[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/d/ - Hentai/Alternative

Name
Options
Comment
Verification
4chan Pass users can bypass this verification. [Learn More] [Login]
File
  • Please read the Rules and FAQ before posting.
  • There are 53 posters in this thread.

08/21/20New boards added: /vrpg/, /vmg/, /vst/ and /vm/
05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
[Hide] [Show All]


[Advertise on 4chan]


>>
File: 102856241_p0_master1200.jpg (596 KB, 600x1200)
596 KB
596 KB JPG
>>
File: 102287099_p0_master1200.jpg (494 KB, 750x1200)
494 KB
494 KB JPG
>>
File: 102227885_p0.png (531 KB, 512x1024)
531 KB
531 KB PNG
>>
File: FhmPcO4UYAAnKCn.png (471 KB, 512x768)
471 KB
471 KB PNG
>>
File: FhHYJrNVsAASKa7.jpg (172 KB, 1024x1024)
172 KB
172 KB JPG
>>
Any crushing please?
>>
File: 1664832087731762.png (621 KB, 512x1024)
621 KB
621 KB PNG
>>
File: 1664833641802859.png (688 KB, 512x1024)
688 KB
688 KB PNG
>>
File: 1665940839802561.png (927 KB, 704x832)
927 KB
927 KB PNG
>>
File: 1665399039986484.png (1.58 MB, 1284x768)
1.58 MB
1.58 MB PNG
>>
File: 1665227715105987.png (642 KB, 563x768)
642 KB
642 KB PNG
>>
File: 1665086471085247.png (939 KB, 646x768)
939 KB
939 KB PNG
>>
File: 1665010555567340.png (850 KB, 561x768)
850 KB
850 KB PNG
>>
File: 1665010715918667.png (676 KB, 563x768)
676 KB
676 KB PNG
>>
File: 1665248377901100.png (1.49 MB, 844x1152)
1.49 MB
1.49 MB PNG
>>
File: 1665021944534893.jpg (254 KB, 844x1152)
254 KB
254 KB JPG
>>
File: 1665075410660779.png (726 KB, 588x768)
726 KB
726 KB PNG
>>
File: 1664846678475530.png (395 KB, 704x512)
395 KB
395 KB PNG
>>
>>10443570
Oh wow this one is particularly bad. If someone wanted to argue why ai art should be banned this would make a strong case.
>>
File: AI Paintover (Large).png (2.49 MB, 1119x1080)
2.49 MB
2.49 MB PNG
>>
>>>/trash/52389076
Ankha AI ass
>>
>>10443639
this can't be AI it's got a normal amount of toes
>>
File: 1664895322511620.png (526 KB, 512x640)
526 KB
526 KB PNG
>>
File: 1666277949960931.png (483 KB, 672x384)
483 KB
483 KB PNG
>>
Dumping
>>
2
>>
3
>>
Tired of doing captcha
>>
Uh... Model?
>>
File: 102668646_p0.png (2.35 MB, 1024x1536)
2.35 MB
2.35 MB PNG
>>
>>10444673
I painted over it, it only had 4 initially
>>
>>10445130
>Uh... Model?
Pretty clear that most of the pics in this thread are made in NAI, not Stable Diffusion.
>>
File: 1667977267133579.jpg (87 KB, 512x768)
87 KB
87 KB JPG
>>
>>10445496

I think most people just use 'AIdiffusion' as a catchall term for AI-generated art.
>>
>>10445597
He's asking for the model, NAI doesn't let you pick (kinda).
NAI has the furry model and the anthro model, so asking what model they used doesn't really change anything.
>>
What prompts are people using? I'm still having issues with any generated size art.
>>
>>10444946
walk me through the prompt for this one?
>>
>>10445734
try "low_angle_pov cityscape with clouds above macro"
>>
>>10445747

Couldn't tell you i'm afraid, I'm not the original author. Most of these are from prior AI-threads. Maybe OG anon will turn up and help you, but I can't.
>>
File: 102462057_p0.png (2.05 MB, 1024x1536)
2.05 MB
2.05 MB PNG
>>
Does anyone have any good prompts and models to share? I'm running the NAI leak via the voldy guide and can't produce anything like this at all. I feel like I'm missing something.
>>
File: Capture.png (816 KB, 461x693)
816 KB
816 KB PNG
>>
File: 1668662530478693.jpg (306 KB, 1060x1024)
306 KB
306 KB JPG
>>10443201

