clothes), (sweaty), (post-workout glow), (natural beauty)

acornspinning11-flux

Mistress, leather boots, leather one piece, black long hair, model figure, tall, high heels, dominant, bold, she is walk on the beach with a pet dragon, teasing viewer

Model name:  ag-flux  (choose this from the easy menu or Tasks > Workflows in the image editor)

To create a similar image, try words like: Seductive woman, The prisoner, Zelda, atmospheric, cinematic, black corset, heart pendant, opening a door, seductive, analog photo

ag-odarka-flux

Cinematic photo - This is a photograph taken at dusk, capturing a young woman standing in shallow ocean water. She has a light to medium skin tone. Her dark hair is wet, slicked back, and tied up in a loose ponytail. She is wearing a black string bikini that accentuates her slender, toned physique. The bikini top is small, barely covering her small-sized breasts, and the bottom features a high-cut design that highlights her flat stomach and curvy hips. The background is a dramatic cityscape /size:768x1024 /seed:205401 /guidance:1.5 /steps:25

acornspinning11-flux

((NSFW, mature audience only)), slutty woman, adult party video game, roommates simulator, erotic sex scenes, customizable avatar characters

acornspinning11-flux

NSFW, A candid, energetic photograph of an 18-year-old girl working on her 1960s American muscle car in a cluttered garage. She’s crouched down next to the car, her hands greasy from hard work. Her face is smudged with dirt, and a few strands of hair stick to her forehead, damp with sweat. Despite the effort, she wears a satisfied smile, clearly enjoying the challenge. She’s dressed in worn-out jeans and a loose tank top, stained from the day’s work. She has amazing breasts, full and large

Intermediate & Advanced – use the hot pink “image editor” to create these

For those of you interested in creating believable influencers, where the same character appears in different scenes but is consistent, you’ll need to master the image editor and check our tutorials about how to manage seed and a consistent but ever-slightly changing prompt

Two examples, a close-up selfie and a body shot:

<donutsdelivery-soft2-xl>score_9, close-up, looking at viewer, holding up V fingers, cute e-girl, solo, (freckles across nose:0.6), grinning, long hair, bleached blonde hair, pale eyes, tiny pink and white striped bikini, lips, eyelashes, thick mascara, heavy eyeliner, portrait, realistic[[score_6, score_5, score_4, source_pony, source_anime, source_furry, source_cartoon, tongue out /clipskip:2

<donutsdelivery-soft2-xl>score_9, looking at viewer, holding up V fingers, cute e-girl, solo, (freckles across nose:0.6), grinning, long hair, bleached blonde hair, pale eyes, tiny pink and white striped bikini, large natural breasts, breasts to the side, breasts apart, lips, eyelashes, thick mascara, heavy eyeliner, portrait, realistic[[score_6, score_5, score_4, source_pony, source_anime]] /clipskip:2

<boltning-hyper-xl> instagram photo, 1girl, black woman, perfect detailed eyes, natural skin [[(face asymmetry, eyes asymmetry, deformed eyes, open mouth)]]

You can also do anime and 3D art styles

1 girl, cute girl on display, bound arms, arms behind back, shy smile, crying, 21 years old, ribbon in hair, short shorts, garter, crop top, slutty rave costume, slut, dirty blonde hair, slender, freckles deep skin, masterpiece, best quality, highly detailed, 24 years old [worst quality, low quality:1.4, watermark, distorted penis, missing limbs, chubby, fat] <kidnapped:0.8><animerge24><better-cute-asian:0.4><victoria-justice2:0.5>

unchained-q6-flux

anime style illustration, niji style, masterpiece, 8k, ultra realistic photo of 1girl, looking at viewer, smile, blonde hair, long hair, curly hair, earrings, blue eyes, large breasts, black jacket, hooded jacket, long sleeves, open jacket, white shirt, striped shirt, black shorts, shopping bag, mall, night

realvisxl2k

3d octane render, video game ninja gaiden female character, slutty leather costume, garter belt, sleeveless, skin tight clothing, the legend of the purple unicorn and the frisky virgin, Highly detailed and amazingly decorated, kinky, bdsdm, in a purple ornate bedroom, floral pattern, wooden furniture

sd3

a future female-only world, futuristic city background; detailed face and body, intricate details

[[<fast negative-xl:-2>]] <midjourney-mimic13-xl:0.3> /steps:40 /nofix /parser:new <boltning-xl> 1girl, a striking young blonde female war photographer kneels confidently in a rugged environment holding her film camera. Loren Gray. She has a determined expression, (her camera slung casually over her shoulder by a worn leather strap). Her appearance is both tough and elegant: her hair in a loose bun and framing her face, dressed in function.

<donutsdelivery2-xl>score_9, in the style of film noir and dark city, cinematic, black and white movie still, a beautiful woman smoking a cigarette in an alleyway [[score_6, score_5, score_4, source_pony, source_anime, source_furry, source_cartoon, from above, overhead shot, tongue out, worst quality, low quality, face asymmetry, eyes asymmetry, deformed eyes, ugly, airbrushed, simple background, cgi]] /clipskip:2

<boltning-hyper-xl>a jazz flapper girl, 20’s woman in jazz club, collar choker, (eye shadow, makeup:1.2), black hair, pale skin, bokeh, sharp focus on subject, fullbody, a wild and sensual dance, dancing in a jazz club, cosy atmosphere, gas light, cinematic, dark, chiaroscuro, low-key [[(face asymmetry, eyes asymmetry, deformed eyes, open mouth)]]<illustration-insanity-xl:0.4>

The images may appear rough on the drafts, so use the UPSCALE tool to boost them.  The first image was 800×900 and the final one shown here was upscaled to 1664×1216.

Tip: Do not start at very high resolutions, allow room on the video card some room or it cannot upscale an already high-res image. A good start range is 896×1152 or thereabouts, or 1040×1024.

amateur cell phone photograph [[<fast negative-xl:-2>]] /steps:30 /nofix /parser:new /clipskip:2 <boltning-hyper-xl>1female, cute, young, dark hair, pixie cut, freckles, freckles, petite body, looking up at viewer, zone-tan, hair over one eye, ahoge, drill hair, skull hair ornament, off-shoulder sweater, black thighhighs

Examples from an earlier version of our software

These still work, and you can use these as guidelines with our newer models and tech!

portrait of a beautiful 22 year old Brazilian woman named Adriana Santos, epic boobs

And click Brew, and you’re done!

