Current 3D generative models are slow and low quality. We present GRM, a large-scale model that reconstructs 3D Gaussians in 0.1s and generates high-quality 3D assets from text or single images in a few seconds. Demo: 1/4

GRM is a feed-forward transformer-based model that efficiently incorporates multi-view information to translate the input pixels into pixel-aligned Gaussians, which are unprojected to create a set of densely distributed 3D Gaussians representing a scene. 2/4

GRM works with various 3D diffusion models: Zero123++, Instant3D, SD3D, and V3D. Generated assets are represented as 3D Gaussians, which can optionally be exported into meshes that are compatible with existing graphics pipelines, such as Blender. 3/4

Amazing team members include @YinghaoXu1, @Vivianszf1, @toomanyyifans, @HanshengCh, @CeyuanY, @pengsida, Yujun Shen 4/4