Here, I did some rudimentary photoshopping to fix some of the most egregious parts since I really like this piece. Hand's still kinda fucked though, and I'm not good at shooping by any standard.
>>
File: 1664775779824751.jpg (357 KB, 640x960)
357 KB
357 KB JPG
>>
>>10446993
looks like you made her wide
>>
art by Artgerm, masterpiece, highest quality, best quality, aerial view, from above, slender giantess female knight walking on a tiny contemporary city, long blonde hair, blue eyes, hair ornament, earrings, high heeled boots, intricate breastplate, (gentle smile:1.2)

Negative prompt: extra arms, extra hands, extra legs, bad hands, deformed hands, loli, ugly face, blurry, jpg artifacts, extra fingers, missing fingers, low quality, cropped, signature, watermark, text, multiple women, stand, pedestal, nsfw

Steps: 28, Sampler: DPM2 Karras, CFG scale: 10, Seed: 2353600557, Size: 512x896, Model hash: 925997e9, Model: Novel AI, Denoising strength: 0.5, First pass size: 0x0
>>
This one also came out well.

art by Artgerm, masterpiece, highest quality, best quality, aerial view, from above, (slender:1.2) giant female knight walking on a mini futuristic city, giantess, long blonde hair, blue eyes, earrings, high heeled boots, intricate blue breastplate with golden filigree, (gentle smile:1.2)

Negative prompt: extra arms, extra hands, extra legs, bad hands, thicc, deformed hands, loli, ugly face, blurry, jpg artifacts, extra fingers, missing fingers, low quality, cropped, signature, watermark, text, multiple women, stand, pedestal, nsfw

Steps: 28, Sampler: DPM2 Karras, CFG scale: 11, Seed: 1956437781, Size: 512x896, Model hash: 925997e9, Model: Novel AI, Denoising strength: 0.5, First pass size: 0x0
>>
>>10447489

Yeah 10%, I prefer them thick. You can just quickly narrow it again in paint.
>>
>>10447622
>>10447850

First one is exceptional, second is alright. The perspective starts to break down when you take her feet into consideration.
>>
>>10447909
Ye, it would be nicer if there was a consistent way to compose a picture. Being able to put together a model and then have the AI draw over it would release the giantess floodgates. Have another.

art by Artgerm, masterpiece, highest quality, best quality, aerial view, from above, (slender:1.2) giant female knight walking on a mini futuristic city, giantess, long blonde hair, blue eyes, earrings, high heeled boots, intricate blue breastplate with golden filigree, (soft smile:1.2)

Negative prompt: extra arms, extra hands, extra legs, bad hands, thicc, hair tied up, deformed hands, loli, ugly face, blurry, jpg artifacts, extra fingers, missing fingers, low quality, cropped, signature, watermark, text, multiple women, stand, pedestal, nsfw

Steps: 28, Sampler: DPM2 Karras, CFG scale: 11, Seed: 2149533373, Size: 512x896, Model hash: 925997e9, Model: Novel AI, Denoising strength: 0.5, First pass size: 0x0
>>
>>10447622
This is what I get with the exact same prompt, seed, and settings
>>
>>10447912
>>10447960

Nice, still seems the best way to handle feet is to try to ignore them though. Still good by AI-standards.
>>
Good thread, I wish my prompts came out half as good as some of these
>>
Giant girls in diapers when?
>>
>>10447960
I tried to reproduce this result but couldn't, even after messing around with .vaes and other settings which aren't reported.

trending on Artstation, masterpiece, highest quality, best quality, aerial view, from above, (slender:1.2) giant female knight walking on a mini cityscape, giantess, long blonde hair, blue eyes, earrings, intricate pink breastplate with filigree, (gentle smile:1.2)

Negative prompt: extra arms, extra hands, extra legs, bad hands, thicc, hair tied up, deformed hands, loli, ugly face, blurry, jpg artifacts, extra fingers, missing fingers, low quality, cropped, signature, watermark, text, multiple women, stand, pedestal, nsfw

