Image to 3D converts photos, text descriptions, or multiple reference angles into fully-textured 3D models you can import directly into Blender, Maya, Unreal Engine, or Unity. The output is a GLB file with physically based rendering materials and an interactive 3D viewer inside Claude.
Turning a product photo or concept sketch into a 3D asset used to take hours of modeling work. This tool does it in 1-3 minutes, producing game-ready or high-fidelity hero assets depending on the model you choose.
What you can do
- image_to_3d — generate a 3D model from a single photo, with choice of high-fidelity PBR or game-ready low-poly output
- text_to_3d — generate a 3D model from a text description — describe the object, material, and style
- multiview_to_3d — produce a higher-accuracy model from 2-4 photos of the same object from different angles
Who it's for
Game developers creating props and characters, product designers visualizing concepts in 3D, e-commerce teams building product configurators, and architects or designers creating object assets for spatial rendering.
How to use it
- Use image_to_3d with a single photo URL for most objects — choose H3.1 for maximum fidelity with PBR textures or P1 for game-ready topology
- Use text_to_3d with a descriptive prompt for objects you don't have a photo of
- Use multiview_to_3d for best results — provide front, left, back, and right views if possible
- Set texture_quality and geometry_quality to "detailed" on H3.1 for print or hero-asset quality; add quad:true for rigging-ready meshes
Getting started
Connect your account to activate 3D generation — models take 1-3 minutes and return a GLB you can view interactively before downloading.
