This model was trained exclusively on POV clips of two women simultaneously licking/kissing the shaft of the penis. It does NOT include any testicle sucking.
This lora works well by itself at strength 1 even without the general NSFW lora. It's better for live-action than anime.
Description
FAQ
Comments (18)
Will there be a t2v variant?
Double Blowjob (Wan2.1) by dtwr434 would pull out a girl actually giving a blowjob, and the other licking the penis or the balls. This LoRA is definitely nice, thanks for the training. But it's more a double licking penis, than a double blowjob.
Consider it a stylistic difference. Personally I prefer this style, which is why I trained it this way. You're free to have your own preferences but I was pretty clear in the description as to what this is. If you've got something more constructive to add I'm happy to listen.
For me personally it helps, when you combine two LORAs together, like this one and another ‚real‘ blowjob one. Haven’t tried on this yet, but it works for me with other’s quite well. You only have to experiment with the order and strength a bit and mostly you get what you want.
***** EDIT *****
Here is a quickly made example: https://civitai.com/posts/21902326
@juniorrmk173 Thanks for posting - I'm new to all this and that helps heaps.
@crju I‘m basically new to this too, I just have the freedom (GTX 5090) to experiment a lot with my videos. 😜
Might be my fav 2.2 Lora so far
Is there s trigger word?
epic cool
This is super nice ! I do have an hard time getting the girls to look at the camera, by default they just look at the penis after less than a second 🤔
I tried various things including : (both women are looking at the camera from beginning to end:2)
Am I doing it wrong ?
There's a good amount of eye contact in the dataset, maybe try the phrase "They look at the camera, making eye contact with the viewer" and see if that helps.
@TwoMoreLurker Thanks for your reply, just a follow up, it was because the workflow I was using had a "Florence LLM" node which was interfering with my prompts.
I start my prompt with "(They never stop looking at the camera, making eye contact with the viewer:2)" and it really helps
Token string? Default prompt?
Check the sample images for a prompt. Though this lora does basically one thing and one thing only, so as long as your prompt is sufficiently descriptive it should work fine.
Nice. I suggest change the lora name to licking penis
This is my fav 2.2 lora for sure, it's incredible. Was curious if you would make one with one girl sucking the other licking. Trying to combine loras to achieve this never seems to work out. Great job!
Anyone know why this is creating a segmented double penis instead of a single?
I get that too, my guess is it's either the seed or because it's blending with a NSFW model, for example I use the NSFW Remix lighting model and with this lora on top of it, it gets confused and gives two d***s instead of one. I think the model already has the capabilities to do this it's just harder to get it to do so because of not having the proper prompt setup. The only thing I found that works most times is to change the strength of the lora so it doesn't conflict as much with the model or when you find a seed that this doesn't happen with fix it in.
Details
Files
WAN-2.2-I2V-Double-Blowjob-LOW-v1.safetensors
Mirrors
WAN-2.2-I2V-Double-Blowjob-LOW-v1.safetensors
WAN-2.2-I2V-Double-Blowjob-LOW-v1.safetensors
WAN-2.2-I2V-Double-Blowjob-LOW-v1 2.safetensors
low.safetensors
WAN-2.2-I2V-Double-Blowjob-LOW-v1.safetensors
WAN-2.2-I2V-Double-Blowjob-LOW-v1 (1).safetensors
WAN-2.2-I2V-Double-Blowjob-LOW-v1.safetensors
WAN-2.2-I2V-Double-Blowjob-LOW-v1.safetensors
WAN-2.2-I2V-Double-Blowjob-LOW-v1.safetensors
WAN-2.2-I2V-Double-Blowjob-LOW-v1.safetensors
WAN-2.2-I2V-Double-Blowjob-LOW-v1.safetensors
WAN-2.2-I2V-Double-Blowjob-LOW-v1.safetensors
WAN-2.2-I2V-Double-Blowjob-LOW-v1.safetensors
WAN-2.2-I2V-Double-Blowjob-LOW-v1.safetensors
WAN-2.2-I2V-Double-Blowjob-LOW-v1.safetensors
Available On (2 platforms)
Same model published on other platforms. May have additional downloads or version variants.