Show-1

Show-1 is a research text-to-video model from Show Lab at the National University of Singapore that combines pixel-based and latent-based video diffusion. It uses pixel diffusion to produce coherent low-resolution motion and latent diffusion to upscale, which improves text-video alignment over latent-only approaches. Base, interpolation, and super-resolution checkpoints are published on Hugging Face for self-hosted inference on a GPU.