Steps: 28, Sampler: DPM2 Karras, CFG scale: 10, Seed: 3142261442, Size: 512x896, Model hash: 925997e9, Model: Novel AI, Denoising strength: 0.5, First pass size: 0x0
>>
>>10448621
where do you get the novel ai model?
>>
>>10448141

NO

Even /d/ has limits.
>>
File: FiH2KtPXoAMcFTj.jpg (1.95 MB, 1200x2000)
1.95 MB
1.95 MB JPG
>>10443186
>>
>>10448933
Not that anon but I'm using the leak and guide from voldy.
>>
File: 06079.png (2.07 MB, 2048x1024)
2.07 MB
2.07 MB PNG
>>10443186
>>
File: 06078.png (1.72 MB, 2048x1024)
1.72 MB
1.72 MB PNG
>>
>>10449229
>>10446241

Really dig these ones. Is it possible you could try a prompt that generates a background such as a cityscape or something smaller in scale relative to the girl?
>>
I'm still trying to figure out how to get a specific character into the ai.
>>
File: FiOsfD6aYAAG8Rn.jpg (1.27 MB, 1024x1536)
1.27 MB
1.27 MB JPG
I'm not ready to let go just yet
>>
File: FeUc6a8WYAMeOpr.jpg (461 KB, 1024x1536)
461 KB
461 KB JPG
>>
File: FgEoErdUAAAOpSQ.jpg (86 KB, 512x768)
86 KB
86 KB JPG
>>
File: FgEoErfVEAIhnOT.jpg (84 KB, 512x768)
84 KB
84 KB JPG
>>
File: FgEoErmUoAA4w9J.jpg (84 KB, 512x768)
84 KB
84 KB JPG
>>
File: FgEoEruVsAAZi3n.jpg (71 KB, 512x768)
71 KB
71 KB JPG
>>
File: FfLqsM2UUAEzTZc.jpg (65 KB, 512x768)
65 KB
65 KB JPG
>>
File: FfLqsM8VQAAVeuc.jpg (86 KB, 512x768)
86 KB
86 KB JPG
>>
File: FiEzfrqaEAEBkSx.jpg (67 KB, 512x1024)
67 KB
67 KB JPG
>>
File: FiFSEmfagAA2_Hz.jpg (98 KB, 768x1024)
98 KB
98 KB JPG
>>
File: FhRYi0jWIAEQpp5.png (716 KB, 563x768)
716 KB
716 KB PNG
>>
>>10450379
>>10450380
>>10450382
>>10450384

HOSHO!!!!

Best Yamato Nadeshiko kancolle girl~!
>>
>>10443186
Retard here. Can somebody give me a qrd on what this is and how I can set it up?
>>
>>10451213

You feed lines of code into a program and make it spit out lewd art after a few iterations.
>>
File: FiNtKR3UUAAgXq2.jpg (79 KB, 512x768)
79 KB
79 KB JPG
>>10451213
>>10451331

Check out Voldy's guide to stable diffusion in Novel AI, which is the program in question. It walks you through the set-up, but it will take a bit of commitment.


Note that there are other AI-programs, Novel is just the main waifu generator I think they even advertised it as such on one occasion.
>>
>>10451213

Oh and Novel isn't free either, but there's a pirated version out there circulating.
>>
File: download (19).png (750 KB, 640x704)
750 KB
750 KB PNG
Been trying to mess with other models than novelai.

The anything model seems promising for this sorta stuff.
>>
File: 00004.jpg (3.96 MB, 3636x4000)
3.96 MB
3.96 MB JPG
>>10451405
>when prompt looks kinda alright so you get the AI to shit out 100 and see how they are overall
>>
>>10451411

A lot of those came out well I'd say.
>>
File: image:343977.png (3.68 MB, 1800x1763)
3.68 MB
3.68 MB PNG
>>
>>10449292
I tried getting creative but no, not really, doesn't work. I feel like there are two totally different NAI models out there.
>>
File: 102903815_p0.png (1.96 MB, 1024x1478)
1.96 MB
1.96 MB PNG
>>
File: 102823082_p0.png (2.41 MB, 1024x1536)
2.41 MB
2.41 MB PNG
>>
>>10447960
You have your CLIP skip/ "Stop At last layers of CLIP model" setting set to 2. That's what the NAI guide instructs you to do in order to replicate NAI outputs.