To choose a different look, click the Advanced menu and instead of one of our easy recipes, choose the model you want.  Try the one called Photon.

Your result should look like the image below:

Video: How to use the Stable Diffusion tool

Examples from our community

You can use these prompts using your private Web UI or in our Telegram Chat Bot.

In Telegram, simply start the prompt with /render

#boost award winning portrait photograph ((full body)) salvadorean 18 years old woman, long black hair, wearing sexy lace lingerie, natural breasts, stockings, highly detailed skin, soft smile:0.8, detailed hands with five fingers, sharp focus, photorealistic, gods rays, cinematic lighting, 8k, film grain

((masterpiece)), (photograph), Korean, huge breasts, ((tiny mouth)), ((smiling)), in the dark, deep shadows <majic6>

(((Cute British adult 18 year old girl, UK, shy, tight, Insta, naked, warm light, Instagram, dilapidated house, golden hour, brunette hair, Britain, sunlight pouring in, selfie, touching belly, ruined building, memories, phone camera photo, healthy weight, confident, nude, small breasts, belly)))

shirtlift Latina, curly hair, hoop earrings, white sweater, big tits, sitting in desk, classroom chilloutmix style  #boost <urpm13>

((Masterpiece, best quality, extreme hentai, rule 34 porn, gimp porn still)) of ((1girl, petite, hourglass figure, tightly bound in latex gimp suit, glowing eyes, bdsm, constricting latex, pain, large breasts puffy nipples, perfect gorgeous face, long beautiful hair)) <urpm13>

GlowingRunesAIV2 orange, woman, casting spells over army in an open field for battle, masterpiece, good quality, realistic,  <deliberate2>

(Real, amateur, college sluts, explicit, candid party sluts, CHi omega sorority porn), ((1girl, sexy, naked, at basement party, college, alcohol, post sorority rush party)), ((medium perky breasts, puffy nipples, giggly, beautiful puffy labia, slutty)), ((cute face, drunk but happy)), ((sexy ass, delicious butthole, pussy)),(soft warm indoor basement lighting, global illumination, dynamic shadows, fine details, wealthy basement)  <realflex2>

((( a pretty indian model, edge style))) best quality, RAW photo, realistic photo  <edge>

8kUHD,extremely beautiful young woman, Semi-Nude <paragon>

(((Cute British adult 18 year old girl, UK, shy, tight, Insta, naked, warm light, Instagram, dilapidated house, golden hour, brunette hair, Britain, sunlight pouring in, selfie, touching belly, ruined building, memories, phone camera photo, healthy weight, confident, nude, small breasts, belly))) <degen>

 /size:768x1024 ((( hyper realism, Pascal Blanche Rutkowski Repin artstation. 1990’s Nickelodeon, ((Penelope)), big ass, (Toha Heavy Industries), feeling alone in a crowd, cargo pants, black top, alabaster skin, cracked background, doubtful pose, perfect face, hyper detailed, jungle theme, sensual smile, black pupil, inked, splash art, minimal, masterpiece, best quality, highly detailed, intricate, art by Carne Griffiths Wadim Kashin, fantasy concept art, 32k resolution, heavy strokes, paint dripping, bokeh, hyper realism, photorealistic, beautiful detailed intricate, insanely detailed, natural skin, award-winning photograph)))))<buttlift:0.6><vivy>

RAW photo, Onlyfans nudes of latina girls, (side view, turned away from camera). crowded goa festival, extremely detailed background, tribal clothes,dancing together, teasing curves, (big breasts), hard nipples, cute smile angle from above <degen>

NSFW, a young 18yo girl, cute face, in a t-shirt, undressing, no underwear, with a medium body and young breasts)), best quality, high quality, high resolution, realistic, (masterpiece), (over-detailed), (hyperrealism:1.2), extremely detailed eyes, (extremely detailed face), (detailed skin), textured skin, (photorealistic, over-detailed:1.2), crazy details, seductive look, cute face, stunningly beautiful, (stands far away from the camera), (gray braids hair), ( pissing in the livingroom), skin details, pores, (female ejaculate from pussy:1.3), ultra realistic, sharp focus, 8K [worst quality:2], [low quality:2], [normal quality:2], [text, a man, a boy, too many person], ((thin body)), ((slim body)), visible ribs, selfie, posters in a room, [panties], [[[hairy pussy]]] <photon>

((( standing vaginal, creampie, close up, bending over standing vaginal close up, photo of beautiful brunette side swept hair skinny woman supermodel wearing black shirt getting fucked from behind by man in a office, (masterpiece:1.3), (perfect face:1.2), professional photoshoot, incredible detail, (best quality:1.3), hd, 8k, (1girl), (1boy) [(creampie:1.2), poorly drawn face, anime, cartoon, 3d render, low quality, blurry, hands, multiple men, poorly drawn lips, poorly drawn mouth, ugly, table, desk, shelf, bending over, extra fingers, poorly drawn hands] ))) [[frame, border, signature, watermark, text, letter, number, deformed iris, deformed pupils, semi-realistic, cgi, 3d, render, sketch, cartoon, drawing, anime, text, close up, cropped, out of frame, worst quality, low quality]] <photon>

cybersex in virtual reality, 8kUHD, Airbrush impressionism <paragon>

sex afterparty, 8kUHD <photon>

internal affairs, masterpiece, 8kUHD <paragon>

What are Recipes?

Recipes (the thumbnails pictured below) are easy presets to give you a distinct art style without a lot of prompt engineering.  For those that want to create something more exact, try this video tutorial from our technology vendor. Our software works in the exact same way.

