File: ComfyUI_03321_.png (732 KB, 1280x1280)

732 KB PNG

I've got a weird prompting issue, maybe someone can help.

File: ComfyUI_03320_.png (1.65 MB, 1280x1280)

1.65 MB PNG

>>3090610

So they only prompting difference between these two is adding "large pectorals" to it, but it gets much more... smooth and detailed? I wanted to keep the flatness of the first one. I've tried multiples variations of big/large and pecs/pectorals/chest, all of them resulted in this change.

Here's my prompt:

hiromi_higuruma, 1boy, solo, male focus, adult male, male, short hair, black hair, white collared shirt, open shirt, white briefs, black necktie, loose necktie, bulge, cigarette, standing, skinny, muscular, tall, looking away, BREAK white background, simple background, BREAK simple, cartoon, (beitemian:0.7), (idoraad:0.9),

neg:

lowres, worst quality, bad quality, bad anatomy, signature, username, logo, watermark, artist name, old, bad hands, weibo watermark, patreon username, bad face, blurry, messy background, disfigured, extra limbs, watermark, 3d render, realistic, colored skin, ui

Using HomoSimile v4 and this lora:

https://civitai.red/models/2449284/hiromi-higuruma?modelVersionId=2754009

File: ComfyUI_03350_.png (665 KB, 1280x1280)

665 KB PNG

>>3090610
>>3090611

So I'm kind of dumb, but still confused lol. The flatness of the first one is because for some reason the (beitemian:0.7) isn't triggering even though it's included in the prompt. I should've realized that based off the art difference.

Still not sure why it's not prompting but whatever I guess?

File: ComfyUI_03447_.png (721 KB, 1280x1280)

721 KB PNG

Which version do you all like more, this one?

File: ComfyUI_03464_.png (1.67 MB, 1280x1280)

1.67 MB PNG

>>3090615

Or this one?

>>3090615
Both are hot, just my type, but I think I like this one more.

File: 00117-3304760521.jpg (1.45 MB, 1296x1728)

1.45 MB JPG

>>3090612
>>3090611

Full disclosure: I don't pretend to know everything...But:

You are pulling the model into a lot of different directions with the Lora, 2x BREAK commands, and the fact that Beitemian is at 0.7 strength. I am pretty sure that is why the beitemian is not triggering.

Just like if you prompt in opposite directions (describe a characters face, then prompt for the head to be out of frame or facing away etc...) you are making the model choose between prompts that are in opposition to each other. The model doesn't inherently know to "mix" two styles. It just knows you want one style more then the other, so it is skipping one. Also, keep in mind that BREAK is very strong, and I don't think you are using it in the intended way. Try getting rid of the BREAKS altogether and see if that helps, would be my first step.

File: snow ball earth.png (2.88 MB, 1917x2122)

2.88 MB PNG

something about the guys from Snowball Earth anime. the big one is the bottom and he's teaching tetsuya the smaller guy what sex is.

File: ComfyUI_03545_.png (928 KB, 1280x1280)

928 KB PNG

>>3090619

Interesting, I'll keep that in mind and try playing around with it. I probably don't need all the breaks, that's for sure. Never really understood their purpose lol.

For what it's worth, I've been using some variation of this prompt for almost a year and this is the first time it's happened. HomoSimile XL v4 can definitely mix two artists styles.

File: 00018-1060439272.jpg (1.22 MB, 1296x1728)

1.22 MB JPG

>>3090621

Both SD 1.5 and SDXL models are famous for being able to mix. But that mixing is an unintended consequence of the model being imprecise. In fact, if you read what people are saying about "newer" models like flux, they complain it can't mix the way older models can.

