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

seedance 2.5 is released, but for 10 platforms only for now. Not sure when Venice will get their hands on it and whether it'll be uncensored or not.
 
seedance 2.5 is not released yet, not even on ByteDance's own platform.
the release date is 25th July, and the release date for API is unknown. so assume it is coming to Venice like in next month.
 
how would yall rate audio reference aspect of seedance? personally i find it underwhelm as the moaning sound does not really reflect voice in original audio
 
video and audio are getting better but still way behind grok1.5, we will see what happens with 2.5. The quality is improving pretty fast so I think maybe in a year we will have what we are all looking for
 
when using a Ref vid for Seedance 2.0 R2V Enhanced does the ref vid need to be exactly lets say for example 8.0 seconds or could there be milliseconds?
Do I round up on the prompt option/selection?
 
Reference video doesn't need to be as long as your output afaik. Unless you have dialogue and even then it's a toss up. You could use a 5 second reference video and choose a 15 second output.
 
Thanks for the help!
I keep getting this error anyway to fix it?
Seedance 2.0 R2V Enhanced
Could not process the face asset. Please try a different image or video. You were not charged or your credits have been refunded.
 
I get those too. Even with assets that worked just hours earlier. You're better off just trying a different image or video than trying to find a way to fix it. It'll probably work another time. It's just one of those things.
 
So no viable alternatives then. Grok I2I and I2V were my only reason for using this, the other models need an essay as a prompt for a result that doesn’t even compare…
 
They never gave an official release date, there was only speculation and a target for july
 
I want to revise this opinion on Krea2.
If you're using this locally, it can be brilliant. I trained a few character loras and grabbed a NSFW enabling lora. It's still just a text to image model, with no reference capability, but I'm having fun.
 
is there any other platform that offers SD 2.0 uncensored? I feel like Venice is little bit expensive.
 
Its about the same price everywhere. Maybe even a little more elsewhere if it's purely pay per use without subscription
 
I looked too but couldn't find anything cheaper; plus, on some sites you can't include multiple elements—just a starting image and an ending one. When I asked AI about it, it mentioned cheaptokens.ai and APIs, but I don't understand any of it, lol.
 
qwen image pro 2.0 and seedance v5 pro keeps blocking my prompts that recently worked any prompt suggestions?
 
You must log in or register to reply here.
Back
Top