Free Photo to STL Converter

From a picture to a printable STL

Most "image to STL" converters only make a flat relief or a lithophane: the picture is pressed into a slab, darker pixels become thicker. This tool does something different. The AI looks at the photo, works out the full 3D shape of the object in it, and builds a complete model you can hold in your hand after printing: the front, the sides and the back. You then set the real-world size and download an STL that any slicer opens, such as Cura, PrusaSlicer, Bambu Studio or OrcaSlicer.

How to convert a photo to STL

1. Upload a photo (PNG, JPG or WEBP, up to 10 MB) of a single object: a toy, a figurine, a character, a pet, a sculpture or a product. A plain background works best.
2. Press Generate. The page is already set to the Detailed style with 50,000 triangles, which keeps fine features for printing. You can lower it if you want a smaller file.
3. Check the preview. Rotate the model. If the shape is wrong, a different photo usually helps more than changing settings. If it is only rough, try a higher quality level.
4. Set the size in millimetres under "3D print" and press STL. The longest side of the model is scaled to that size.
5. Download the STL and open it in your slicer. The status line tells you whether the mesh is watertight (closed), which is what slicers want.

Tips for a clean print

• Choose the angle carefully. A three-quarter view shows depth better than a flat front view. Extra photos from the left, back and right (under "More angles") help when the back of the object matters.
• Think about thin parts. Ears, tails, swords and antennae can come out thinner than your printer can handle. Scale the model up, or thicken those parts in your slicer or in Blender.
• Plan supports. Characters usually need supports under arms and chins. Printing the model lying on its back sometimes avoids most of them.
• If the STL is not watertight, most slicers can repair small holes automatically. You can also press Generate again: each run is a little different.

For wall thickness, orientation and slicer settings in more depth, read preparing an AI-generated model for 3D printing. If you are unsure which format you need, see STL vs GLB vs OBJ.

What works and what does not

Single, solid objects give the best results: toys, figures, vases, shoes, fruit, small statues and cartoon characters. Busy scenes, landscapes, groups of people, glass, mirrors and very thin wire-like objects are hard for any single-photo AI. Read how to choose the best photo before you start, it makes the biggest difference.

Questions