I do know that after a certain point SDXL will just straight up ignore prompts once it hits a certain limit, which, again, is what is happening to you (I bet). The "padding" that BREAK commands put in will make you hit that limit faster. Try something simpler like:


hiromi_higuruma, 1boy, solo, male focus, adult, short hair, black hair, white collared shirt, open shirt, white briefs, black necktie, loose necktie, bulge, cigarette, standing, skinny, muscular, tall, looking away, plain white background, (beitemian:0.7), (idoraad:0.9)

File: ComfyUI_03583_.png (627 KB, 1280x1280)

627 KB PNG

>>3090622

Yeah, I'll definitely play around with skipping the breaks. It was just weird it wasn't mixing before adding "large pectorals". Just kind of weird. The mix works fine though, both of these were mixed too:

>>3090615
>>3090616

I was just trying to figure out a flatter/less soft gen. But I'll definitely try it your way! Thanks for all the tips and information! <3

File: 1777145823008092.jpg (565 KB, 4096x3072)

565 KB JPG

>>3090593
I've updated and polished the LoRA for my draenei character
Care to give it a try?
>https://files.catbox.moe/mefsg6.safetensors
the trigger word should be "blue_draenei"

File: Sholmes.jpg (812 KB, 1536x1536)

812 KB JPG

Slowly transitioning from team Ryunosuke x Kazuma to team Ryunosuke x Sholmes

File: 09-33-49-706.jpg (2.62 MB, 2000x2666)

2.62 MB JPG

does anyone know how this is achieved? multiple characters and complex composition

File: Sholmes x Ryunosuke.jpg (806 KB, 1536x1536)

806 KB JPG

>>3090632
Sketch out your characters and composition and use img2img. And/or use NovelAI's character boxes/positioning feature.

>>3090628
why not post the uncensored version?

File: ComfyUI_temp_oxnff_00001_.jpg (659 KB, 4096x3072)

659 KB JPG

>>3090635
I was just reusing something I used in a blue board
here you go

>>3090645
Prompt?

>>3090636
Thanks m8, I'd actually appreciate if you posted more. Love the blue ayy goats.

File: 00007-931982423.png (3.15 MB, 1160x2184)

3.15 MB PNG

AI still struggle with text in anima, any tips?

>>3090668
>any tips?
pay for a gpu cluster and fine-tune it, it isnt fully trained, and even when it is I have no idea how capable the model can be with text

>>3090682
but, for whats it worth, in other models (Flux klein) I had better luck when the text in the prompt was in full caps

File: 00025-4022314459.png (3.08 MB, 1608x2144)

3.08 MB PNG

>>3090683
That worked, it's a lot better. Thanks!

File: ComfyUI_03369_.png (750 KB, 1280x1280)

750 KB PNG

>>3090622
>>3090619

You use Stability Matrix right? Considering switching away from Comfy with their new investment/worried about their future. Any recommendations? I used A1111 until a few months back when I nuked it somehow messing with extensions. Something simple, private and easy to install would be dope.

File: 00053-219522897.jpg (1.16 MB, 1728x1296)

1.16 MB JPG

>>3090726
Yes, I have always used Stability Matrix, and I currently use Forge-Neo. Forge-Neo is based on the old A1111 UI and functions pretty similarly. If you liked A1111 you will be right at home.

There is a regular "Forge" branch that you can ignore, it is no longer updated. Forge-Neo is active and updating all the time. It can handle Flux/Anima/Wan/Zimage and video generation, though I personally don't bother to make videos.

A couple of tips for ForgeNeo:

There is a "UI Preset" box in the very top left corner. Self explanatory for the most part when you want to change it, but remember that Pony/Illustrious are all build from the old SDXL, so choose "XL".

Clip skip setting is hidden unless you dig through the full settings list. It defaults to 1, which is fine for Noob/Illustrious, but if you are using old SD or Pony, you may want to dig it out and set it so the option will appear on the main screen.

The Only extensions I use are:

"A1111-sd-WebUI-Tagcomplete" (a autocomplete for the prompt box that automatically pulls the tags from Danbooru (or elsewhere, you can pick) can really help you prompt if you know exactly the word the checkpoint was trained on.

"Adetailer": I don't remember how it works in comfy, but you will need to download this extension and different Adetailer detector models for it to work. I only use "Anzhc Face seg 640 v2 y8n.pt" (for face) & "CockAndBallYolo8x.pt" for penis. Although the penis one works best if you build a separate prompt, both positive and negative, which can be a pain.

