#TripoSR

TripoSR is an open-source tool designed for efficient 3D reconstruction from a single image, developed by Tripo AI and Stability AI. It implements the Large Reconstruction Model approach to achieve fast processing and superior quality, producing models in under 0.5 seconds using NVIDIA A100 GPUs. TripoSR has shown better performance than other open-source models on several datasets. The project, distributed under the MIT license, provides source code, pre-trained models, and a user-friendly demo, facilitating advancements in 3D generative AI for researchers and developers.

Stable Fast 3D facilitates quick 3D mesh reconstruction from a single image using UV unwrapping and illumination separation for enhanced texture quality. Based on the TripoSR model, it produces artifacts-free meshes with detailed material prediction, ideal for game assets. Compatible with Python 3.8 and supports both CUDA and MPS backends, including experimental settings on Mac and Windows. Features a Gradio demo and ComfyUI integration with customizable remesher options for specific mesh configurations.