Jewelry CAD for the AI era.

Yes. This was research-grade in 2023 and is production-grade today for most ring and pendant categories. Current state of the art produces a printable STL or GLB from one or several photos in under a minute, and the output is good enough for casting on the first or second pass for most pieces. It is not yet good enough to skip a CAD jeweler's review on complex micro-pavé or unusual setting work.

For standard rings, pendants, and stud earrings: accurate enough to cast directly after a thickness check and a quick mesh review. For pieces with unusual settings, fine pavé, or interlocking parts: accurate enough as a starting point, with a CAD jeweler refining the last 10–20%. The floor of "what works on the first pass" rises every quarter.

Still the dominant parametric engine, and likely will be for the rest of the decade. Most serious jewelry CAD tools (including ours) are Rhino-backed underneath, because parametric edits (change a shank, swap a head, adjust a stone size) still require Rhino-grade geometry that a pure mesh tool cannot match. The shift is that you are less often opening Rhino directly; you are driving it through a layer that knows about jewelry.

One base 3DM, many variations. This is the workflow that has quietly become standard. From a single ring file, teams generate variations across shank style, stone shape and size, head type, and metal at the click of a button, then render each one. The studios shipping the most product are doing this for catalog launches; bespoke jewelers are doing it for client option sets.

Both, in that order. Photo-to-STL handles the cases where you need a manufacturable file fast and do not want to model from scratch. Manual CAD or parametric editing handles the cases where the piece is unusual enough that AI gets you 80% of the way and you need the last 20%. The jewelers ignoring CAD entirely are the ones whose practice still hits a ceiling.

Four cover most situations. STL for 3D printing. 3DM for parametric editing and most casting houses. OBJ for tools that do not read 3DM. GLB for web-based viewers and storefront customizers. If your CAD tool cannot export all four, you will hit a wall whenever you need to hand off to a vendor or a storefront.

The CAD jeweler's day used to be 60–70% modeling from scratch; today it is closer to 30%, with the rest spent on review, refinement, and parametric variations. The volume of work each CAD jeweler can ship has roughly doubled. The shops we see thriving have shifted billing from per-hour modeling to per-piece or per-collection, because the value they deliver is no longer time-on-keyboard.

Yes. The photo→STL workflow takes one or several photos of a piece and produces a printable STL or GLB. You get free retries until the mesh is right.

Yes. The CAD environment is Rhino-backed and ships with jewelry-specific libraries. You can swap settings, shanks, stones, and prongs while keeping the rest of the file intact. Full parametric release is scheduled for June 2026.

Mesh evaluation, thickness checks, undercut detection, layer organization, geometry repair, and clean exports per format. The goal is a hand-off file your casting house or printer accepts on the first try, so you stop paying for the round-trip.

CAD feeds directly into the photoshoot tools (drop a 3DM or STL in as the seed; the geometry is preserved) and into the customizer tools (one 3DM, swappable settings/shanks/stones/prongs, rendered on demand for shoppers). The same file powers both your catalog imagery and your on-store configurator.