"sd-forge-extra-sampler" & "sd-webui-agent-scheduler": They add newer Sampling Methods and Scheduling Types. I prefer the 2 combos: Euler SMEA Dy Negative with SGM Uniform ; and "Extended Reverse-Time SDE" with Simple. Keep in mind with Hires.fix you will need to change the sampler method to something basic (Like Euler), as the newer ones don't play well. That goes for "Adetailer" as well.

File: 00109-1306352258.jpg (1.05 MB, 1728x1296)

1.05 MB JPG

>>3090790
This is very good. Was it made using Anima?

File: 00061-2146637746.png (2.88 MB, 1248x2144)

2.88 MB PNG

>>3090816
Yeah, animaika merge, I'm testing lately the different mixes and it's pretty fun. Hires fix works now on neo with it also.

File: 00087-2765709578.jpg (1.14 MB, 1728x1296)

1.14 MB JPG

This board is so slow

File: 00092-1735149645.jpg (1.32 MB, 1296x1728)

1.32 MB JPG

Just posting some random stuff to give people new things to look at

File: 00093-1735149646.jpg (1.28 MB, 1296x1728)

1.28 MB JPG

I mean, maybe not random, it is all Duo, but still...

File: 00108-2048017563.jpg (1.36 MB, 1536x1536)

1.36 MB JPG

Just had a hankering for some nice testicles this morning.

File: 00105-1769347566.jpg (1.31 MB, 1536x1536)

1.31 MB JPG

Hope your day is a little better after seeing them

File: ComfyUI_06285_.png (1.78 MB, 1280x1280)

1.78 MB PNG

>>3090880
>>3090881
>>3090882
>>3090883
>>3090884

Love Duo and his balls. Yeah, this board has slowed down a lot. I used to post a lot, but I was worried I was posting too much lol. I'll ramp back up the next few days.

File: 00113-1698577855.jpg (1.32 MB, 1728x1296)

1.32 MB JPG

>>3090886
Yeah, I feel the same way, I don't want to dominate the whole thing, and I like to see what others are coming up with.

And I don't just mean the AI posting slowed down. I mean the entirety of /Y/ itself.

I remember back in the day checking for the latest leaked Sakimichan (The 9s series was one of the first pieces I ever saw, and I loved it so much) and Bludwing uploads. When Camp Buddy first came out and it was all over the place. All the different google drives different people would maintain of different artists.

Oh well...

File: 4195522201.jpg (667 KB, 1560x2080)

667 KB JPG

Considering how little goes on on this board you'd be doing it a favor by making it your own blog/image dumpster, I wouldn't really worry about it

File: 00051-683138177.jpg (1.07 MB, 1728x1296)

1.07 MB JPG

>>3090888
You got a point really. But I only have AI images to share, and I don't really want to dump them in non-ai threads. Maybe not that big a deal these days, but I do like to stay on topic

>>3090889
yeah I meant on this thread not the whole board kek

File: 00060-1830263447.jpg (1.27 MB, 1728x1296)

1.27 MB JPG

>>3090890
LOL maybe I will go back to uploading 5 a day or so just cuz.

File: 00007-2333858515.png (3.49 MB, 1248x2144)

3.49 MB PNG

>>3090892
https://anima.mooshieblob.com/

If anyone is also using anima and it's interested

File: 00026-1298346238.png (3.44 MB, 2144x1248)

3.44 MB PNG

>>3090895

>>3090891
Please do

File: gen.png (1.52 MB, 1280x1280)

1.52 MB PNG

>>3090888
Agreed, the more the better
>>3090775
Anyone have anything else of this variety? Perturbed young men, perhaps restrained, in a basement /dungeon or otherwise?

File: 00027-1916889964.jpg (1.03 MB, 1296x1728)

1.03 MB JPG

>>3090911
Some experimental stuff. Civitai doesn't allow noncon, and dungeon stuff isnt really a kink of mine

File: 00026-3793714816.jpg (1018 KB, 1296x1728)

1018 KB JPG

>>3090911

File: 00177-1492019120.png (2.73 MB, 1296x1728)

2.73 MB PNG

>>3090911

>>3090916
Much appreciated. I can only guess at the intricate prompt that produces something like that...