Advanced Mode

The two major “tricks” to making AI pictures look realistic are (1) pick the right AI base model, and (2) pick good negatives, meaning the things you don’t want to see, like blurry pictures and gnarly hands. Try it out:

Example prompt

A portrait of a beautiful 21 year old Swedish woman drinking a beer next to the river in Amsterdam

Choose a recipe, or add a strong negative

1. Copy the prompt above into the “Describe Anything” box, this is the “make this” area, the positive.
2. Click the Advanced button (bottom left)
3. Click “add concept” and choose a base model. This will determine the overall art style.  You can also add a character or background later, but let’s keep it simple for this tutorial.
4. In the “but not this” text area, paste this:

[[[<verybad-negative:-1.5><fast-negative:-1.5><bad-hands-v5:-1>]]]

You can also add your own words like [big hairy nose] to suppress anything that you don’t want. We’ll explain what those specific “very bad” codes above mean in the advanced tutorials later.

Please note that we are always updating our models, so changes in versions may result in slightly different images. If you aren’t getting good results, or have a question, shoot us a DM on Patreon so we can help.

Welcome!  We are the official website of MyAiGF, as seen on Reddit and Instagram.

You can design your dream GF, or talk to our premade girls ready to go.

Everyday, we’re posting the hottest AI-generated girlfriends, and now we’re also sharing our tools so you can create them with your web browser, on your phone or on your computer. Everything is rendered on the cloud by our hosting company (Graydient) so it doesn’t matter if you have an old phone or computer.

Our service is unlimited and open-ended, you can create anything you want. Ask for selfies and chat all day, there’s no additional costs or hidden fees or “tokens” or any such nonsense. Our service is truly unlimited and endlessly customizable

Start your 7-day free trial, Satisfaction Guaranteed, with unique models that only MyAiGF has created, and one of the most powerful AI.  You’re going to be blown away, so what are you waiting for?

STEP (1) Sign up

Create a PRO Account to get instant access

STEP (2) Receive a login email link

Unlock our service to meet these lovely ladies right away.  Check your spam folder if you don’t see your login link.

Talk to our pre-made GFs or create your own

Also Include: Easy Pose

Choose from easy preset buttons and choose the position you want your character to appear in.

Over 500,000 possible combinations!  Here’s just a few (the real page is 5x as tall)

Also Included: Full-blown AI Image Editor “Stable Diffusion”

You can also create free-form NSFW images of any kind. We have many templates for tricky poses and situations, hot stuff!

Start with our easy-to-use widget generator, example prompts, or for experts you have over 6,000 unique AI models with all of the popular features of Stable Diffusion at your fingertips.  You can use any device, even an old smartphone, as we handle everything in the cloud.

The best of Stable Diffusion SDXL, Pony Diffusion, and SD15

Creating NSFW AI images takes a little practice, but it’s not rocket science. You can do this:

Pictured below: (a) Type what you want to see (b) Pick a recipe, which are premade popular art styles created by our community and (C) hit Brew.  It’s that easy. This is the PC Browser version…

You get a private drive, just like this.  On the top right I have tabs for creating, my favorites, my files, and my account.

Mobile Version

Here’s what it looks like on your phone, it’s the same platform. You can start something on the computer and take it to go on your mobile without missing a beat, everything is always sync’d up and backed up.

This is my own private space. Nobody can see what I create.

After the image comes back, there’s a Tools menu to request changes.  You can paint things on and off, change colors, and more.

And there’s a lot more features when you click Advanced.  We’ll cover that in the next guide. Isn’t that easy?  And that’s the basics!

Create an account and try it yourself

For Beginners – Using Recipes

We’ve made it super easy to get realistic images – simply choose a realistic recipe first (b) and describe the girlfriend of your dreams (a) and click Brew (c).  This is the easiest way to do it and 99% of people should do it this way. Just get creative with your keywords.  Check out our Reddit to get inspired, really the only limit is your imagination.

For Experts: Positives and Negatives

For AI enthusiasts, click the advanced tab and get really specific with the use of positive and negative tags.  These big blocks of keywords are what are hidden behind recipes.

You can also craft your own manually like this:

Make This (positive tags)

A positive tag means how strongly you want to emphasize something.  For example, the more parenthesis you add, the stronger the idea is. For example, to make the word boobs more important, we can do two things:

1. Move it the front of the prompt, because order matters most, and
2. Add parenthesis like this:

(boobs) – 1.1x important

((boobs)) – 1.2x important

(((boobs))) – 1.3x important

You can also use commas like this.  ((big boobs, perfect abs))

Now, it’s important not to overdo it, as too many parenthesis will create blue noise and artifacts

You went too far, or have conflicting information, if blue boxes appear.  Just tone it down a bit or set your weights lower to fix it. Setting a lower guidance (bottom menu pulldown) can also help reduce those blue boxes that appear when prompts have conflicting information.  Lower guidance gives the AI more wiggle room to fill in the blanks.

But not this (negative tags)

You can also use NEGATIVE prompts to boost realism. While it may seem obvious to humans, the computer doesn’t know what a good picture is for the mood you’re in.  Sometimes people ask for zombies, you know?

Example: [[mutated fingers, fused hands]]  that will improve the hands

Use negatives like a Boss: Textual Inversions

There are two categories of Textual Inversions: Positive and Negative. 

Textual Inversions are found in the concepts section of our system. They are a type of “mini model” that support and focus the result of a base model, like a powerful prompt.  Another common name for Textual Inversions is “embeddings”.  It means the same thing.

Positive TIs  (pictured above)

When you see models like celebrities, people, anime characters, and art effects, you are looking at either LoRA mini models or positive Textual Inversion mini models.  If you’re not sure, you can click into each model to find instructions.

Put models like this in the “make this” positive prompt area and adjust the effect with a weight like <jenna-ortega:0.7> (70%) or <jenna-ortega:0.2> meaning 20% influence.

Negative TIs are you ultimate cheat code to great images

(Pictured above: click the negatives tag to find them)

To get the highest quality images possible, you must learn to master negative TIs. You’ll know right away when you are looking at a negative textual inversions, because they have negative, counter-intuitive sounding names to remind you not to plop them into the positive box.

Read these names for example, and you’ll begin to understand.  For example, look at the first one. Who would want to use a model called “Bad Artist? Oh right, that’s a negative.”

Placing [[<bad-artist-anime:-1.5>]] in the negative prompt box means “don’t give me bad artist results, by the power of 1.5x”.  Powerful stuff.

