InstantMesh
Open-source framework that turns a single image into a 3D mesh in seconds.
About
InstantMesh is Tencent ARC Lab's open-source feed-forward framework for image-to-3D generation. It pairs a multi-view diffusion model with a sparse-view large reconstruction model to produce diverse 3D assets from one photo in about 10 seconds, with code, weights, and a demo released on GitHub under Apache 2.0.
It is a research-grade tool aimed at developers and 3D researchers rather than casual users: running it needs a GPU and Python setup. The paper notes real limits too, including a 64x64 triplane resolution bottleneck and weaker results on tiny thin structures. Still, it is one of the fastest open paths from a single image to a usable mesh.
What it can do
- Single-image to 3D mesh in about 10 seconds
- Sparse-view LRM reconstruction architecture
- Differentiable isosurface mesh extraction
- Apache 2.0 open-source code and weights
- Hugging Face demo and model card
Setup
Clone github.com/TencentARC/InstantMesh, install dependencies, and run the demo or inference script on a GPU machine.
Prerequisites
Python environment with PyTorch and a CUDA-capable GPU.
Use cases
Works with
Trust score — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 85; Trending score 71/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.