File: image(14).jpg (211 KB, 784x1168)

211 KB JPG

>>3090911

File: 00068-2016366366.png (1.25 MB, 1536x896)

1.25 MB PNG

Sorry for the spam

File: 00071-3537516131.png (1.42 MB, 896x1536)

1.42 MB PNG

>>3090938

File: 00077-1083750125.png (1.22 MB, 1536x896)

1.22 MB PNG

>>3090939

File: 00065-1668176677.png (1.38 MB, 1536x896)

1.38 MB PNG

>>3090941

File: 00089-2678684178.png (1.23 MB, 896x1536)

1.23 MB PNG

>>3090942

>>3090938
>>3090939
>>3090940
>>3090941
>>3090942
Lt. Surge especially in the style of Dizdoodz...

File: 00081-4069142320.png (2.8 MB, 1536x1920)

2.8 MB PNG

Posting some generations I did a couple of years ago when Pony v6 first came out. What a wild time, everyone was having so much fun with anime porn, really seeing the potential.

File: 00196-871915196.png (1.97 MB, 1296x1728)

1.97 MB PNG

These are actually repost I think, but again from a year and a half ago I think. Never uploaded them to Civitai either, a lot of them contain the "teenager" tag, or "school" tag, which are pretty much instantly rejected

File: 00266-2583264012.png (2.47 MB, 1536x1536)

2.47 MB PNG

The newer models are more "precise", but that precision takes some of the magic away. The older models were able to just have some spontaneity to them that is lacking nowadays.

File: 00249-3740876410.png (2.63 MB, 1536x1536)

2.63 MB PNG

I might go back to Pony one of these days, try some of the newer techniques for prompting I have learned over the years. But I don't think that I could ever really recapture the particular magic again.

File: 00099-1637248789.png (2.04 MB, 1536x1536)

2.04 MB PNG

Wish that Pony 7 wasn't such a failure, but Anima looks like it will be a lot of fun, if not so chaotic.

>>3090955
which model do you mean when you say Anima?is this Anima?

If someone can also help out with the way to see the prompts that would be great - unless 4chan does just strip them out.

File: 00032-2976622544.png (1.18 MB, 896x1536)

1.18 MB PNG

>>3090945
I tested it, not an artist I follow or know well, if you have an idea for a scene maybe I'll try it for fun

Wanting to replicate the Kishimoto style (plenty of pics on rule34 are doing that). I'm so confused by all the tools, very new to ai. How should I proceed? Thank you

File: anima_00094.png (1.2 MB, 1216x832)

1.2 MB PNG

>>3090962
the easiest way is to get 10 images that best represent the style and make a text file for each of the same name (make sure these names aren't crazy with weird characters or punctuation), then manually describe each image in the files. then you just put the folder into the oven (use a lora trainer) at a high learning rate and rank and wait until they're baked

good loras take effort and time but if you just want the style of something you can brute force it with the above method

do both the training and generating on anima https://huggingface.co/circlestone-labs/Anima

File: 5772a0ae-0eb7-4fe7-aba6-9(...).jpg (247 KB, 1216x832)

247 KB JPG

>>3090956
https://civitai.red/models/2458426/anima-official

Anima is a new image generation model currently in open testing phase that shows a great deal of promise. The individual posting the more complex scenes with captions is using it right now (for example): >>3090790

Anima does not use the "Stable Diffusion XL (SDXL) base, it is newer architecture. Both "Pony" and "NoobAI/Illustrious" Models are actually retraining done on the original SDXL base released a couple of years ago. They used so many millions of new pictures (anime porn from danbooru and e621) to retrain SDXL that those 2 forks are basically new things.

