Image to 3D model: one photograph, one mesh
Turn a photo or picture into a 3D model. Upload one image — a photograph, a product render or concept art — and get a textured mesh back. The most predictable route into 3D, because you choose the silhouette instead of asking a model to invent one.
What it does
Your subject, not an interpretation
Reconstruction works out the shape of the object in your picture rather than inventing an object from a description.
Works with what you have
A photograph, a render, a concept sheet, a product shot. If you can picture it, you can reconstruct it.
Background removal built in
A backdrop gives the reconstructor edges that are not part of your object. Removing it is usually the cheapest quality improvement available.
Four engines
Run the same image through Zetta N1, N2.5, Np1 or Np1.5 — different reconstructions, and often a faster fix than a better photograph.
How it works
- Pick a good photographEven diffuse light, a plain background, the whole object in frame, and a three-quarter angle rather than straight on.
- Remove the backgroundOn by default, and worth leaving on for anything shot against a surface.
- ReconstructTextured output, with PBR materials available on Zetta N1.
- RefineA reconstruction is dense and machine-built. Retopology is what makes it usable.
What people use it for
- Digitising a real object you can photograph
- Turning concept art or a product render into geometry
- Rebuilding an asset you have a picture of but not a file for
- Catalogue and visualisation work where the subject already exists
What it cannot do
- The back is invention. Symmetrical objects survive it; distinct rears do not
- Transparent and reflective subjects reconstruct badly — the camera recorded the reflection
- Hard shadows read as shape and can reconstruct as creases
- Wide-angle distortion is inherited faithfully by the mesh
- Cropped edges become cut-off geometry, not inferred continuation
The details
- Input
- One image
- Models
- Zetta N1 and Zetta N2.5
- Background removal
- Available, on by default
- Output
- GLB, with OBJ and FBX available
Common questions
Can I turn a 2D picture into a 3D model?
Yes — that is what this does. One picture of one object, whether a photograph, a product render or a piece of concept art, is reconstructed into a textured 3D mesh. The side the picture did not show is inferred, so add reference views with multi-view to 3D when the back matters.
Multi-view to 3DIs image to 3D free?
You can start for free. The free plan includes image to 3D on Zetta N1, with a one-off grant of credits on sign-up and no card. The finishing tools — retopology, texturing and rigging — and the other engines come with a paid plan. Exporting as GLB, FBX, OBJ or STL is free.
See pricingWhat kind of photo makes the best 3D model?
Even, diffuse light, a plain background, the whole object in frame and a three-quarter angle rather than straight on. Background removal is on by default, and it is usually the cheapest quality improvement there is.
Can I 3D print a model made from a photo?
Yes. Generate the model, then export it as an STL from the Studio — free, on every plan — and take it to your slicer.
Image to STLLast reviewed