But you wouldn’t use something called “artist” to make a realistic picture.  The ones with more general-sounding names like [[<verybad-negative:-1.5>]] are recommended for realism.

A few favorite negatives, explained

[[<verybad-negative:-2>]]   – a powerful, general purpose choice

[[<fast-negative:-2>]] – similar to the above, seems to compliment it

[[<negative-hands:-1.5>]] – it’s not perfect, but it seems to help hands

Activating Negatives and Trigger Words

To activate that properties of a negative textual inversion in the Web UI, simply click the “apply to prompt” button as shown below:

… and slide until it has a negative value.  Between -1.5 to 2 is usually the sweet spot.

Re-using your favorite negatives quickly

Tip: You can re-run your prompt by clicking Advanced > History

No need to navigate the menus and re-pick the models every time. Just click the history button. If the history is blank, it’s because you’re logged out. Click profile > send yourself an email to login.

Another way: When you find a pleasing result, click the prompt button below the image and copy the prompt to repeat it.  This way, you don’t have to remember things like model names, trigger names, seed, samplers, guidance, etc that went into making the best image.

Clicking the Prompt button under your results lets you quickly copy and repeat your last result.

Knowing your Base Models

Use the concepts picker to choose “Realistic” from the tags menu.  There can only be one “base”, this controls the overall art style. You have over 100 preloaded base models on the cloud to choose from, each with different interpretations of beauty. Some look like studio photography, some look like photos snapped right from an iPhone. There’s anime, comics, and everything else inbetween.

How to tell Base Models Apart?

This part just takes patience and experience. If you click into each one, there is detailed information on how to use it. The thumbnail images will give you a pretty good clue.

If I’m going to make a beautiful Japanese woman, it’s much easier if I start from a base model that specializes in Asian beauties.  Use our tags to find them. In this case, BRA 5 (Beautiful Real Asians) is an obvious choice.

You can combine a “full model” (aka full model) with up to 3 mini “LoRA” models.  There are many categories of LoRAs that can reduce the amount of prompting you need to type to get a good effect.

Put it all together and you’re going to get stunning images!

Introduction to After Detailer

If you’re using a Stable Diffusion 1.5 model like <photon> you can also add the parameter /adetailer anywhere your prompt.  It helps with faces and hands. Unfortunately, it doesn’t work with SDXL yet.

Back to Tutorials

Use what you learned here to make any kind of Ai GF, like this!

Its easy to upload and edit your own photos, use your own photos as a starting point, and change anything about them.

Important: To see the upload button, you must be logged in.  It is possible to browse your private website as a guest, so to login click > profile > switch and send yourself an email.  Click that and the upload button will appear.

Recommendations:  Photos at about 1000×1000 work best, and 100% high quality JPGs work fastest.  If you try uploading an uncompressed iPhone it also works, but it’s slower to upload because the files are so large.

On the bottom left hand corner, there’s a button for Advanced.  Click on that to reveal the advanced editor, which has two modes.

Free Text Mode (default)

Two boxes will appear, and now you can type what do don’t want to see. For example, I don’t want freckles or a big nose.  You can correct any image and remake it as many times as you want.  Our service is unlimited, no credits.

Weights Mode (shown below)

Alternatively, you can use Weights (below, top right) which puts a slider on every word. So if you want to see a 50% Russian and 50% Chinese girlfriend, try the weights mode, it’s a riot.  Weights deserves its own guide as there’s a lot to say, but for now just remember this 1 rule:

In Weights, less than 0 weight means to hide that word or concept.

Change the character

We have many diverse, gorgeous premade characters you can choose from, including celebrities (if they are in Stable Diffusion). Using the sliders, you can blend characters together.

You can choose from hundreds of premade characters, and use the “weights” system to blend them together to create a girlfriend that’s unique to you.  Every prompt is saved in “history” so its easy to go back and edit the same characters into different new scenes and stories.

Change Anything: Appearance, Poses, Art Styles, Expressions

You can mix characters together, set a pose, haircut, environment, add more than one character to a scene, and more.  Over 1000 options that you can mix and match in every way imaginable, in every art style.

For example, putting your girl in bondage play

There’s a bit more to it, but those are the basics!

This is an advanced tutorial. We recommend the beginner guides first

Q: I’m able to create the girls I want, but they look like different characters from one scene to another. How do I fix this?

We can definitely help.

All the tools are here.  Try the tactics in this guide, step by step.

Step 1: Lock the character in

Click advanced, and paste this template:

/seed:342342

OK, but what is a seed? 

A seed is a made-up reference point. The number is random, but it acts as an anchor, just like GPS coordinates.

The seed number itself doesn’t mean anything per se — it doesn’t refer to a specific person.  It doesn’t mean to look up a specific photo in a database, either.  What seed does is help the AI understand that you’re talking about the same pattern across different prompts, in an abstract way. Seeds are great especially useful when re-running a prompt, and sharing one on the web. Seeds were built for repeating similar results.

For example, “12345” or “3948394” or “6969” are three examples of valid seed numbers. Use whole numbers without zeroes. You can go into the trillions if you’d like, but a bigger number doesn’t make it better or worse, it’s just a random number.

FIGURE 1: How to use Seed

Step 2: Locking the characteristics

Word order matters. A lot.

Describe your girl at the start of your prompt with very specific details: Age, Race, Country

Putting age and some ethnicity or other specific characters at the start of the prompt also makes a big difference. Let’s say I want to create a “20 year old half-Asian woman”. But what kind of Asian?

The more specific we are, the better the chances the AI can repeat that result.  So “20 year old half-Korean red-haired woman” is the better choice. You can also wrap the most important characteristics in nested positive prompts to really drive this home. Now, that’s already a pretty unique idea for a character, let’s see how we did.

So far, our prompt looks like this

 /seed:342342 (A 20-year old half-Korean red haired woman)

Don’t worry about things like size and resolution or quality at first, this just slows the render down. You can boost things at the very end when you have the perfect repeatable prompt. Select Draft to get back 9 images fast. Render your idea and see how you did.

