/vt/ - Virtual Youtubers

>>64091217
txt2img is still growing as well, this is all txt2vid (txt2img, then using multiple stable diffusion videos in a chain to build the framework, then fed into animatediff) the main issue is getting sdv to do anything you actually want, so its alot of rerolling for every 25 frames, and the longer the chain, the less coherent it becomes.