Anima is very much focusing on better "prompt understanding" (using newer text encoders), unlike sdxl that can get a little soft. That means (theoretically" you could just type to anima: "I want a beautiful explicit picture of link from breath of the wild being fucked Bokoblins, the master sword strewn carelessly in the tall grass as link grits his teeth and bends over taking the Bokoblins large cocks up his ass, the caption above reads: JUST ANOTHER DAY IN HYRULE

(actual picture from this prompt included; negative: worst quality, low quality, score_1, score_2, score_3, artist name)

File: 00203-4274476352.png (1.87 MB, 1296x1728)

1.87 MB PNG

>>3090956
As for the prompts, 4chan does strip out all metadata that you could read at home (using PNG Info for Forge-Neo, not sure about comfy). If there is a specific picture you like, it can be uploaded to catbox with data intact so you can read it at home, or sometimes it can just be copy and pasted here with the picture. Both are a pain in the butt though so not every picture can have a full prompt.

However, if you browse CivitAI, there are all kinds of yaoi pictures, although things like rape, bestiality, shota are not allowed (not that they are necessarily allowed here either). Most everyone has generation data fully shown for every picture if you click on it.

https://civitai.red/user/gtgiygus is my profile if you want to take a look (you will need a free account).

You are also free to ask for advice on specific prompting tips here or at CivitAI.

>>3090951
>>3090952
>>3090954
>>3090955
You've mastered prompting for the 'twink' physique of late teens impressively. I haven't run any models since the early SDXL days but I always struggled to strike the middle ground between adult and (unpleasantly) young appearance. Beautiful work

File: tmpjsmy04x0.png (1.67 MB, 1280x1280)

1.67 MB PNG

>>3090911

lol that's one of mine haha, i've got a bunch more from this set, i'll zip em and share them later.

File: tmpaly7zns_.png (1.52 MB, 1280x1280)

1.52 MB PNG

gtgiygus here posted a rex pic that made it look like he had ribbed underwear due to his body suit and i loved the ribbed aspect lol

File: tmpr46d0jcw.png (1.49 MB, 1280x1280)

1.49 MB PNG

https://files.catbox.moe/ydhqrv.zip

File: 00106-3776040597.png (2.65 MB, 1536x1536)

2.65 MB PNG

>>3090968
Positive prompts: "teenager", "19 years old" (or 16 years old if no censorship), "slim body" (You can add "toned male" or "pectorals" to give back some muscular definition).

Negative Prompts: "facial hair", "mature", (adult if no censorship). Avoid "young" "shota and/or child" (for obvious reasons). "makeup" in negative can push it back to more masculine character if it drifts over to female without compromising the age. If you put "female, women, girl" it will add age and heft to the character while getting rid of feminine traits.

Usually that puts the age right in solid teenager without making it 12 years old.

>>3091007
isn't there a board for trans? /d/ or something?

>>3091010
Don't be rude. It's a genderbent girl, so it's 100% male.

File: 00007-2855393465.png (3.02 MB, 2144x1248)

3.02 MB PNG

Fire emblem men bonding after a battle

>>3091017

File: 00013-1216306293.png (3.5 MB, 2144x1248)

3.5 MB PNG

>>3091019
Forgot the pic somehow D:

>>3090979
Much obliged
>>3090977
Love that one, maybe the best in the set

>>3091017
>>3091020
most of their faces aren't familiar to me unless I recalling the hair colors. but nice settings of orgies that I always imagine in FE. :DD

File: tmpcrc9z5ly.png (1.63 MB, 1280x1280)

1.63 MB PNG

>>3091030

happy to share! :3

File: 00104-2100805774.png (2.41 MB, 1296x1728)

2.41 MB PNG

Back again with some older stuff to help liven up the board, though I see that there are some new pictures up! Always good to see.

File: 00140-1322841950.png (2.57 MB, 1728x1296)

2.57 MB PNG

This particular post was one of my most upvoted of all. Not really sure why everyone liked it, but that's why I am not a real artist, because I don't know what people like...

File: 00135-2166892755.png (2.66 MB, 1296x1728)

2.66 MB PNG

I always have a hard time with femboys. I want them to be masculine but, like, trying desperately to be feminine? I want to see the psychological distress in their presentation if that make sense...

