/bbwai/ - bbwai

Here is a prompt:由四个平凡无奇的瞬间组成变胖对比图，展现爆肥带来的从脸到身体道四肢的所有部位的肥满化巨变。 丰腴微胖->超重显胖->大胖妞->SSBBW。 保持画风服饰与动作统一，强调变胖爆肥逐步变化。女主长相十分普通但是身材极其丰乳肥臀。从左到右排布四个阶段。 现场采访的截图。主角是17岁的少女，穿着暴露，手持话筒，纯真微笑。偏侧面角度。 第一张图:.主角身材有曲线，肉感丰满，整体微胖丰腴。字幕:我要增肥 第二张图:主角身材发福变胖，轮廓圆润了很多，全身明显发福。此时女主体脂率上升，体重达到超重级别，身材走形，显得更加肥满。腿变粗变胖，并且脸部变大变圆，有显著的双下巴，肥肉覆盖每个部位，肚腩明显。外观显胖了。字幕:还不够胖 第三张图:在保持画风不变，角色整体、细节设定不变的条件下，角色到达肥胖的体型。身材完全走形，但还未到达ssbbw的程度。 脸要更大更圆更双下巴，补充肉鼓鼓的脸颊。 腹部至少要有两层游泳圈，靠下方的那层会更加向外突出。字幕:理想身材 第四张图:人物爆肥，胖得轮廓圆润丰满，高体脂率高体重，赘肉肥硕向外溢出，体型庞大肥满，四肢粗胖，臀部腿部和手臂胖得没有缝隙，大肚腩圆滚滚，面部又大又圆胖乎乎的，脸颊过度饱满，整体膨胀肥胖。极端严重肥胖，夸张extremely obese，远超最胖级别的SSBBW。字幕:控制不住

I'm no good at AI. And everything is "free" where I check. So, I got her consent and shes pretty curious too on what peeps can come up with. Go all out. Niche fetishes, normal fetishes, weird fetishes, anything under the sun. I tried my hand at it a little bit but All I could get was a pic to vid of her throwing her ass around.

The original "Photorealistic cartoon pics" thread was ruined from users getting the wrong impression, so I started a new one. The original thread supposed to focus on AI-generated realistic-styled screenshots from animated shows.

Post you're favorite current AI wg games and give you're opinions on them.

Let's bug Ai's with proper fat questions and share the results. Some things might come out hot, some fun, some educational, some inspirational and some could actually be practical. Here is a link to my fun gemini dialog log. https://gofile.io/d/juunFs Use this folder or add your owns.

Looking for some sort of free tool, like stuffer.ai, that can generate somewhat standout and unique faces. I use chatGPT and u find it has ‘same face’ syndrome pretty often.

Ssbbwilover charges for the nude and immobile photos of their AI are which is ridiculous. I've been trying to find out how they do it but keep failing. Any tips? They get a bunch of questions on their page asking how they do it but none of it really tells you how to do it. All I know is they use stable diffusion 1.5 and merge models biggergirlsv5 with Customs LORAs I'm not even sure If im using the proper site. Here's the creators page https://www.deviantart.com/ssbbwllover

I’ve been playing with Nano Banana Pro the past few days and I’ve had a lot of fun joining two of my great loves – classic video games and all things fat. I tried to capture the style and feeling of SNES, GBA, PS1, and new 2D games. I have a few dozen images to share and it’s a grab bag of different subject matter. This thread is for everyone and all types of pixel art content: male or female, loli/shota, furs, inflation, blueberries, etc. Anything that you care to share is welcome as long as it adheres to this art style. Enjoy!

Past several months I've been using Grok for erotic stories and I'm ruined for normal porn now. Here's a short example: grok.com/share/c2hhcmQtMg_0f11728f-89ee-45bb-b8e1-1c94d5bdf2f2 Feel free to use that as a starting point or start your own from scratch and then share here, would be fascinated to see what people come up with as I've run out of ideas now.

Since you can do very precise regional prompting, this allows you to have a lot of control over the poses, and the body shape, which allows you to do stuff which was very hard to prompt for in other models. I've made these images without any loras or control-nets. It sometimes adds a bit too many skin imperfections for my taste, so adding a second pass with Z-Turbo works great to make the skin more idealized.

Thought we make a Thread for Jscammie.com https://www.jscammie.com/

post ai BELLIES all shapes and sizes are welcome

someone else posted this image and was wondering if there were any more of this type. Women with huge bellies to the point of being unrealistic

I've trained two Loras for Z-Image Turbo, and created some easy to use Workflows for ComfyUI to use these for Image and Video generation. [Feedai Lora](https://civitai.red/models/2545806/feedai) Produces realistic small to large BBW Women. Can be used with a bunch of different triggerwords including: feedai_lora_token: the basic activation trigger dOuble_b3lly_lora_token r0und_b3lly_lora_token tu88y_lora_token: on the chubbier rather than bbw side cut3_r0und_belly_lora_token B008s_lora_token: large boobs Unfortunately the model only loosely follows these triggerwords, and tends to bias towards soft double bellies. I've found leaving out any trigger words tends to be the most consistent for getting smaller, rounder bellies. [SSAIBBW Lora](https://civitai.red/models/2545842/ssaibbw) Produces unrealistically large SSBBW Women with the triggerword ssaibbw_lora_token. Each come with their own workflow (attached to the images on CivitAI), containing a wildcard prompt which will generate an infinite combination of BBWs and scenarios. The image generation workflows first produce a rought draft, and then refine this in a second polish stage. This allows for more control over the size of the BBWs without compromising realisitc detail. The strength of the Lora can be adjusted individually for the two stages, for e.g in the SSAIBBW workflow the rough draft has very poor detail, but can produce extremely large BBWs, which the model otherwise struggles with, and then can be polished to be more realistic in the second stage. The video generation workflow (attached to the videos on CivitAI) is designed to take the images and produce varried 5s videos in a batch. You select the first image in your input folder, and select the number of runs, and it will iterate through all the images. I use a NSFW lora, as this produces much more realistic jiggling. The higher the resolution, the better the detail, but this also results in much longer render times. I run these workflows on my 8GB GPU (64GB RAM), the images take ~2min, and the vidoes ~4min. I usually just queue a bunch, and let it generate overnight.

Anyone had any luck doing vore stuff? I know it's a bit too weird for some. I've tried using stuffer.ai but I can at best get implied vore and no actual vore. It seems like descriptions work ok but I'm finding the actions to be difficult to get the AI to perform. These are some attempts I have tried.