LibreAI / AI 3D / modly
// ai-3d · lightningpixel/modly

modly

Modly — a desktop app that generates 3D models from images or prompts, running entirely on your own GPU.

★ 8.0k stars⑂ 739 forksMITupdated 2026-10
▦ visual — this project's README ships no embeddable screenshot,
so here's a placeholder. The guide below is complete regardless.
// project facts

Project facts

modly is an open-source ai 3d project by lightningpixel (lightningpixel/modly on GitHub). It has 7,964 GitHub stars and 739 forks, is released under the MIT license, and was last updated on GitHub on 4 October 2026.

GitHub stars
7,964
Forks
739
License
MIT
Last updated on GitHub
4 October 2026
Category
AI 3D
Listing
Full install guide
Official site
Official site
// how it works

How it works

  1. Clarify your goal

    Check your GPU (NVIDIA, or AMD via ROCm) and decide whether to take the release installer or run from source for development.

  2. Get the project

    Clone lightningpixel/modly from GitHub (or use the install path in the run steps below) and skim its README for the parts relevant to your goal.

  3. Install and configure

    Download the latest installer from GitHub Releases for Windows, Linux or Apple Silicon macOS — or clone and launch with the provided scripts.

  4. Run your first task

    Generate a first model from a single image, inspect the mesh in the viewer, then smooth or decimate it before exporting.

// features

Why people love it

Local GPU generation

3D generation runs entirely on your own GPU — no cloud generation service and no per-model fees to a host.

Image or prompt input

Create models from a reference image or from a text prompt, depending on the model extension you install.

Extension system

Models are installable extensions; one model node can pull multiple Hugging Face repositories and separately installable weight variants, with shared weights inside an extension.

Workflow wiring with validation

Node workflows are validated before a run; invalid graphs stay in place and surface inline/toast warnings instead of losing your mesh view.

Mesh cleanup in-app

Imported meshes can be smoothed and decimated inside the app, with optimised results written back into the workspace.

Cross-platform desktop

Installers for Windows, Linux and Apple Silicon macOS, AMD ROCm support, and a live RAM indicator while you work.

Advertisement
// run it

Run it yourself

  1. Clone the repository

    Get the source from the official GitHub repository for modly.

    git clone https://github.com/lightningpixel/modly.git
    cd modly
  2. Get the app

    Download the latest installer for your platform from the GitHub Releases page, or clone the repo to run without installing (launch.bat on Windows, ./launch.sh on Linux/macOS).

    git clone https://github.com/lightningpixel/modly.git
    cd modly
  3. Run from source

    Install the JavaScript dependencies, set up the Python backend, then run in development as the README's getting-started steps describe.

    npm install
    pip install -r requirements.txt  # then run the dev command — see README
// use cases

What you can build

Advertisement
// faq

Questions, answered

Who maintains modly?

lightningpixel, through the lightningpixel/modly repository, with the site at modly3d.app.

Is modly free to use?

Yes — the project is open source under the MIT license as recorded for this listing. You can use, modify and distribute it under the MIT terms.

Does modly need a paid API key?

No API key is needed — generation is local on your GPU, and model extensions download their weights from Hugging Face; Modly itself is free.

How does Modly compare to a hosted image-to-3D service?

Hosted services generate on their GPUs, charge or rate-limit, and see your images. Modly generates on your own GPU with installable local models — you need the hardware, but generation stays private and unlimited by a plan.

How do I get started with modly?

Clone the repository (or use the documented installer), follow the install and configure steps on this page, and run the smallest real task first — generate a first model from a single image, inspect the mesh in the viewer, then smooth or decimate it before exporting.

// keep exploring

Similar tools

AI 3D

360Extractor

High-performance desktop application for 360° video preprocessing.

★ 154 starsAGPL-3.0
View guide →
AI 3D

Hunyuan3D-2-WinPortable

📦Portable package for running Hunyuan3D 2.0/2.1 on Windows. | 混元 3D 2.0/2.1 整合包

★ 1.3k starsGPL-3.0
View guide →
AI 3D

text-to-cad

Give your AI agent CAD superpowers - generate real STEP/STL models from text

★ 18.4k starsMIT
View guide →
AI 3D

depthsplat

Feed-forward 3D from images: Gaussian splatting meets depth.

★ 1.3k starsMIT
View guide →
AI 3D

Hunyuan3D-2.1

From Images to High-Fidelity 3D Assets with Production-Ready PBR Material

★ 4.1k starsUnknown
View guide →
AI 3D

ooosplat

One-click desktop app turning videos into 3D Gaussian Splats.

★ 1.8k starsApache-2.0
View guide →
// alternatives

Alternatives in AI 3D

The most-starred other AI 3D projects in the LibreAI catalogue — the realistic shortlist if modly is not the right fit.

Get modly

Free and open — grab the source, star the repo, and start creating.