Image to STL methods
AI image to STL vs heightmap — pick the right method
Searching image to STL turns up tools that do very different jobs. Some map brightness to height (relief / lithophane). Some extrude a silhouette. Others run single-image AI image to 3D and export a mesh. PhotoSTL is in that third group: one photo → print recipe → STL/3MF in millimeters. 30 free credits (≈3 drafts) when you sign in (Google or magic link).
Three routes from image to STL
| Method | Best for | Catch |
|---|---|---|
| Heightmap / lithophane | Backlit panels, shallow photo reliefs | Flat panel — no reconstructed back or undercuts |
| Silhouette extrude | Logos, icons, stamps, cookie-cutter outlines | Constant thickness — not a 3D object from a photo |
| AI full 3D (PhotoSTL) | Figurines, busts, product props, miniatures | Back is inferred; finish walls in your slicer |
Why many “AI converters” are still heightmaps
Browser tools often label themselves AI while still reading luminance as height. That is useful for lithophanes and embossed plaques. It is the wrong tool when you need a pet bust, a tabletop miniature, or a product you can rotate. If the download looks like a raised photo on a flat plate, you used a heightmap path — not full reconstruction.
What PhotoSTL actually does
- You upload one clear JPG, PNG, or WebP and pick a print recipe (general, pet, logo, miniature, jewelry, product, architecture).
- The pipeline runs image to 3D in about 30 seconds (varies with queue), then prepares a print draft in millimeters.
- You download STL or 3MF (plus raw GLB for preview) and finish walls / supports in your slicer.
We do not offer a local heightmap or cookie-cutter mode. If that is your job, use a dedicated relief tool; if you need a mesh object from a photo, start on photostl.com.
FAQ
What is the difference between heightmap and AI image to STL?
A heightmap maps pixel brightness to Z height — you get a relief or lithophane panel, not a free-standing object with a reconstructed back. Single-image AI image to 3D estimates depth and unseen sides, then exports a mesh you can download as STL. PhotoSTL is built for that AI path, then prepares millimeter STL/3MF for slicers.
When should I use a heightmap or lithophane tool instead?
Use a dedicated heightmap or lithophane converter when you want a flat panel lit from behind, a shallow photo relief, or a cookie-cutter silhouette from line art. PhotoSTL does not claim those pipelines — we reconstruct a full mesh draft from one photo for makers who need something closer to a 3D object.
Does PhotoSTL invent geometry that was not in the photo?
Yes, within limits. The model infers the back and sides from one view. That is useful for figurines, product props, and busts, but it is still a guess — check thin walls and supports in your slicer before you print.
Is there a free way to try AI image to STL?
30 free credits (≈3 drafts) when you sign in (Google or magic link) on photostl.com. After that, Hobby starts at $9 for 1,500 credits (~$0.06 per draft on our price list).