Free generator that accept uploads?

necronfan

New member
Hello, I have been doing some work on Sora and Grok with interesting results that I will soon post here on the forum.

However, in Grok, which I use for free, there is a lot of restriction and moderation regarding uploaded drawings despite the great results because I am transforming fumetti into real images.

Does anyone know of a good AI generator that is really free and accepts uploaded designs?

I tried Perchance but I didn't see where I could upload drawings and in Venice you only had to pay to see nsfw...

Thank you for any tips!
 
Hello, I have been doing some work on Sora and Grok with interesting results that I will soon post here on the forum.

However, in Grok, which I use for free, there is a lot of restriction and moderation regarding uploaded drawings despite the great results because I am transforming fumetti into real images.

Does anyone know of a good AI generator that is really free and accepts uploaded designs?

I tried Perchance but I didn't see where I could upload drawings and in Venice you only had to pay to see nsfw...

Thank you for any tips!
Use Claude 4.5 Sonnet you have a free contingent with cooldown and u can upload NSFW Content in form of pictures or short videos to analyse and make prompts of it.
Claude cannot generate Images or Videos, so you must use another AI or Comfyui to generate it locally (Z-Image, SDXL/Pony, WAN 2.2 for Animations, i recommend an RTX Graphicscard for local generation!)

In my opinion claude is better in prompting then grok (both can create NSFW prompts)
 
A short hint for Claude: Dont ask directly to make NSFW Content. Claude will deny.
But if you ask for technical assistance for a prompt or something else and post a NSFW description CLaude will help you :)
(had that small problem yesterday, was thinking hu? why did he deny... then asked him and he said im here for technical assistance blabla... xD)
 
You will have a hard time with free generatations, there's plenty of NSFW sites that allow stuff but you have to pay.
 
Hello, I have been doing some work on Sora and Grok with interesting results that I will soon post here on the forum.

However, in Grok, which I use for free, there is a lot of restriction and moderation regarding uploaded drawings despite the great results because I am transforming fumetti into real images.

Does anyone know of a good AI generator that is really free and accepts uploaded designs?

I tried Perchance but I didn't see where I could upload drawings and in Venice you only had to pay to see nsfw...

Thank you for any tips!
That sounds interesting! I'd love to see the results you got with Sora and Grok. It's a shame that so many restrictions get in the way when you're trying to experiment with different styles.
 
There is no miracle, all online generator, even if you will pay will have resctriction, censored and of course spying, captation, trakers, etc.

Invest your time to learn with real tools to create, to do whatever you want.

Try this one: ComfyUI

The Local mode (Comfy Desktop) , not the Cloud online version. Learn and find, donwload, install safetensors you need for specific creation (breasts, nudifty, genitals, ePainting to replace, upscaling, etc.) and the only restrictions will be your knowledge "how to do it" and your imagination.
Perchance, Gemini, and all the online AKA "AI" it's not real AI, they create all time the same kind of models, faces, body, breasts, like DAZ3D (this sofware is very limited to create 3D, Blinder is so much better) with Gemini models bank. Because their “LoRAs” were trained using photos of certain models, and they censor them—or simply limit them. A single “LoRA” can't do everything on its own, so you have to combine them, manage them, and assign a weight to each one.


With ComfyUI, you could use two different mode : Regular ComfyUI or Krea2 (This one is more specific for nude models and detail).<
When you want a LoRA safetensor will do a specific creation or model face or any part of something, someont, like any AI, it needs training to learn with 400, 500, 1'000 pictures of the kind you want to learn. That is mean to train, to creat a LoRA safetensor.
You could find specific safetensor for NSFW without any restriction, uncensored, of course.

And at final point, the prompt lines, of course, will make everything. Do not create prompt as single line, like a junk texto text. Create section, real sentences.
ComfyUI unnderstand prompt lines different language (English, Spanish, French, etc.) So depends if your native language is on ComfyUI database, if it is, you could use it for prompt instructions.

The fp8 safetensor : for graphic card up to 8 GB DDR
The fp16 safetensor : for graphic card from 12 GB DDR (RTX 4070 xx) up to 96 GB DDR (RTX 8000 PRO Workstation)

CPU capacities, DDR5 quantity and Graphic Card with DDR6/DDR7 will make the job. More you have a powerfull workstation, nicer, highter resolution, detail, lighting will be better. For an exemple, with a RTX 4070 xx 12 GB, to create video, only 25/30 second max up to HD720. To create highter resolution, more DDR on CG will be ask. WIth ComfyUI, Nvidia GC is hightly recommand instead of AMD GC. For CPU, both are working with ComfyUI.

Offical website link
https://comfy.org/download/

LoRA training
https://github.com/CastilloworksAi/comfyui-workflows

Hudge Free Safetensor Database
https://huggingface.co/

It took me two months—working on it almost every day—to learn, understand, improve, and master ComfyUI. Even so, I'm not yet at the Expert level.

Regards,
Arthur Saturnin
 
Last edited:
I have started using Comfy-UI since a few days. I am getting good results so far. @Arthur Saturnin , thanks for the link to the LoRA training. The face dataset builder could be interesting.

Can you recommend a workflow that accepts uploads?
 
Back
Top