three.ws Forge for desktop
A 3D studio that runs on your machine.
Turn a photo into a 3D model with open-source generators on your own GPU, or hand it to the three.ws cloud when you have no GPU at all. Clean it up, run it through a node workflow, and publish it to your three.ws account in one click.
Free and open source. Based on Modly by Lightning Pixel (MIT license).
What is inside
-
Your GPU or ours
Install open-source generators such as Hunyuan3D, TripoSG and TRELLIS from the Extensions tab and run them locally, with no queue and no upload. The three.ws cloud generator is built in for machines without a GPU, and takes a photo, a text prompt or a sketch.
-
Node workflows
Chain generation, mesh repair, remeshing, smoothing, optimization, auto-rigging and export in a visual graph, save it, and run it again on the next image.
-
The CC0 library
Search the three.ws object library from the Import menu and drop a public-domain model into the viewport as a starting point, a reference or a workflow input.
-
Publish in one click
Sign in once through your browser. Publish sends the open model to your account as a shareable 3D page you can embed, rig or give to an agent.
All downloads
| Platform | File | Size | SHA-256 |
|---|
Opening an unsigned build
Run it from source
Needs Node.js 20 or later and Git. The first launch builds Forge's Python environment, which takes a few minutes.
git clone --depth 1 https://github.com/nirholas/three.ws
cd three.ws/apps/forge-desktop
npm install
npm run prepare-resources
npm run devprepare-resources downloads a standalone Python for your platform so Forge never depends on the one your system ships. The source and build guide covers packaging your own installer.
System requirements
- Windows
- Windows 10 or 11, 64-bit
- macOS
- macOS 12 Monterey or later on Apple silicon (M1 and newer)
- Linux
- 64-bit Ubuntu 20.04+, Debian 11+, Fedora 39+ or another current glibc distribution
- Local generation
- An NVIDIA GPU (CUDA), an AMD Radeon GPU (ROCm, on Windows and Linux) or Apple silicon. Each model in the Extensions tab lists its VRAM needs.
- Cloud generation
- Any of the above, no GPU needed. Uses your three.ws plan.