FIGURE 2: Our first draft

Step 3: Reducing inconsistencies

Imagine you’re talking to a police sketch artist. This girl just stole your heart and your wallet. You tell the police that “she looks like a 20 year old Korean college student with red hair”…is that specific enough to catch a crook?

Figure 3: Close but too much variation

Here’s the lineup. They’re all similar, but notice sometimes the hair has bangs, sometimes its long and parted, the nose bridge is different sometimes, and the general shape of her face is sometimes triangular or heart-shaped.  It’s my fault, my prompt sucked. I wasn’t specific enough.  Remember, a seed is just a general pattern, and not a specific person. So we are 50% there.

Go way more specific

Does she have big nose? A pointy chin? Long eye lashes? There’s so many possible chances to arrest the wrong girl just based on that information. We still haven’t really narrowed it down for a human sketch artist, much less for the AI, so inconsistencies are still possible. 

Of course, you can prompt all of these characteristics every time, but the longer the prompt is, the less the AI listens to the words at the end, we still haven’t included things like clothing and what’s happening in the scene. Long detailed prompts are good for creating single images, but not a series of them. So where do we go from here?

LoRAs

Low-rank Adaptation models (LoRAs) are an instant fix to getting what you want. Our system has 1100+ of these in every style, whether it is a likeness of a celebrity or a shortcut to getting someone in certain clothes or in a pose. You don’t have to use celebrities at all, there are hair, clothing, fashion, all kinds of models in here:

Figure 4

Step 4 (Optional) add a resemblance

Unless you’re making your own AI model (we’ll get to this later), the next best thing is blending characteristics from other models, to strongly inherit the same traits every time you want to make an image.

Here’s how to do it:

• Click Advanced
• Paste our prompt again, same characteristics, same seed, same base art style  /seed:342342 (A 20-year old half-Korean red haired woman
• Click Add Concept
• Click the NSFW tab to reveal all 1100+ models
• Click on Females
• Add 1 or 2.  Three tops is recommended. Save space for pose and clothing slots.
• Use the slider to increase or lower the effect
• Low weights are recommended to start with. LoRA characteristics are stronger than your prompt. You may need to add more (positive) brackets after a LoRA is applied.
• Optional:  Change the base art style if you want.  You can combine a recipe with a different base concept.  I went with Realvis13, for example.

Examples:

<gadot:0.15> – she will look 15% like Gal Gadot

<koreandoll2:0.47> – she will have 47% of the face and body a “Korean Doll”, a premade curvy hot Korean AI girl

Figure 5

Let’s look at the resulting images. Much more consistent already!

As you can see, compared to Figure 2, our results are a lot more consistent already across six images. They’re almost certainly the same person in every image, and we didn’t have to do a lot of prompt engineering for this, either. We didn’t have to prompt anything about parted hair, straight to wavy hair, large breasts, bright red lipstick, and painted eyebrows, those are all characteristics strongly inherent in the LoRas and base model that I selected.

I rendered this using 15 steps, so the realism and detail isn’t as high as it could be.  At this point we’re still developing the basic concept, so we’re not shooting for quality yet.

Time-Saving Tip:

Hit the “Prompt” button under the photo to get the exact prompt that generated the picture.  You can then easily tweak the variables without having to go through the visual menus every time.  Just paste this into the positive box and click the recipe or pick a different recipe if you’re not 100% loving this art style.

Click the Prompt button from your render (or) My Archive screen to continue to Step 5, poses.

Notice that she looks a tad older than my original prompt. This is because the LoRAs I selected influenced her age.

We can stop here and make some creative decisions.  If the base model isn’t doing it for you, try a different one, or make the age a stronger part of the prompt to get what you want.  I can say (((20 year old))) and she will look younger than this image, but still similar. And that is how you can tweak things like hairstyle or lipstick without changing the character.

Common mistake: changing the base model or recipe and expecting the same girl. 

Each AI model was created by a different model maker, so the character is unlikely to appear across different models.  The same prompt will produce an entirely different person if we switch from <realvis30> to <chilloutmix> because the training photos for each model are totally unrelated. The same prompt on a different base model will be a completely different girl.

Step 5: Same Girl, Different Pose

OK, let’s get to the good stuff, how about some fellatio?  

Copy the prompt and add “sucking a man’s dick, oral sex, intercourse” etc.  The more specific you are, the better.

Tip: Be specific when creating sex scenes. If you simply say “sucking a dick” the AI will take that literally and make a dick appear out of thin air. To help you get this right, there are also LoRAs for poses. Go back to the concepts browser, turn on the NSFW toggle, click porn, and add the pose to your prompt.

Here, I turned on NSFW, selected “porn” first and then “oral”.

When you add poses they also have a slider (weight) to control how strong the effect should be.  Starting from 0.7 or lower is recommended. If you get glitches try a different value. Some LoRAs do contain face data, so sliding it to high values may impact your character.

Still pretty consistent results!  Same nose, same hair, and so on.

How about a change of clothes and a different situation next?

Step 6: Different clothes, different situation

Put it all together.

Oh, the places you will go!

Notice how much room we still have in this prompt to describe a much more specific character, but the results are generally consistent. The short hair bangs and high nose bridge never came back, but I aged her a little by mentioning the old men on the cruise. To avoid that, I should have inpainted the old men instead, isolating those words from my girl.

/seed:342342 (A 22-year old half-Korean red haired woman), wide shot, (floral dress), eating a sandwich on the rail side of a luxury cruise out at sea, old men in the background, sunny cloudy sky <gadot:0.42><koreandoll2:0.69>

Remember, we’re still looking at draft quality here. Go back to your prompt and try running it at a higher resolution, using HighDef tools and Facelift.

Even at 512×512 she’s pretty cute:

TLDR Review

Use these tactics to create consistent, repeatable AI characters:

• Make up a number and add it to your prompt, aka a “Seed” number
• Use very specific language to describe them at the beginning of the prompt, such as age and nationality
• Stick to one base model and one aspect ratio when possible
• Keep a journal of prompts you liked, don’t lose the seed number, etc.
• Clothing, hairstyles, and poses with LoRAs
• Reinforce by adding other celebrity names in the positive prompt box
• Use Inpainting when there is more than 1 character in a scene to separate characters

Inpaint, also called Generative Fill, is when you are painting something new into the scene, like removing clothes, changing some anatomy, and things like that.

