AI Image to PBR Material Generator
Turn a single surface photo into albedo, normal, and relative-height maps.
Generation Queue
What this tool does
Image to PBR uses Marigold V2 to directly predict albedo, normal, and relative height. The maps are material starting points to review and refine in your DCC or game engine.
- Generate three coordinated PBR maps from one source
- Use direct Marigold V2 predictions for each map
- Accelerate material blocking for games, visualization, and 3D scenes
How to use it
- 1Upload a sharp, front-facing surface image.
- 2Choose seamless processing if needed for the target workflow.
- 3Generate the PBR set, inspect each map, and download the outputs.
Frequently asked questions
Which maps does Image to PBR create?
It creates albedo, normal, and grayscale relative-height maps. Ambient occlusion, roughness, and metalness are not generated and should be authored separately.
Is the generated material physically exact?
No. The maps are AI estimates from one image, so hidden geometry and real reflectance cannot be measured exactly. Treat the output as a strong starting point and validate it under varied lighting.
How should I check an AI-generated PBR material?
Preview the maps together under several neutral lights and inspect the material at its intended scale. Check that large forms read correctly in the normal and height maps, and that the albedo no longer contains distracting baked illumination.