#13 · Primary category： 3D Generation & Asset Creation
MeshAnything
[ICLR 2025] From anything to mesh like human artists. Official impl. of "MeshAnything: Artist-Created Mesh Generation with Autoregressive Transformers"
Project last updated：04/28/25
GitHub Stars
2.3K
Forks
108
Contributors
6
License
Other
Why we included this project
MeshAnything turns rough 3D scans into clean polygon meshes with topology that resembles what a human artist would build. The official implementation of the ICLR 2025 paper, it uses an autoregressive transformer to predict mesh structure rather than relying on the noisy triangles typical reconstruction pipelines output. The repository includes training and inference code, pretrained weights, and a live demo, so you can test it on your own data quickly. Teams that care about downstream rigging or UV mapping will find the cleaner topology helpful, but the non-commercial S-Lab license means you should treat it as a research tool and plan accordingly before any product use.
Articles for this project
No articles for this project yet.
To suggest a topic or contribute an article, contact us.
Related projects in this category
modly
Desktop app to generate 3D models from images or prompt using local AI — runs entirely on your GPU
LichtFeld-Studio
Train, inspect, edit, automate, and export 3D Gaussian Splatting scenes from a single native application.
map-anything
MapAnything: Universal Feed-Forward Metric 3D Reconstruction
ComfyUI-3D-Pack
An extensive node suite that enables ComfyUI to process 3D inputs (Mesh & UV Texture, etc) using cutting edge algorithms (3DGS, NeRF, etc.)
dust3d
Dust3D is a cross-platform 3D modeling software that makes it easy to create low poly 3D models for video games, 3D printing, and more.