Change it to 1 if you want to replicate that result.
>>
>>10451576
Yes, I've been using clipskip. Doesn't help with sizey images.
>>
>>10451411
>3rd from the top 5th from the left
You have it separatly?
>>
Share your tech, people, not just the results. For my part I've found that generating an X/Y plot with varying amounts of denoise (I use .5, .6, .7, .8, .85, .9, and 1) produces a general theme with a decent variation of results having the same overall composition. Note that higher levels of denoise tend to "wash" colors together more strongly, which often means that the first listed color in your prompt can predominate.

trending on Artstation, masterpiece, highest quality, best quality, (aerial view:1.2), (from above:1.2), (slender:1.2) giant princess in focus kneeling in mini village, sunrise, petite giantess, long [blue:green:.6] hair, blue eyes, golden tiara, necklace, earrings, intricate pink sundress with ruffles, hair ribbons, (gentle smile:1.2)

Negative prompt: extra arms, extra hands, extra legs, bad hands, thicc, hair tied up, deformed hands, loli, ugly face, blurry, jpg artifacts, extra fingers, missing fingers, low quality, cropped, signature, watermark, text, multiple women, stand, pedestal, nsfw

Steps: 28, Sampler: DPM2 Karras, CFG scale: 10, Seed: 2132696650, Size: 512x896, Model hash: 925997e9, Model: Novel AI, Denoising strength: 0.9, First pass size: 0x0
>>
Remaster of some old images




>>
>>
File: bbsd.jpg (369 KB, 640x704)
369 KB
369 KB JPG
>>10451699
This one?

Eye is off but you could probably clone and shop it around with the other eye.

>>10451729
img2img ?

Swear I've seen an artist with this sorta style before.
>>
>>10451718
>>10451729
I'm going to cum, you faggot. Does that turn you on?
>>
File: Ai wink.jpg (124 KB, 640x677)
124 KB
124 KB JPG
>>10451733
Is this better?
>>
Is there a wiki how to start with this stuff? I bought a new gpu maybe I will give it a try
>>
>>10451733
Ye, prompt for first one

dramatic shadow, ray tracing, ((best quality)), (dynamic lighting), depth of field, ((masterpiece)), 1girl, ((an extremely delicate and beautiful)), beautiful detailed eyes, (sweat), (perfect face), (black hair:1.1), thick thighs, gigantic breasts, hanging breasts, arm clenched hand, side ponytail, giantess, bent over,(embarrassed:1.0),(looking at viewer:1.0), (open mouth:1.1), (school background:1.1), black skirt, white shirt, perfect details, heavy breathing, skindentation, red eyes
Negative prompt: lowres, bad anatomy, ((bad hands)), text, error, missing fingers, extra digit, fewer digits, cropped, normal quality, jpeg artifacts, signature, watermark, blurry, deformed, fat, (bad hands:1.1), extra hands, extra ears, ((monochrome)), (mutated hands:1.2), extra fingers, extra legs, loli,
Steps: 31, CFG scale: 10, Seed: 2876675704, Size: 704x768, Model hash: 1a7df6b8, Denoising strength: 0.65, Clip skip: 2, ENSD: 31337, Mask blur: 4

the second on is a variation of the same prompt
>>
File: lrPgHQl[1].jpg (206 KB, 851x1024)
206 KB
206 KB JPG
>>10451729
Found the original
>>
>>10451718
>>10451729

I thought those pictures (especially the last one >>10451816) looked familiar!
>>
it's a real shame that the leaked novelAI prompt isn't that great at doing giantess like the actual novelAI model is, at least from what i've seen
>>
I combined the anything 3.0 and hentai model using addative.

Here's the prompt for the image I made.
Trying to figure out some better ways for shrink scenarios.

