3D Topia Welcome to the 3D Topia homepage. 3D Topia is mainly developed by Shanghai AI Laboratory. We keep working in the field of versatile 3D content generation, open-sourcing the datasets, the models, and benchmark suites.

🚀 Updates

🔥 3DTopia-XL

• [09/2024] We release 3DTopia-XL, a high-quality 3D generative model with 1 billion parameters, which supports the generation of 3D assets with high-resolution geometry, texture, and material into GLB format.

🔥 3DTopia

• [03/2024] Technical report of 3DTopia is released.
• [01/2024] We release 3DTopia, a two-stage text-to-3D generation model that uses hybrid diffusion priors for high quality 3D assets generation.

🍿 Datasets

• Objaverse Caption from 3DTopia.
• OmniObject3D for real-scanned objects.

🎁 Models

• 🎉 [new] 3DTopia-XL: Comming soon...
• LGM: Large Multi-View Gaussian Model for high-resolution 3D content creation from text prompts or images.
• 3DTopia: Large Text-to-3D Generation Model with Hybrid Diffusion Priors

🏔️ Bench

• 3DGen-Arena: Human preference for text-to-3D and image-to-3D generative models.
• GPTEval3D: GPT-4V(ision) is a human-aligned evaluator for text-to-3D generation.