Desktop app to generate 3D models from images or prompt using local AI — runs entirely on your GPU

★ 8,529⑂ 795👁 60▤ 14
TypeScriptCreated 2026-03-18modly3d.app
8,529
stars
+0+10+595
795
forks
60
watchers
–
prs
–
issues
14
releases
–
discussions

README

Modly

Local, open source, AI-powered image-to-3D mesh generation.
Turn any photo into a 3D model using open source AI models running entirely on your GPU.
Modly is a desktop application for Windows, Linux, and Apple Silicon macOS.

 Created by Lightning Pixel


  


---


 Download

Head to the Releases page to download the latest installer for Windows, Linux, or Apple Silicon macOS.

Alternatively, you can clone the repository and run the app directly without installing:

bash
 Windows
launch.bat

 Linux / macOS
./launch.sh


---



 Getting started

 1. Install JS dependencies

bash
npm install


 2. Set up Python backend

bash
cd api
python -m venv .venv
.venv\Scripts\activate      Windows
source .venv/bin/activate   Linux / macOS
pip install -r requirements.txt


 3. Run in development

bash
npm run dev


 4. Test

bash
npm test
./node_modules/.bin/tsc --noEmit -p tsconfig.node.json
npm run build


 Platform notes

- AMD GPUs are supported through ROCm on Linux and Windows: a Radeon card is detected
  automatically and extensions are steered to ROCm PyTorch wheels, with no ROCm install
  required. See docs/running-on-amd-rocm.md.
- macOS support targets Apple Silicon only.
- macOS uses native window controls. Windows and Linux keep the existing custom controls.
- The top bar includes a live RAM indicator sourced from the main process.
- Workflow wiring is validated before run; invalid graphs stay in place and surface inline/toast warnings instead of dropping the current mesh view.
- Package Apple Silicon macOS with npm run package:mac.
- Imported meshes can be smoothed and decimated in-app; optimized results are written back into the workspace.

---

 Extension system

Modly supports external model and process extensions. Each extension is a GitHub repository containing a manifest.json plus the runtime entry files required by its type.

 Official extensions

| Extension | Model | URL |
|-----------|-------|-----|
| modly-hunyuan3d-mini-extension | Hunyuan3D 2 Mini | https://github.com/lightningpixel/modly-hunyuan3d-mini-extension |
| modly-hunyuan3d-mini-turbo-extension | Hunyuan3D 2 Mini Turbo | https://github.com/lightningpixel/modly-hunyuan3d-mini-turbo-extension |
| modly-hunyuan3d-mini-fast-extension | Hunyuan3D 2 Mini Fast | https://github.com/lightningpixel/modly-hunyuan3d-mini-fast-extension |
| modly-triposg-extension | TripoSG | https://github.com/lightningpixel/modly-triposg-extension |
| modly-trellis2-gguf-extension | Trellis2 GGUF | https://github.com/lightningpixel/modly-trellis2-gguf-extension |

 How to install an extension

1. Go to the Models page and click Install from GitHub.

!Install from GitHub

2. Enter the HTTPS URL of the extension repository and confirm.

!Enter extension URL

3. If the extension exposes model nodes, download the model or one of its variants. Process extensions are ready once installation and setup complete.

!Install models

 Multiple Hugging Face repositories per model node

A model node whose w

Snapshot: 2026-11-04T14:00:00Z · Data license Apache-2.0 · JSON