File: 00157-1080935503.png (2.51 MB, 1296x1728)

2.51 MB PNG

I only uploaded 40 pictures over two posts, you can take a look here if you want:

https://civitai.red/posts/8783338

https://civitai.red/posts/8824703

File: 00173-2364135106.png (2.64 MB, 1536x1536)

2.64 MB PNG

This particular Lora did a lot of heavy lifting in terms of how the cute androgynous face came out. Vixon had so many good Loras for Pony mostly, but there are a few I use even today for illustrious. Should take a loot at their stuff if you want to try some experimentation.

File: 00082-1468670022.png (2.56 MB, 1296x1728)

2.56 MB PNG

>>3091057
oops forgot link...https://civitai.red/models/888213/vixons-pony-styles-detailed?modelVersionId=486749

File: Gemini_Generated_Image_6v(...).png (1.34 MB, 768x1376)

1.34 MB PNG

made using gemmy with some brute-force method :DD

File: Gemini_Generated_Image_qp(...).png (1.35 MB, 768x1376)

1.35 MB PNG

>>3091078
the closest as it gets to depict benis.

File: 051943.png (820 KB, 832x1056)

820 KB PNG

impossible to gen homri without it being a black cock >:(

File: 00004-3446810630.png (2.05 MB, 1248x2144)

2.05 MB PNG

Anima 1.0 released, any of you have tried it yet?

File: 00001.jpg (532 KB, 824x1000)

532 KB JPG

>>3091093
I'm liking it so far, excited that there's going to be models and loras on this base. I don't have enough of a gpu to really take advantage of the resolution without it being slow, and pretty much every addon feature in forge neo is broken for me, but even without added tools genning seems as good as its ever been.

File: 00019-295636005.jpg (1.12 MB, 1296x1728)

1.12 MB JPG

Anime is neat. Will have to play around to try and get a good style. I really loved the more realistic styles I was able to produce with Illustrious Lora. I also hate score prompting, but things will change fast.

Positive: masterpiece, best quality, score_7, final fantasy vii advent children, realistic style, male focus, 1boy, solo, kadaj, medium hair, silver hair, green eyes, teenager, 16 years old, male masturbation, masturbating, erection, testicles, glowing green precum drip, seductive, confident, looking at viewer, black jacket, black bodysuit, pants pull, leaning back against wall, outdoors, night, moon, cityscape, science fiction, motion lines, male pubic hair, transparent speech bubble saying "FREE MAKO ENERGY"

Negative: mature, facial hair, large erection, long erection, worst quality, low quality, score_1, score_2, score_3, artist name, signature, sketch, unfinished, flat color

ER SDE/Simple ; 40 steps ; Hires.fix Upscaler:8xNMKDSuperscale_150000G ; Hires steps: 18 ; Denoising strength: 0.45 ; 864x1152 ; upscale by: 1.5

File: 00021-1329615025.jpg (1.01 MB, 1296x1728)

1.01 MB JPG

Same settings as >>3091099 but using natural language:

Positive: a beautiful masterpiece illustration of kadaj from final fantasy vii advent children, the illustration is done in a realistic style, the illustration is male focus of a 16 year old teenager, kadaj has shoulder length silver hair with dimly soft glowing green eyes, kadaj is looking at the viewer with a seductive and confident expression, he is wearing a black bodysuit unzipped allowing his erection testicles and male pubic hair out, kadaj is masturbating his erection and there is dimly soft glowing green precum dripping from the head of his erection, kadaj is leaning back against a wall outdoors at night, a science fiction cityscape in the background, kadaj is saying with a transparent speech bubble "FREE MAKO ENERGY"

Negative prompt is the same

File: 00044-3762588808.jpg (1.58 MB, 1840x1840)

1.58 MB JPG

Tired pushing Anima a little harder to see what it would do. This picture was done at 1536*1536 with 1.2 Upscale, everything else the same:

