Does anyone know what AI is used to do this?
You must be registered to see the links
You must be registered to see the links
Does anyone know what AI is used to do this?
- Thread starter alextraszas
- Start date
Cpt.Sus
Newbie
- Aug 2, 2023
- 28
- 10
- 80
You're right, wan doesn't do that, in fact if you explore the page in the second link more, you'll see that there are several similar ones, but I don't know what they're made with.
Of course WAN generates NSFW vids.
And yes, I'm fairly certain the video in the OP is WAN, you can clearly see WAN's "pingpong" feature in some of the sequences.
Last edited:
Cpt.Sus
Newbie
- Aug 2, 2023
- 28
- 10
- 80
I tried generating an NSFW vid but it hit me with an error, but a SFW vid was generated no problem
- Sep 7, 2025
- 131
- 308
- 63
Im curious too.. are there sites like Midjourney .. etc .. that can take an image and animate NSFW?
Where did you try? A website?
I'm taking about a local install on your PC, you can get plenty of NSFW Loras for WAN on civitai.
Seanthiar
Active Member
- Jun 18, 2020
- 624
- 836
- 237
You can do that with Hunyuan and with WAN - for example Civitai has enough loras and modified models to make NSFW content. Problem is only you need time and do some postwork because the video you can make with it are shorter by default.
And you can do take an image and make a video out of it without problem at home. Just need a decent game PC if you want to do it under windows.
And you can do take an image and make a video out of it without problem at home. Just need a decent game PC if you want to do it under windows.
Middling Emu
Newbie
- Nov 1, 2016
- 76
- 894
- 216
Found the Spaniard!
LoRA is Low Rank Adaptation, it's a thingy used to fine-tune AI image generation.
Seanthiar
Active Member
- Jun 18, 2020
- 624
- 836
- 237
I guess you mean LORA? LOw Rank Adaptation model. A small model that adds to a big diffusion model. In laymen terms you could say it is a template for the big stable diffusion models like Pony to get better results and they react to code words in the defining text prompts. Lora's are 'trained' on specific things like a specific race, specific body type etc.. and the lora is often connected to a codeword you can use in the text prompt to get results trending into that area. For example there is a lora for to get women with big chests as a result with the code word bigtiddies in the text prompt. And there are LORA that are for real looking content or specialized for anime. Lora can modify the big models like FLux that does not support NSFW to include NSFW content and many lora give you an option to combine them if you want more than one 'feature' like NSFW and big breast.
LORA are most of the time model specific. For example a LORA trained for the PONY model does not work with the FLUX model.
Oh and don't forget model with similar names does not do the same - for example FLUX DEV is very different from FLUX KONTEXT
Wow, it's definitely more complex than I expected, it's clear that it requires a bit of dedication to create that type of videos, the truth is I didn't understand very well haha, in my ignorance I thought it was just a matter of putting words and in a "magical" way the AI would do it, but I see that's not the case, from this point on it's now clear that it's undoubtedly an effort on the part of the creators
- Sep 7, 2025
- 131
- 308
- 63
Alex 
If you want to go the route of using methods like Stable Diffusion / ComfyUI to generate AI images - and then convert them into videos with what these guys are talking about - or downloading Loras, creating your own Loras - you can get download additions and link them with your AI generating setup to animate them .. but even then, there are issues with that - too many to describe. It is, very complicated, to say the least, every step of the process - but some people love doing it, and find it rewarding, and if you are dedicated enough, and dont mind spending days, weeks, learning, trial and error this and that - and finally become decent at it, you really could have an unlimited amount of custom images and videos, that you kind of 'create' yourself and even find it is an art of some sort. And really, you dont even need to pay for it. But it is a lot. And it really does take a beefy computer. Its one of those things that when you purchase an expensive video card, and you think to yourself - will I ever use this? - well, there you go. I have a super computer system, and it is running hot all the time, and running out of VRam constantly. Some people enjoy this process, I dont really, but I need assets for the NSFW games Im creating. Its not like I can hire somebody to draw 500 images for me. And random AI from sites is way too fickle. So I did take the time to learn it all. It was annoying - and I hated it. But now I know it - and it is kind of rewarding. If you are looking for a decent hybrid approach - setting up Stable Diffusion on your system isnt that hard - chatGPT can give pretty decent step my step instructions how to do it. You could download Loras that you like .. and generate in infinite amount of free images. There are a lot of next extensions, including cropping the images. You can generate images and characters you like, and if you dont want to take the extra step to get into animating them - (I think you would need to use ComfyUI, personally I dont like the node system, and I prefer SD), you can take your images, and try a site like makefilm.ai - (But it costs $$), but you can get a pretty decent animation from it pretty easily. With this hybrid approach, there is a lot you can do really. You can generate characters with SD, on a plain background.. crop them out, use a graphics editor to put them on a greed background .. give that to makefilm.ai , animate the characters doing what you want .. bring them into a video editor.. green screen out the background .. and have fun with your animated character that way. Well.. that's one thing I do.