{{masterpiece}}, {{best quality}}, {{ultra-detailed}}, {{Bremerton}}, {{Azur Lane}}, プロンプトはPOV, open mouth, tongue out, macrophilia, looking down from below,{{masterpiece}}, {{best quality}}, {{ultra-detailed}}, {{looking down from below}}, [home background], smile, nude, {{dynamic angle}}, pink hair, macrophilia|POV on table, poker card, light over head
Negative prompt: lowres, bad anatomy, ((bad hands)), text, error, missing fingers, extra digit, fewer digits, cropped, normal quality, jpeg artifacts, signature, watermark, blurry, deformed, fat, (bad hands:1.1), extra hands, extra ears, ((monochrome)), (mutated hands:1.2), extra fingers, extra legs, loli,
Steps: 23, Sampler: Euler a, CFG scale: 7, Seed: 3733700231, Size: 640x960, Model hash: 6569e224, Variation seed: 4271772840, Variation seed strength: 0.25
>>
File: download.png (718 KB, 640x960)
718 KB
718 KB PNG
img2img that one above into this and it works a lot better for other images in the style
>>
>>10452237
>combined the anything 3.0 and hentai model using addative
You combined them using what?

Ooof but anything 3.0 gives great results with minimal effort. Holy FLUX. The eyes were the only reason I added a third statement.
>>
>>10449682
Look up Dreambooth
>>
File: FfLqsMxVQAAPgXu.jpg (84 KB, 512x768)
84 KB
84 KB JPG
>>10451518

Thanks for trying, anon. That was more than I expected.
>>
File: FfLqsMwUcAIWhRs.jpg (90 KB, 512x768)
90 KB
90 KB JPG
>>
File: FiFSFFdaEAElvJd.jpg (151 KB, 896x1216)
151 KB
151 KB JPG
>>
File: FiJuy9eVEAE59gC.jpg (151 KB, 943x1280)
151 KB
151 KB JPG
>>
>>10452734
>>10452735
Definitely would like the model and prompt for this one.
>>
File: FiY1DrdaUAAo3Cz.jpg (140 KB, 896x1216)
140 KB
140 KB JPG
>>10452941

The original author is this guy I just reposted here.

twitter.com/Claire_Miyamoto
>>
File: tmp671ncbwz.png (890 KB, 1024x512)
890 KB
890 KB PNG
>>
There needs to be a guide for total beginners, as in FROM SCRATCH with no prior knowledge, on how to generate your own artwork. Cos this is massive
>>
>>10450385
>>10446993
I've noticed AI nails hot midriffs. I wonder if there is a research paper on why it sucks so bad at feet
>>
>>10453221
1: go to novelAI
2: pay for account
3: write "huge breasts, giantess, city" and hit generate ad nauseum
>>
>>10453265
I suppose because the shape of feet is more complicated than a midriff and they have more fine details to them, though that may be completely wrong.

>>10453361
It really is that simple. No need for a guide.
>>
File: geez.jpg (3.63 MB, 3717x4055)
3.63 MB
3.63 MB JPG
Tried making a batch with Artoria and it was just thicc and big boobs. While chilling on rooftops/railings.

Throw BB in there and she starts working, she is just made for this stuff. Which makes alot of sense considering the official content for her and the artwork I'd imagine they inspire heh.
>>
>>10453548
What was your prompt? Ive gotten some great Fate art but it falls apart when I attempt making size content.
>>
>>10453548

I don't know if this is too much ask but could you upload a folder with all prompt results separated?
>>
File: 5.jpg (183 KB, 1024x1024)
183 KB
183 KB JPG
any mini gts here
>>
>>10453265
https://colah.github.io/
Based on my knowledge of colah's work, I'd say it has to do with how deep the structures are in terms of nested features. Basically what >>10453405 says, a line detector/generator is easy, and you can make circle generator out of line generator, which is already pretty close to midriff generator, but a foot generator would need to be composed of heel generator and toe generator that can compose exactly 5 toes and each toe generator has to be composed of whatever. It's definitely a much deeper structure. But it's already impressive how deeply into the "giant woman" structure an ai can go, it just doesn't get all the way to fingers. It's interesting though that it can do so well with faces in comparison.
>>
File: tmpqz085nk4.png (1 MB, 960x960)
1 MB
1 MB PNG
Using AI to illustrate my interactive stories. Shit is tight.
>>
>>10453864
No, only omnipotent gods.
>>
>>10454170
please man what is the prompt
>>
so everyone's using novelAI now? what about stable diffusion
>>
>>
>>10454338

As someone who's completely out of the loop, are any of these usable in a reasonable capacity for free? I know Novel AI is paid but I though AI diffussion was like a subsystem for it, but it appears to be an entirely independent AI thing?
>>
>>10454338
Stable isn't as good at anime girls
>>
>>10454338
I only use Stable Diffusion
>>
File: tmpg0x4uic3.png (1.2 MB, 960x960)
1.2 MB
1.2 MB PNG
>>10454330

