Creating a 3D model from an image is now easier than ever, thanks to AI tools like ComfyUI and Hunyuan3D 2.0. In a recent tutorial, Adán Martín shows how to turn a single image into a textured 3D object — without traditional modeling or photogrammetry. This is a game-changer for architectural visualization artists looking to speed up asset creation.

What You Need to Get Started

• A centered, well-lit image of the object
• ComfyUI installed and configured
• The Hunyuan3D 2.0 model loaded as a custom node
• Basic knowledge of image prompts and node-based workflows

How to Create a 3D Model from an Image

1. Load Your Image
   Choose a photo with minimal background clutter and good contrast.
2. Use ComfyUI for Image-to-Depth Mapping
   Combine Stable Diffusion and depth estimation nodes to guide reconstruction.
3. Run Hunyuan3D 2.0
   This model infers geometry and texture directly from the image, generating a clean .obj file with UVs and texture baked in.
4. Preview and Export
   Open the result in a mesh viewer or import it into 3ds Max or Blender. The model is ready for lighting and integration into your archviz scene.

This workflow helps you generate 3D models from images quickly — great for props, concept designs, or architectural ornamentation.

Don’t forget to check out more video by Adán Martín here.

And if you would like to explore more about creating 3D models from images, take a look at MeshyAI and Rodin.