/vt/ - Virtual Youtubers

>>39107157
And for comparison, here's someone the other models were already not too bad at, again with the prompts from the OP. Definite improvement here as well but not nearly as drastic in my opinion.

That said, it's more than just about being lazy. The less stuff you have in your prompt just to get it to draw the right person, the easier it is to get the AI to actually focus properly on the other descriptors you have, like clothing or scene or setting or so on. And in my experience the longer a prompt is the harder it is to make small adjustments to something that's almost right without the whole thing blowing up in your face