First, create an image and then click the tools button

There are different kinds of inpaint model brushes you can choose from, shown below.

• Realistic Vision, Rev Anime, and Uber Realistic Pornography Merge are exactly what they sound like, and they’re really powerful.
• SD 1.5 means Stable Diffusion.
• Dreamlike is an illustration style.
• Hassaku is Anime Hentai.
• There are more, most of them realistic. We add new inpaint models all the time. Try “Epic”, one of our favorites!

Choose your brush

The text box below then tells the AI what to fill in next.  So for example, I can pain over her body and then describe a different outfit, and then press the Inpaint button.

The AI is very powerful but it sometimes gets confused with things like fingers, so its easy to paint to fix it, or to re-prompt those hands in a different position.

Literally use your mouse or finger to paint over the image, drawing a “mask shape” on the image

Then hit inpaint, and you’re done!

Removing Clothes Tutorial

1. First Login
2. Then click Inpaint in the top menu
3. Then upload your photo

Example: 

This woman doesn’t exist, we created her.  Let’s get her naked using Inpaint

See the Brush size on the bottom right?

Think about the exact words. In the case of a naked woman, what color are her nipples, is she athletic, do you want to also change her hair, is she wearing something, and so on.  You can also cancel visual ideas.  For example, if you want a woman shaved, the “shaved” goes in the positive box, and words like “hairy, pubic hair” go in the negative prompt.  You can also fix common anatomy issues in the negative box.

Positive and negative prompts

(((nsfw))) a nude woman, small breasts, fit, pink nipples, abs 
Copy

Negative

[[[man, male, clothing, shirt, dress clothes]]]Copy

(Optional) You can go really nuts with the prompts.  Here’s a notable one we found on the Internet to boost realism:

High pass filter, airbrush, portrait, zoomed, soft light, smooth skin,closeup, Anime, fake, cartoon, deformed, extra limbs, extra fingers, mutated hands, bad anatomy, bad proportions , blind, bad eyes, ugly eyes, dead eyes, blur, vignette, out of shot, out of focus, gaussian, closeup, monochrome, grainy, noisy, text, writing, watermark, logo, oversaturation , over saturation, over shadow, floating limbs, disconnected limbs, kitsch, cartoon, penis, fake, [black and white], airbrush, drawing, illustration, boring, 3d render, long neck, out of frame, extra fingers, mutated hands, monochrome, [[poorly drawn hands]], [[poorly drawn face]], [[[mutation]]], [[[deformed]]], [[ugly]], blurry, [[bad anatomy]], [[[bad proportions]]], [[extra limbs]], cloned face, glitchy, bokeh, [[[long neck]]], [child],[childlike], 3D, 3DCG, cgstation, red eyes, multiple subjects, extra heads, close up, man, text, bad anatomy, morphing, messy broken legs decay, ((simple background)), deformed body, lowres, bad anatomy, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low jpeg artifacts, signature, watermark, username, blurry, out of focus, old, amateur drawing, odd, fat, morphing, ((simple background)), artifacts, signature, artist name, [blurry], disfigured, mutated, [poorly hands], messy broken legs, decay, painting, duplicate, closeup

It’s time to literally paint

Paint over the entire body with your mouse, or using your finger.  Touch the photo and the paint brush will begin.  There is an eraser and reset button if you make a mistake.  It doesn’t have to be super exactly precise, you can be a little sloppy.  The AI will figure out the rest.

Almost finished!  Last check

Make sure everything is selected – the URPM model (or whichever you prefer) and that we have a prompt in the positive and negative box

And now, hit the inpaint button

And you’re done!  You’ll get five variations.

Repeat the process to fix any imperfections, or use the /facelift command again to upscale it to HD

Tada!

Experiment with different positive and negative prompts to change the results to your liking.  Another example:

You can also undress AI images that you create in the app

First, render an image.  Try this prompt:

/render /sampler:dpm2m /steps:75 /guidance:7.5 /portrait 8kUHD, award winning photography, light art by Daniel muller, raw, 23yo, brunette, swedish woman, happy, gorgeous face, highly detailed facial features, technicolor erotic body lighting, beautiful body proportions, sweater, high detailed skin, realistic skin texture, realistic colorful light patterns on body, real hair movement, 16k, dslr, sony a7 mark iv, (beautiful lights on naked body, silhouette), euphoric, (dynamic shadows), (erotic body lighting) [[[nsfw, Sketch, painting, cartoon, anime, animation, cgi, rendered, doll, lifeless, sad, depressing]]] <realvis20>Copy

When the results appear, click Tools > Inpaint 

The “Edit Inpaint” browser allows you to select the AI model on the left, expand positive and negative prompts (click refine inpaint prompt), and draw where you want your changes with a brush. Check our models page to get familiar with what these code names mean. This soon becomes second nature.

As this is a realistic image, we’ll go with URPM13-Inpainting.  Only two models are shown here, but there are over a dozen now to choose from.

Last step:  Refine your prompt, paint over the area with your finger or mouse cursor, and click the inpaint button.  You can then close the window.

Finished inpainted images should look seamless. Inpaint is amazing!

You can inpaint as many times as you like.

Go ahead, add a mustache.

Troubleshooting

Q:  The face changed

Stable Diffusion doesn’t handle low resolutions very well.  You’ll first see your high-megapixel photo from your phone, and then a 512×512 image from Stable Diffusion, which is the dimensions it was trained on.  You can upscale your image but the face will change.

To avoid this: If your person is a full size image, try adding the parameter /size:768×768 when inpainting.  You can go a bit higher, but going too far from 512×512 may cause doubles.  It’s better to upscale the finished result than trying to go too high at once.

Q:  The body style or features look different than I imagined

This is what different inpaint models and LoRAs are for. Each inpaint model will respond differently to your prompts. Instead of saying “naked” you may want to get more specific:  “naked 22 year old Brazilian” etc — and that will produce a slightly different result depending on the inpaint model that you choose.

You can add get extremely specific in your prompt, but to really get the results you want, use LoRas.

