Back to Pixels: Lifting Latent Video Diffusion into Pixel Space
Frontier video generation models work in a VAE latent space that compresses video in both space and time. This compression made training tractable, but now limits high-resolution, controllable, and interactive generation -- doubling resolution quadruples token count, conditions apply to groups of frames rather than individual ones, and each response must wait for the rest of its multi-frame block. We present PixVid, a simple yet effective method that lifts a pretrained latent video model to directly generate every frame in pixel space. PixVid keeps the diffusion transformer and replaces the VAE decoder with a convolutional head that renders its outputs into pixels. We find that predicting clean video rather than velocity, together with a larger noise shift, is key to quality. Finetuned from Wan2.1-14B, PixVid is competitive with the latent baseline within $\sim 1.3$M training clips. Working framewise in pixels, it generates at $2\times$ higher resolution with the same per-frame token c
Related Stories
Anthropic CEO Amodei admits "there's a huge uncertainty in what's going to happen next year" days before a record IPO
The Decoder · November 14, 2026
Gemini 4 Carbon takes the MathArena crown but costs 15 times more than GPT-6.1 Sol
The Decoder · November 14, 2026
Anthropic says outdated prompting habits are slowing down Claude Opus 5.5
The Decoder · November 14, 2026