Stable Fast 3D (SF3D) is a large reconstruction model based on TripoSR, which takes in a single image of an object and generates a textured UV-unwrapped 3D mesh asset.

• Developed by: Stability AI

• Model type: Transformer image-to-3D model

• Model details: This model is trained to create a 3D model from a single image in under one second. The asset is UV-unwrapped and textured and has a relatively low polygon count. We also perform a delighting step, enabling easier asset usage in downstream applications such as game engines or rendering work. The model also predicts per-object material parameters (roughness, metallic), enhancing reflective behaviors during rendering. The model expects an input size of 512x512 pixels. Please check our tech report and video summary for details.

Model Sources

• Repository: https://github.com/Stability-AI/stable-fast-3d

• Tech report: https://arxiv.org/pdf/2408.00653

• Video summary: https://youtu.be/uT96UCBSBko

• Project page: https://stable-fast-3d.github.io

• arXiv page: https://arxiv.org/abs/2408.00653

Community License: Free for research, non-commercial, and commercial use by organizations and individuals generating annual revenue of US $1,000,000 (or local currency equivalent) or less, regardless of the source of that revenue. If your annual revenue exceeds US $1M, any commercial use of this model or derivative works thereof requires obtaining an Enterprise License directly from Stability AI. You may submit a request for an Enterprise License at https://stability.ai/enterprise. Please refer to Stability AI's Community License, available at https://stability.ai/license, for more information.