For example, there are breast size adjusters, nipple size and color, and so on.

Q:  The photo isn’t undressing

Make sure you selected an appropriate inpaint model for the job, like URPM.  It’s also important that the input photo is bright and clear.  Try redrawing the mask from scratch again and copy one of the preset prompts in this lesson above.

Advanced options menu

Hit the advanced button on the bottom left to reveal the aspect ratio, guidance and steps options.

Here’s what each of these do:

• The default image proportions are Square, and for a good reason. Our AI was trained on square images.  Going too wide or two tall without describing other things in the image will sometimes get you twins, so square is a good place to start. You can change this to tall, wide, or landscape.
• Prompt Guidance:  The lower the number, the more freedom the AI has to interpret what you mean, but things can get blurry.  If you’re getting blue squares, stepping down in guidance usually makes unwanted artifacts go away.  The higher that number goes, the more assertive the AI is about what you want, resulting in a sharper image.  You can now understand why we hid this in the advanced menu.  This is a delicate setting for people with the patience to master it.
• When you think about Steps, think about how your printer works.  The more passes it does, the better the quality is. But if your printer does too many passes on too many instructions, things end up blotchy. So while increasing the step count can give you a more detailed image, writing a clear prompt at 30 steps can also get you there with less wait time.

High Def and Facelift Upscaling Explained

Tons of tools are available after the image is generated:

• HighDef doubles the pixels of the image, similar to anti-aliasing
• Facelift 4x’s the pixels and also applies a face restoration, similar to the beauty mode on your phone
• More will give you five more variations of the same picture, with minor changes
• Remix lets you re-prompt into another base model, destroying the image and making a new one
• Edit is a chat editing program, where you can ask questions like “what if her hair was green?”.  Important: Edit is experimental and only for low-resolution use, it uses its own AI model and doesn’t work within the concepts system
• Inpaint is generative fill, you can literally paint on the photo with your finger or mouse and re-prompt into the mask

Deleting your own images

Quick Flag Method

At the bottom of the page, there’s a FLAG button.  Hit that and nothing will happen, but in a few minutes the image will be gone.  We just added this feature and will make it easier to use in the next software update.

Full File Manager Method 

On the top right of the menu, you’ll find the “Your Archive” button.  This is a full file manager. You can create folders, move files, set favorites, and batch delete to keep only the pictures you want.  You can also download multiple images at once.  It’s quite easy.

Just select the images you want, and action buttons will appear at the bottom.  Use the action buttons to control your files. That’s all there is to it.

Your files are on the cloud, so you’ll never lose your prompts or images, and you can resume from any device.  Just keep your login (email) safe, and always log out or someone can control your account.

Click the Organize button to delete or move your files.

Requesting an image takedown from this site

Send us a message on Reddit

There are two basic ways to use poses:  Use positive prompts to describe the scene, and using a concept that already strongly influences the pose.  Some “base models” are much better at poses than others, especially if more than 1 person is in the scene.

If you’re going for realistic sex, the first choice of many is to set the base model to URPM 1.3

Next, choose a pose.

Use the advanced editor to find the pose you want.  Make sure NSFW is turned on, or you won’t see everything that we offer.

Going into the Poses or Porn section, you’re going to find an incredible amount of options.

Even the ones that look like anime can work with realistic people, or can first be rendered as anime and remixed into reality later, or vice versa. 

Let’s pick a random one just as an example:

Be sure to read the description of each model because they have a ton of extra options sometimes.

That’s really all there is to it.  If the pose isn’t working, try adjusting the weight, or moving the pose instructions to the pose of the prompt.

Troubleshooting tip:  Remember, its never a good idea to add more than 3 loras to 1 base model, and sometimes these concepts are too strong and don’t well together. If you’re getting blue blotches or jaggy people, try lowering the weight of the lora, or using less.  Build your image slowly step by step for best results.

What is Hentai, anyway?

First, a quick lesson: The word “hen” in Japanese means unusual, and “tai” means want or desire, whereas “anime” literally just means cartoons. Hentai is literally referring to the kinkier stuff, whereas Anime just means cartoons.

We have SFW, NSFW anime, like this effect that makes clothes explode:

… and then there’s NSFW “Hentai”, the much harder stuff.

The Easy Way: Recipes

The basic way to get started is to choose an anime recipe.  Just scroll sideways on our recipes menu and you’ll find them.  There’s quite a few.

Advanced Mode:

If you’re going with advanced mode, Base Models like Hassaku or AOHard are a good place to start.  For more, look for the Hentai Tag in the concepts picker.

Not all anime base models can do hentai well, so that’s why there are two tags to guide you.

Creating AI anime is easier than you think, but when you start to explore the advanced options, you’re going to be blown away. Without having to type a word, you can get into over a thousand of these poses, characters, and effects.

We also have literally hundreds of anime effects and poses, far to many to name here.  Hit the models area to explore them, and create an account to render them!

For advanced users that want more control, there’s more power under the hood. You can add additional commands to your prompt to activate various settings.

WHAT IS THE WORD LIMIT?

Stable Diffusion stops paying attention after 75 tokens, and the word order matters. We don’t limit how much you can enter in a prompt, but it does stop mattering after the 75 token mark. To complicate matters, every model responds differently to long prompts.

For beginners, you’ll be surprised how well <level4> understands your prompts. For experts, <deliberate2> is famous for how well it can understand a long prompt — but you’ll have to work harder at balancing the weights and word order. Compare the base models and find the one that works best for you. You’re on an unlimited plan after all, so don’t worry about wasting renders.

To make the most of the word limit, get your most important ideas at the start of the prompt, and save on tokens by using Textual Inversions so you don’t lose tokens to negative prompts.

IMAGE COUNT

When testing, you can add /images:1 to your prompt to speed up trial and error, so you’re not waiting for 5 images to come back. Use the presets in the advanced editor to get up to 9 images back at once in low steps.

Pictured bottom right: steps and image count easy presets:

WHAT ARE STEPS?

Steps means fidelity. The more steps, the less noise the image will contain. An average number of steps is 25. Our system is set to 50 by default.  You can speed up your image by lowering steps, and increase detail by increasing steps. The upper limit is 200, which is nice but very slow. There are better ways to get a nicer image than always working in the slowest step count, like using facelift.