Positive: a beautiful masterpiece nsfw illustration of 2 boys Kadaj and Cloud Strife from Final Fantasy VII Advent Children, the illustration is done in a realistic style, the illustration is a male focus work of a toned 16 year old teenage boy, Kadaj has shoulder length silver hair with dimly soft glowing green eyes, Kadaj is looking at the viewer with a seductive and confident expression, Kadaj is wearing an unzipped black bodysuit allowing his erection male pubic hair and testicles out, Kadaj is presenting his erection and there is a dimly soft glowing green precum dripping from his erection, Kadaj is saying with a transparent speech bubble "FREE MAKO ENERGY", Cloud Strife has blond hair and blue eyes showing only his upper body near Kadaj, Cloud Strife is staring at Kadaj's erection with a memorized and worried expression with a open mouth and saliva dripping down, Cloud Strife is caressing Kadaj's testicles, the scene is showing imminent fellatio, Cloud Strife is wearing a dark blue ribbed sleeveless turtleneck, Kadaj is ruffling Cloud Strife's hair, a science fiction cityscape in the background at night

Overall it is pretty good. The text seems to generate worse at higher resolutions, but prompt adherence is decent. I particularly like that it was able to generate a well trained in character (Cloud) side by side with a far lesser trained character (Kadaj). Although I did use an old trick of describing the lesser trained character in much more detail and they are both from the same show. I will also point out that the uploaded picture is cherrypicked, but I only had to do 4 pictures to get one that was pretty good. Took about 4.15 minutes with 16 gigs Vram, including Hires.fix and ADetailer face per picture.

File: 00045-855975101.jpg (1.56 MB, 1840x1840)

1.56 MB JPG

Just to finish the set off:

Positive: a beautiful masterpiece nsfw illustration of 2 boys Kadaj and Cloud Strife from Final Fantasy VII Advent Children, the illustration is done in a realistic style, the illustration is a male focus work of a toned 16 year old teenage boy, Kadaj has shoulder length silver hair with dimly soft glowing green eyes, Kadaj is looking at the viewer wincing and biting own lip, Kadaj is wearing an unzipped black bodysuit allowing his erection male pubic hair and testicles out, Kadaj is saying with a transparent speech bubble "WATCH YOUR TEETH", Cloud Strife has blond hair and blue eyes showing only his upper body near Kadaj, Cloud Strife is sucking Kadaj's erection with a memorized and worried expression with dimly soft glowing green precum dripping down Cloud's chin, Cloud Strife is caressing Kadaj's testicles, the scene is showing deepthroat oral sex and fellatio, Cloud Strife is wearing a dark blue ribbed sleeveless turtleneck, Kadaj is grabbing Cloud Strife's hair, a science fiction cityscape in the background at night

File: file.png (806 KB, 1024x1024)

806 KB PNG

>>3090964
this is remarkable

>>3091085
the training data for anima is bad for a lot of artists

this is why you should avoid loras that use @artist tags if possible, because they'll always be tainted by the bad data

>>3091129
I mean, don't think it has to do with the data, homri just makes black dicks, lots of artists are like that, if they only draw interracial you get interracial when using them

>>3091093
I love these crotch shots, great work.

File: 00010-2377384185.png (2.92 MB, 1344x2304)

2.92 MB PNG

>>3091144
Thanks I don't know how to prompt it very well, it seems random sometimes, this was another attempt. If someone has some advice it will be great.

File: 00084-814503452.jpg (1.32 MB, 1728x1296)

1.32 MB JPG

>>3091152
"from below" ; "urethra" ; "upshorts" ; "crotch focus" ; in addition to flaccid penis/erection, testicles, male pubic hair/body hair...

File: 00009-3394767947.jpg (1.18 MB, 1536x1536)

1.18 MB JPG

Also, sleepy boys are cute...

File: 00015-360784316.jpg (1.17 MB, 1728x1296)

1.17 MB JPG

...and vulnerable

File: IMG_9694.jpg (167 KB, 768x1280)

167 KB JPG

I want to become one of those AI yaoi artists who posts image packs on Patreon and gets paid. Where do I start?