Cosmos-H-Surgical
Surgical World Model
Cosmos-H-Surgical is a surgical world model built on NVIDIA Cosmos3-Nano. It supports video prediction, generating the next 92 frames of surgical video from a starting image and a text description, and simulation-to-real transfer, turning a blur, depth, segmentation, or edge control video into photorealistic surgical video at 480p. A distilled four-step DMD2 checkpoint is included for faster transfer.
No agent skill for this model yet. Use it directly via
HuggingFace
or GitHub.
Engineering & research use only — not a medical device, not for diagnosis or clinical decision-making. Skills follow the open Agent Skills spec.