Workflow Venice AI Workflows and Models (DON'T POST REAL PEOPLE!!)

Remove anything with age or family. It's Venice blocking not the models
 
Is there currently discount code or promo codes that can increase the amount of credits when subscribing?
 
Found this online JBOOGX (-15%), just tried it with a subscription, and it worked ;)
 
yes, it can't do accurate genitals by itself and does some weird tongue stuff while sucking things, but it knows breasts out of the box and even porn (the movement and general setup, like I said, the details don't look too great), but that will easily be ironed out by the community with a couple of loras
 
If LoRA's needed for a true NSFW experience then Venice might take at least a month to provide the uncensored model. Am I wrong?
 
minimax h3 uncensored is already out on wavespeed but their playground section applies promp-level moderation so you have to use the API.
We have to wait for Venice but is there any other alternative? they are pretty slow.
 
I have tested it and still censored unfortunately. They probably mislabeled or something.
I am assuming Venice still uses the model from the official API because there should've been an 768p option (there is only 2K option right now) if it were the actual open sourced model.
I might be mistaken about it of course, but according to my testing the model is still censored on Venice and they mislabeled it as "Uncensored".
 
I did like 4 videos and only one was uncensored with fully breasts exposed. Other 3 were censored but yea its way to expensive when its forcing you to use 2k option. Not worth the tests for now
 
Yes exactly. It is way more censored and expensive than SD 2.0. It doesn't even perform like SD 2.0 imo. Not worth using.
It seems we have to wait for FLUX 3.
 
It works locally. You can't expect an API to be uncensored. Its an amazing tool for porn.
 
Yes, but I expected Venice to host the model in their own servers and serve to us. Instead they use the official API from Hailuo which makes using the model completely pointless.
Also the model has 66 GB size with 33B parameters. There is no way an average person can run the model locally. We have to use a provider like Venice.
 
Yeah trying it out, it takes forever for even a basic gen. And the quality isn't THAT great, I have to assume most of the stuff we're seeing is edited or super cherry picked.
 
The api works fine fully uncensored on wavespeed. Works best with references just like seedance does
 
You must log in or register to reply here.
Back
Top