ByteDance Seed Team announced the launch of the 3D generative model Seed3D 1.0, achieving end-to-end generation from a single image to high-quality simulation-level 3D models.

Seed3D 1.0, adopting the Diffusion Transformer model architecture widely used in generative AI, is trained with large-scale data, capable of generating complete 3D models including fine geometry, realistic textures, and physically-based rendering (PBR) materials.

The team plans to introduce multimodal large language models (MLLM) in the future to enhance the quality of 3D generation and promote the large-scale application of 3D generative models in world simulators.