TripoSR: 3D generator — what it does and what you need to run it

TripoSR is an open-source tool for quickly creating 3D objects from a single image. It tackles the image-to-3D task, i.e., it turns a flat picture into a volumetric model, and is geared toward high processing speed.
What it does
As described by the author, TripoSR reconstructs the 3D geometry of an object from a single input image. The authors emphasize reconstruction speed, which is reflected in the project's name. Details about supported output model formats and textures are not disclosed in the fact sheet.
What you need to run it
- Platform: as described by the author, CUDA is required. This means an NVIDIA GPU is necessary.
- VRAM: the authors state that processing a single image with default parameters requires about 6 GB of VRAM.
- Language and environment: the code is written in Python. To install dependencies, the authors recommend installing CUDA if it is available on the system.
- Model files: the model weights take up 1.6 GB. The largest file also weighs 1.6 GB.
- License: both the code and the weights are distributed under the MIT license. This permits free use, including commercial use.
- Hosting: the code is available in the repository github.com/VAST-AI-Research/TripoSR, and the model weights are on huggingface.co/stabilityai/TripoSR.
Who it is for
TripoSR will be useful for developers and researchers who need fast draft reconstruction of 3D objects from images and are willing to run Python scripts on a local machine with an NVIDIA GPU. The permissive MIT license allows integrating this generator into your own projects without restrictions.
The tool is not suitable for users without a discrete NVIDIA GPU with less than 6 GB of memory. It is also not designed for those expecting a ready-made application with a graphical interface — work is done via the command line and Python scripts. The authors do not provide a cloud version, and running TripoSR locally on computers with AMD or Intel GPUs is not possible unless they support CUDA.
TripoSR is a compact tool with clear system requirements. Its main feature, as stated by the creators, is the speed of getting a result from a single snapshot, which can be critical in tasks where processing time matters more than detail.
Fact sheet
Repository · Model on HuggingFace
| Task | image-to-3d source |
|---|---|
| Code license | MIT source |
| Weights license | mit source |
| Platform | CUDA per the author’s description source |
| VRAM | 6 GB per the author’s description source |
| Largest weights file | 1.6 GB source |
| All weights files | 1.6 GB source |
| Language | Python source |
| Last code change | 2026-06-04 source |
| Repository created | 2024-02-07 source |
| GitHub stars | 6806 source |
|---|---|
| Forks | 874 source |
| Downloads per month | 166842 source |
| Model updated | 2024-08-09 source |
Values are collected automatically from official sources and were checked on 2026-08-19. Each one links to its source, and values taken from the developer’s pages also carry a verbatim quote — hover over the note. Pricing and versions are shown as of the check date and change most often; verify on the vendor’s site before buying.