The number of images, steps, and resolution add up to higher memory on the graphics card on the server, which has a finite limit. High numbers of images at high resolution will likely crash the render. This is why we offer safe presets in the advanced pulldown menus.

The values and their meanings:

wayless – 15 steps
less – 30 steps
normal – 50 steps
more – 100 steps
waymore – 200 steps

WHAT IS A SAMPLER?

For 99% of people, you’ll never have to change the sampler. This stuff is squarely into the territory of AI inside baseball. But it is interesting, and we love to teach you about it.

Samplers are a part of the AI diffusion software, a modular piece of software that plays a part in creating images. There are thousands of samplers available on the Internet as open source downloadable projects. To make an oversimplification, each sampler uses a different mathematical approach to arrive at the final image.

The differences between samplers, to be frank, can be very minor to the average person. But for the very discerning creator, you may find you have a favorite. There are real differences to detail, speed and performance tradeoffs when taking into account the number of steps, as each sampler reacts to step counts differently. For those that study samplers carefully, you’ll find strong vocal opinions on which one is best all around Stable Diffusion enthusiast forums.

Our system has many samplers installed and selects one at random when you are creating your image. We do this to give you maximum variety. Only 1 sampler is at play at any given time when creating your image.

How do I know which sampler was used for my image?

After your image was created, click the prompt button. You can see which one was selected by clicking the prompt details of the finished image. For Telegram users, this appears in the output as well.

Pictured: A sampler called “k_euler_a” played a part in creating this image  

How to override and manually select a sampler

That said, you can choose your favorite one and override the random selection.  To override the sampler, copy this command at the start of your positive prompt

/sampler:upms

The cheat sheet to sampler codes:

SEED and CCF (classifier free guidance)

As you saw in the image above, and is repeated here below, you can also specify a random seed value to “lock” a result. A seed is just a random number and doesn’t have any specific meaning, it is just a means to repeat. It isn’t like a photo ID in a database.  Learn how to use seeds in this tutorial.

Example:

/seed:112697

Clicking on a prompt shows what random seed was used, and also how much guidance an image has. This picture also has a guidance of 8.41, which we’ll cover in detail below.

Classifier Free Guidance, also known as CFG or just Guidance, is the parameter that determines how loosely or how strictly to follow your prompt instructions. A value of zero is too much freedom, producing creepy paper-mache images from another world. A maximum guidance of 20 is too literal, where every word MUST happen, resulting in jaggies and distortion.

Unless you select guidance in the Advanced GUI, we’ll assign a random value.  See where it says “Artistic:8” below?  That means a guidance of 8 — not too sharp, not too foggy. It’s a safe number for illustrated works or soft photos.

Random Guidance

Our system selects a random guidance value between 7 and 12 to give you a decent amount of accuracy and softness, as things look more in focus the higher we go up (at risk of blue boxes or jaggies appearing).  You can specify an infinite number of levels of granularity for guidance like 8.4123317823719237 but most people stick to one decimal or no decimals.

To override guidance, type it like this:

/guidance:12

ADVANCED “MORE”

Our “more” tool is just a quick shortcut to increasing guidance by locking the same seed value, that’s all it does. You can take matters into your own hands using Remix instead and get exactly what you want.

REMIX AND STRENGTH

Strength is a parameter that is only applicable when there is an image as an input. Strength determines how much influence the existing image has over the new prompt.

The average prompt has no use for strength, because strength isn’t used in text-to-image. This is an important and overlooked fact that many people mistake.

When strength is added to a text to image prompt, it is treated like a token or ignored. Something may change, but only for that reason.

Instructions:

1. Click on one of your favorite rendered images of a person
2. Use the remix command
3. And paste this into the remix command*

/strength:0.3 /guidance:7.0 /size:1200x1200 /images:2 <fc-portrait> [[<verybad-negative:-1><unspeakable-horrors:-1> <easy-negative:-1><bad-hands-v5:-1><bad-prompt:-1> <boring-e621:-2>]] <photon>

Translation: Keep the image as is, but boost it with this LoRA all of these negatives at a guidance of 7, and this higher resolution.  You should get a higher quality image than your 512×512 typical image.

*for Telegram users, right click and “reply” to an image as if you’re going to talk to a person, and write /remix (paste the prompt here)

Q: Why doesn’t my image look realistic?

• Use recipes to get quick realistic results. You can also go manual and do it, but don’t forget to use a strong negative. This guide will teach you how.
• Make sure your base model or recipe is set to realistic.  A recipe like Libreal will make your images look like they were taken with an iPhone, or a base model like <lazy> or <realvis51>
• Add keywords for things like sweat, skin pores, freckles, pock marks
• Choose your words carefully.  Don’t mention anything in your prompt that could be misinterpreted as fine art.  For example, the word “masterpiece” can refer to paintings or a type of concept art popular in anime, so the AI thinks you want illustrations
• Click advanced in the editor, and increase the guidance.
• You should also study how to use negative prompts.  For example, a positive prompt for “realistic photograph” is made stronger with a negative prompt for “[[anime, illustration, 3d, cgi]]” because we are saying we definitely don’t want any of those things, with strong brackets.

Q: Why are my images small?

• We send rough drafts in 512×512 but you can get near 4k quality in two steps:
• Use highdef to double the pixels, and then….
• Use 4X facelift to boost the pixels.  Step by Step Tutorial

Q: Why am I getting blue blocks or static noise

• Use the advanced menu and lower the guidance to 4, then work your way up. Long prompts and high guidance are the most common culprit.
• Reduce the number of negative prompt brackets
• Simplify your prompt.  Put the most important words at the front and use less negative and positive prompt
• If you copied a prompt from the internet, check that the lora names match our system and that they negative prompts and weights are properly terminated
• Make sure you don’t have keywords that might contradict each other
• Try reducing the number of loras
• Make sure the lora is compatible with the base model, not all of them are
• Reduce all your lora weights to 0.1 and work your way up

Q: Can you help me with my prompt?

• Yes actually.  send us an email and we’ll help you one-on-one.