Modly brings AI generation of 3D models to the PC: free, open source and completely offline
A desktop alternative to cloud services for generating three-dimensional meshes via artificial intelligence arrives. The open source project allows transforming images and texts into 3D models using only the local GPU.
Most AI generators of 3D models work through cloud platforms. The user uploads an image, the remote server processes it and returns the result. Modly, developed by Lightning Pixel, proposes a radically different approach: bringing the entire process to the user's computer.
The application allows generating three-dimensional meshes from images or textual prompts. It works completely offline once the program and necessary models are installed. It requires no credits or subscriptions.
- Open source software with MIT license for local generation of 3D models
- Completely offline processing via local GPU
- Support for multiple open source AI models (TripoSG, Trellis2, Hunyuan3D 2 Mini)
- Compatible with Blender, Unity, Unreal Engine and 3D printing
A unified environment for open source AI models
Modly does not introduce a new proprietary generative model. Its value lies in accessibility: it makes various open source AI models usable without complex configurations.
The project works as a desktop environment that eliminates the need to manually configure Python environments, CUDA dependencies, and different repositories. Users can install extensions to integrate models such as TripoSG, Trellis2, and Hunyuan3D 2 Mini.
The source code is distributed under the MIT license. It can be modified, integrated into other projects, or adapted to specific workflows while respecting the license terms.
Simplified workflow: from image to mesh
The generation process requires few steps. The user provides a visual or textual input, selects the AI model, and starts local processing.
The workflow accepts images in PNG, JPG, or WebP format, or textual descriptions. After choosing one of the installed AI models, processing is performed by the local GPU. The resulting three-dimensional model appears directly in the Modly interface.
The results can be exported to modeling software such as Blender, game engines like Unity or Unreal Engine, or prepared for 3D printing.
The process is not equivalent to a metrological three-dimensional scan. With a single photograph, the AI only knows what is visible in the image and must estimate the geometry of non-visible parts.
Hardware requirements
article written with the help of artificial intelligence systems
Q&A
What is Modly and how does it work?
Modly is an open-source desktop application that generates 3D meshes from images or text prompts using the local GPU. It works completely offline without requiring subscriptions or cloud credits.
Which AI models does Modly support?
The software acts as a unified environment to integrate various open-source models such as TripoSG, Trellis2, and Hunyuan3D 2 Mini. Users can install specific extensions to utilize these generative engines.
Can Modly be used without an internet connection?
Yes, once the program and necessary models are installed, the entire generation process occurs offline on the user's computer. There is no dependency on remote servers for data processing.
In what formats can created models be exported?
Results can be exported to modeling software like Blender, graphics engines such as Unity and Unreal Engine. They are also suitable for preparation and use in 3D printing.
What is Modly's distribution license?
The source code is distributed under the MIT license, allowing users to modify, integrate, and adapt the software to specific workflows. This license guarantees maximum freedom of use while respecting standard conditions.