I just copy paste salty twitter posts from size artists whining that AI will never be good and SD just returns these gens. It's fuckin weird
>>
>>10452251
This is really good. Honestly the most impressive thing about this gen are how nice and even all those pokerchips look lol. This is how you tell I do to much of this. Hot gts and Im focused on the poker chips.
>>
>>10443186
We need a guide posted for how to get pics like these
>>
File: 1.png (2.2 MB, 1408x1024)
2.2 MB
2.2 MB PNG
>>10454408
yes Novel AI model has been leaked, if you got a gaming gpu you can just generate these for free
>>
File: tmp_y_k9s9q.png (757 KB, 768x1024)
757 KB
757 KB PNG
>>10454411
SD by itself is not good at anime, but once you start merging models, Waifu diffusion, SD and even NAI model too, you can get pretty amazing results, NAI looks too samey for me, you can have very different anime style results on SD
>>
>>10454461

I wanna give it a try. Where can I find this?
>>
File: FgD8MkhVEAE1Nxd.jpg (89 KB, 512x768)
89 KB
89 KB JPG
>>10454652

google it fren
>>
File: FgD8MkkVsAAPlz9.jpg (92 KB, 512x768)
92 KB
92 KB JPG
>>
File: FgD8MkqUUAAhtMJ.jpg (90 KB, 512x768)
90 KB
90 KB JPG
>>
File: FgD8MlIUcAAUkDr.jpg (83 KB, 512x768)
83 KB
83 KB JPG
>>
>>10453602
>BB(fate), giantess, middle of tiny city, giant, massive, colossal, destruction, smug, sudden growth spurt, damaged maid uniform, illustration, single red ribbon in hair, skyline, official art, purple hair, sexy pose, exposed nipples, sitting on several buildings at once, movie

Go nuts, I find getting results to be a mix of using the right model and getting in the head of the AI by suggesting placements that mess with perspective.

>>10453637
Here ya go anon
https://www.mediafire.com/file/zz7lky09nr8jn1k/bigBB.rar/file
>>
Is it possible to generate tiny/shrunken people with Novel AI? Haven't found a way yet myself
>>
>>10455107

Thx, much appreciated
>>
File: 1669153384137369.webm (2.83 MB, 960x540)
2.83 MB
2.83 MB WEBM
>>10453864

Got an AI-generated growth sequence from normal to mini that was posted in the other thread. This feels like it has real untapped potential.
>>
>>10455413
Looks nice. Could be interesting to see more of this, especially for non-mini sizes.
>>
File: 101975608_p0.png (543 KB, 768x512)
543 KB
543 KB PNG
Commencing dump
>>
File: 101975608_p1.png (518 KB, 768x512)
518 KB
518 KB PNG
>>
File: 101992701_p3.png (529 KB, 768x512)
529 KB
529 KB PNG
>>
File: FeS1U6oaUAAJAZW.jpg (87 KB, 1024x512)
87 KB
87 KB JPG
>>
File: FeS1XoZakAEc9ep.jpg (85 KB, 1024x512)
85 KB
85 KB JPG
>>
File: FhuDrpsaUAAFddU.jpg (159 KB, 1024x1024)
159 KB
159 KB JPG
>>
File: FfxHhgVUcAE-B2z.jpg (112 KB, 768x1152)
112 KB
112 KB JPG
>>
File: FfgM8GIagAAG9wv.jpg (82 KB, 512x1024)
82 KB
82 KB JPG
>>
File: FgFY4oHUoAAL7gb.jpg (82 KB, 512x768)
82 KB
82 KB JPG
>>
File: Fe29FLmaAAEONy6.jpg (120 KB, 512x1024)
120 KB
120 KB JPG
>>
File: FgK0kVMacAACOs6.jpg (137 KB, 768x1152)
137 KB
137 KB JPG
>>
File: FgK0nQ0aEAAiRw7.jpg (89 KB, 512x1024)
89 KB
89 KB JPG
>>
File: FgmCQAUUUAApVcJ.jpg (79 KB, 512x768)
79 KB
79 KB JPG
>>
File: FgmCQAZUcAA6n1x.jpg (92 KB, 512x768)
92 KB
92 KB JPG
dump end
>>
Following from another asking (as I'm getting into Novel AI), is there a way to generate shrunken/tiny people? I've gotten mostly POV stuff and some macro but I'm more on the people getting shrunk side. Scratching my head trying to figure out the tags
>>
File: demongirl.jpg (522 KB, 960x704)
522 KB
522 KB JPG
>>10456373
>>10456377
>>10456378
>>10456380
>>10456384
I like these ones, I find it hard to get the girl to be further back. Rather than front and glued to the camera.
>>
>>10455320
>>10456692
I'm just now figuring NAI out, but I think it'd be more consistent just making two separate pics and collaging them together. You could put tags like "white background" or "simple background" for the tiny person's pic, though I don't know enough to make sure all the characters in the final pic share the same style.
>>
my laptop is super crappy and only has a few gigs available. is there any way i can do this stuff?
>>
>>10457089