Sorry to be so long winded here. Ya .. it's complicated.. but after enough effort, it becomes worth it.
If you want to go the route of using methods like Stable Diffusion / ComfyUI to generate AI images - and then convert them into videos with what these guys are talking about - or downloading Loras, creating your own Loras - you can get download additions and link them with your AI generating setup to animate them .. but even then, there are issues with that - too many to describe. It is, very complicated, to say the least, every step of the process - but some people love doing it, and find it rewarding, and if you are dedicated enough, and dont mind spending days, weeks, learning, trial and error this and that - and finally become decent at it, you really could have an unlimited amount of custom images and videos, that you kind of 'create' yourself and even find it is an art of some sort. And really, you dont even need to pay for it. But it is a lot. And it really does take a beefy computer. Its one of those things that when you purchase an expensive video card, and you think to yourself - will I ever use this? - well, there you go. I have a super computer system, and it is running hot all the time, and running out of VRam constantly. Some people enjoy this process, I dont really, but I need assets for the NSFW games Im creating. Its not like I can hire somebody to draw 500 images for me. And random AI from sites is way too fickle. So I did take the time to learn it all. It was annoying - and I hated it. But now I know it - and it is kind of rewarding. If you are looking for a decent hybrid approach - setting up Stable Diffusion on your system isnt that hard - chatGPT can give pretty decent step my step instructions how to do it. You could download Loras that you like .. and generate in infinite amount of free images. There are a lot of next extensions, including cropping the images. You can generate images and characters you like, and if you dont want to take the extra step to get into animating them - (I think you would need to use ComfyUI, personally I dont like the node system, and I prefer SD), you can take your images, and try a site like makefilm.ai - (But it costs $$), but you can get a pretty decent animation from it pretty easily. With this hybrid approach, there is a lot you can do really. You can generate characters with SD, on a plain background.. crop them out, use a graphics editor to put them on a greed background .. give that to makefilm.ai , animate the characters doing what you want .. bring them into a video editor.. green screen out the background .. and have fun with your animated character that way. Well.. that's one thing I do. Sorry to be so long winded here. Ya .. it's complicated.. but after enough effort, it becomes worth it.
I appreciate your response, without a doubt it is an art to learn how to do all of this, without a doubt I will have to leave it for later, when I have a PC with which perhaps I can enter this world, without a doubt AI opens a world of infinite possibilities and the sample video that I put made me wonder how they do it, because in a certain way it is at the level of a 3D animation and it is surprising to see how it has advanced so much and I have no doubt that every day it will be more surprising to a level that perhaps, like now, is beyond my understanding, as Bill Gates once said, it is fascinating to think about the things that people will do with all of this hahaAlex
Sorry to be so long winded here. Ya .. it's complicated.. but after enough effort, it becomes worth it.
- Sep 7, 2025
- 131
- 308
- 63
As I was responding to your message, and all day today really, I have been creating assets for my upcoming game. I have been using to the workflow I described, and created all my characters. Mostly, I create the basic ideas with AI tools .. and my wife, is quite talented with graphics and drawings, does all her magic to make them look good. Then i create my own custom Loras for all the characters in the game, so I can put them in various situations, or just as general sprites for the game - and then use AI to animate them. So this one one of the characters in the game. I imagine, it wont belong before a lot of these kinds of games have very elaborate, nice animations.
This is fantastic, it's at the level of a professional animation, when I saw this video, I thought it was an animation in Blender or one of those programs (I don't know anything about this) and then I saw the tags and some answers and the creator says it was with AI and it blew my mind because I said, damn, how I would like to make my own videos too, this of course without knowing the background and all the work that it clearly entails, in my ignorance I only thought it was just putting words and that's it haha