This extension for AUTOMATIC1111's Stable Diffusion WebUI enables the generation of 3D models from text prompts or images, leveraging OpenAI's Shap-E model. It is designed for users of the Stable Diffusion WebUI ecosystem looking to integrate 3D asset creation into their workflow.

How It Works

The extension integrates OpenAI's Shap-E, a diffusion model capable of generating 3D assets. Users can select either a text-to-3D or image-to-3D generation mode. For text-based generation, a prompt is provided. For image-based generation, an input image is used, with a recommendation to pre-process by removing the background for better results. Shap-E's models are downloaded automatically on the first run.

Quick Start & Requirements

• Install via the Stable Diffusion WebUI's "Extensions" tab by pasting the repository URL.
• Requires AUTOMATIC1111/stable-diffusion-webui.
• Shap-E models are downloaded automatically.

Highlighted Details

• Generates 3D models from text prompts or input images.
• Based on OpenAI's Shap-E model.
• Supports batch generation with output files saved in extensions\sd-webui-txt-img-to-3d-model\outputs.

Maintenance & Community

• Developed by jtydhr88.
• The developer also maintains other Stable Diffusion WebUI extensions: 3D Model/pose loader, Canvas Editor, StableStudio Adapter, and 3D Editor.

Licensing & Compatibility

• The README does not explicitly state a license. Compatibility with commercial or closed-source projects is undetermined.

Limitations & Caveats

The extension is a custom integration for a specific WebUI version and may not be compatible with other Stable Diffusion interfaces. The README notes that image-to-3D generation does not utilize text prompts, and vice-versa.