PartCrafter vs TripoSR

PartCrafter is a free, open-source model that generates multiple 3D parts and objects from a single RGB image. It can handle individual objects and complex scenes, producing distinct meshes with meaningful part structure without requiring a pre-segmented image. Its architecture represents parts with disentangled latent tokens and uses hierarchical attention to pass information within and across parts. Local inference scripts can remove backgrounds and optionally apply style transfer to adapt real-world photos toward the rendered-image domain used for training. An optional vision-language feature can suggest the number of parts; Gemini is the default provider, and the provider architecture can be extended. Pretrained object-level and scene-level weights are hosted on Hugging Face. Installation is from the GitHub repository in a Python environment, and inference requires a CUDA-enabled GPU with at least 8 GB of VRAM. The linked Hugging Face demo is paused, though its page says users can ask the authors to restart it. The repository lists an MIT license and provides the software as-is, without warranty.

TripoSR is a free, MIT-licensed open-source model that reconstructs a 3D object from one image. Tripo AI and Stability AI developed it collaboratively for researchers, developers, and creatives working in 3D generative AI and content creation. The project states that generation can take less than 0.5 seconds on an NVIDIA A100 GPU. Its repository provides command-line inference and a local Gradio app launched with Python. The app offers background removal, foreground-ratio and marching-cubes-resolution controls, and OBJ or GLB output. Inference accepts one or more image paths and can bake textures instead of vertex colors, with configurable texture resolution. Default settings use about 6 GB of VRAM for one image. Local setup lists Python 3.8 or newer, PyTorch, and requirements.txt dependencies; CUDA is optional when available. TripoSR supports image-to-3D and AI texturing, but not text-to-3D. Its listed platforms are Linux, self-hosted, and web.