As far as I know the only real requirement is a somehat decent GPU with at least 8GB of vram.
>>
File: FgKvIH6VEAEUmQz.png (578 KB, 512x768)
578 KB
578 KB PNG
>>10456793

Then you might like these as well
>>
File: FetKqTuUoAAzuE-.jpg (297 KB, 768x512)
297 KB
297 KB JPG
>>10456793
>>
File: gts.png (2.46 MB, 2078x1023)
2.46 MB
2.46 MB PNG
Hard vore bros....help me out here in making a better prompt.......

Also when I try other prompts the giantesses always have massive tits kek, is there any particular reason for that
>>
File: bigbb.png (668 KB, 832x640)
668 KB
668 KB PNG
>>10457361
>>10457362
Yeah, I find the poses to be more interesting. The better looking results I get look more like some alright college work.

Started playing around with the x/y plot feature and its making it much easier to find decent results.
>>
File: bigsb.jpg (2.78 MB, 3328x1920)
2.78 MB
2.78 MB JPG
I like the way this model handles bodies, eyes come out super sketchy through. But probably an easy fix if I were to bother with image editing.
>>
File: tmpyfn4nw8q.png (339 KB, 832x512)
339 KB
339 KB PNG
>>10457549

Currently it is challenging even with SD and better models. Your best bet is to use img2img to get the mouth and then inipaint in the prey, though honestly I tend to just do POV so I dont need to bother with showing prey. In general closeups and internals are still challenging. I can pull them off but only after lots of Cronenbergs
>>
File: gts2.png (2.6 MB, 2061x1023)
2.6 MB
2.6 MB PNG
>>10457701
Yeah I might need to learn more about how this all works
>>
File: samplers.jpg (322 KB, 2553x768)
322 KB
322 KB JPG
Whelp now I'm messing with samplers, between this and clip skip productivity is going up up up.
>>
File: xodhp754k9t91.png (789 KB, 512x1024)
789 KB
789 KB PNG
>>
File: FgkHsQ0UAAA76uH.png (645 KB, 768x512)
645 KB
645 KB PNG
>>
File: FfYYbUyVIAAK1h_.png (536 KB, 768x512)
536 KB
536 KB PNG
>>
File: FfYYbVAVQAAM1Ef.png (525 KB, 768x512)
525 KB
525 KB PNG
>>
File: Ffqwgt4aEAAnPPa.jpg (74 KB, 1152x768)
74 KB
74 KB JPG
>>
File: FfqwhKBaEAEtkCG.jpg (62 KB, 768x512)
62 KB
62 KB JPG
>>
File: Fghs_EYWIAY4eIW.jpg (77 KB, 512x1024)
77 KB
77 KB JPG
>>
File: FfT4Ek6XoAMlxni.jpg (82 KB, 512x1024)
82 KB
82 KB JPG
>>
File: MEGUUUU.jpg (158 KB, 768x1024)
158 KB
158 KB JPG
>>
File: MEGUuu.jpg (193 KB, 768x1024)
193 KB
193 KB JPG
>>
File: gts.png (2.45 MB, 2048x1044)
2.45 MB
2.45 MB PNG
giantess, light smile, native american
>>
File: FetKpvpakAA4naZ.jpg (319 KB, 768x512)
319 KB
319 KB JPG



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.