Seedance para Geração de vídeo

Opiniões da comunidade

• 2026-09-08T12:08:29.274287+00:00 — FateOfMuffins

  I've never imagined that LLMs would ever become better than diffusion models at creating any kind of video. Now I can see how just one year from now we could have something like an llm creating cinematic scenes rendered kind of low poly in UE5 but with DLSS5 on top of it to make it realistic. It would be very customizable and competitive compared to Seedance

• 2026-08-03T07:23:26.176476+00:00 — Remove_Forward

  This one apparently did use a diffusion model (Seedance 2.5) for some of it, but Opus 5.5 was the one using the diffusion model. Still very impressive. The other p(doom) music videos floating around didn’t use any diffusion for visuals from my understanding. [

• 2026-07-15T06:56:10.11361+00:00 — echelon

  America doesn't have anything close to Seedance 2.5