- Home
- AI 3D Model Generators
- TripoSR
TripoSR
What it is
TripoSR is a free, open-source model for reconstructing a 3D object from a single image. Tripo AI and Stability AI developed it collaboratively for researchers, developers, and creatives working in 3D generative AI and content creation. The project reports generation in less than 0.5 seconds on an NVIDIA A100 GPU. Its repository includes a command-line inference script and a local Gradio app launched with python gradio_app.py. The app provides background removal, foreground-ratio and marching-cubes-resolution controls, plus OBJ and GLB output. Inference accepts one or more image paths and can bake a texture instead of using vertex colors, with configurable texture resolution. Local setup requires Python 3.8 or later, PyTorch, and dependencies from requirements.txt; CUDA is optional when available. Default inference uses about 6 GB of VRAM for one image. The code, pretrained models, and interactive demo use the MIT license. TripoSR supports image-to-3D and AI texturing, but not text-to-3D.
For and against
For
- Single-image 3D reconstruction
- Local command-line and Gradio interfaces
- OBJ and GLB output
- Configurable baked textures
- MIT-licensed code and models
Against
- Default inference uses about 6 GB VRAM
- Requires Python, PyTorch, and dependencies
- No text-to-3D support
- CUDA is optional, not guaranteed
The verdict
TripoSR offers a free MIT-licensed workflow for turning one image into a 3D object, with local controls and common export formats. Its hardware and setup requirements make it better suited to technically prepared users.
Who it suits
It suits researchers, developers, and creatives who need local or self-hosted image-to-3D reconstruction. Users should be comfortable with Python setup and GPU memory requirements.
Plans and prices
TripoSR
MIT-licensed model and code · default inference uses about 6 GB VRAM per image
Read 8 Oct 2026 · github.com
On the record
- Demo features
- The Gradio app offers background removal, foreground-ratio and marching-cubes-resolution controls, and OBJ and GLB outputs.github.com · read 8 Oct 2026
- Hardware requirement
- Default inference options use about 6 GB of VRAM for a single image input.github.com · read 8 Oct 2026
- How it works
- The model generates a 3D model in less than 0.5 seconds on an NVIDIA A100 GPU, according to its README.github.com · read 8 Oct 2026
- Input
- The inference script accepts one or more image paths.github.com · read 8 Oct 2026
- Intended users
- The project describes its intended audience as researchers, developers, and creatives working in 3D generative AI and 3D content creation.github.com · read 8 Oct 2026
- License
- The source code and pretrained models are released under the MIT license.github.com · read 8 Oct 2026
- License holders
- The MIT license file names Tripo AI and Stability AI in its 2024 copyright notice.github.com · read 8 Oct 2026
- Local installation
- The README lists Python 3.8 or later, PyTorch, and dependencies installed from requirements.txt; CUDA is optional if available.github.com · read 8 Oct 2026
- Local interface
- The repository includes a Gradio app that can be started with python gradio_app.py.github.com · read 8 Oct 2026
- Memory requirement
- The default inference options use about 6 GB of VRAM for one image input.github.com · read 8 Oct 2026
- Output options
- The inference script can bake a texture instead of using vertex colors and lets users specify texture resolution.github.com · read 8 Oct 2026
- Pricing
- The TripoSR repository describes the model as open source under MIT, but does not state a subscription price for TripoSR.github.com · read 8 Oct 2026
- Product identity
- TripoSR is an open-source model for fast 3D reconstruction from a single image, developed collaboratively by Tripo AI and Stability AI.github.com · read 8 Oct 2026
- Requirements
- The setup instructions list Python 3.8 or newer and recommend installing CUDA if available.github.com · read 8 Oct 2026
- Run locally
- The repository provides a command-line inference script and a local Gradio app.github.com · read 8 Oct 2026
- Single-image input
- TripoSR reconstructs 3D objects from a single image.github.com · read 8 Oct 2026
- Speed
- The project says it can generate a 3D model in less than 0.5 seconds on an NVIDIA A100 GPU.github.com · read 8 Oct 2026
- Texture output
- Inference can output a texture instead of vertex colors using the --bake-texture option, with texture resolution configurable.github.com · read 8 Oct 2026
