VGGT-Omega
Feed-forward 3D reconstruction from images or video
Image or text to 3D mesh — professional modeling tool
Depth, surface normals, and albedo from a single image
Streaming 3D reconstruction from video with local context
Generate 3D CAD models from sketches, point clouds, or text
Explore interactive visualizations with Viser
Generate metric depth maps and camera poses from images
Animate a first frame with Cosmos3-Super I2V (64B, 4-step)
NVIDIA Cosmos3-Edge — reason over images & video