We run a 3D printing farm, so you might expect this post to conclude that 3D printing wins. It usually does under 1,000 units, but not always, and we tell people to go mold their parts more often than you would think. Here is the honest version of the comparison, the one we walk through with clients before quoting.

The economics in one paragraph

Injection molding has a brutal entry fee and a beautiful marginal cost. Cutting a steel or aluminum mold costs thousands of dollars before the first part exists, and then each additional part costs very little. 3D printing is the mirror image: no tooling at all, but every single part carries roughly the same material and machine time cost as the one before it. Plot those two lines and they cross somewhere, and everything about this decision is asking which side of the crossover your project sits on.

For most parts, at most sizes, that crossover sits well above 1,000 units. Below it, printing wins on total cost. But total cost is not the only axis, so keep reading before deciding.

Where 3D printing wins below 1,000 units

No tooling bill and no tooling wait. Your first part costs the same as your fiftieth, and production can start days after you send a file rather than weeks after mold trials.

Design changes stay cheap. If revision three of your product ships next quarter, a mold is a sunk cost, but a print profile is a file. Products that are still evolving are terrible candidates for tooling.

Complexity is free-ish. Internal channels, organic shapes, and consolidated assemblies that would need complex molds or multiple molded parts often print as one piece.

Small batches of many SKUs. A catalog of twenty products selling modest volumes each would need twenty molds. A print farm treats that catalog as one queue, which is exactly how our fulfillment clients run.

Where injection molding wins, even at low volume

This is the section most print shops skip.

Materials printing cannot match. Living hinges in polypropylene, optically clear parts, glass-filled engineering resins with certified datasheet properties, soft-touch overmolds. If your part depends on a property that only exists in a molding-grade material, no print process substitutes cleanly.

Cosmetic class surfaces. Molded parts come out of a polished tool with a finish printing cannot reach without heavy post-processing. If the part is the product and the product sells on looks, layer lines are a real cost.

Tiny parts at real volume. A thumbnail-sized clip that molds sixteen at a time in a two-second cycle beats printing economics far earlier than 1,000 units. Crossovers are geometry-specific, and small simple parts cross early.

True production certainty. A validated mold produces the same part on cycle ten thousand as cycle ten. If your program is committed to one frozen design at high ongoing volume, molding is the endgame and printing is the bridge.

When we tell people not to 3D print

We have told prospective clients to go get a mold quoted, and it usually happens in one of three situations. The design is frozen, the volume is confidently in five figures per year, and the part is simple enough that tooling amortizes quickly. The part needs a molded material property, most often a living hinge or optical clarity. Or the unit economics of a very small part simply favour cycle time over layer time. In each case we would rather say so at the quoting stage than sell a run that serves you badly, because companies remember who gave them straight advice.

The reverse referral happens too. Mold shops send people to printing when tooling money is not justified yet, and the healthiest product programs use both: print the first few hundred units, learn from the market, then tool up with a design you trust. The prototype and short-run phase is where printing is unbeatable, and it hands molding a better design than a straight-to-tooling program ever gets.

The bridge production pattern

The most useful version of this decision is often not either-or but when. Bridge production means printing your first production units while a mold is being quoted, cut, and validated, then switching processes once tooling is proven. It turns the mold’s lead time from a launch delay into background noise, and it means the design you eventually freeze into steel has survived hundreds of real customers rather than a focus group. Printed runs routinely surface design problems, an awkward assembly step or a fastener boss in the wrong place, that would have been expensive mold revisions if discovered after tooling. Printing first is cheap insurance on the mold you buy later, and it is the strongest argument for starting every new product on a printer even when molding is the known destination.

A practical way to decide

Ask three questions. Is the design still changing? If yes, print. Does the part need a property only molding materials deliver? If yes, mold, regardless of volume. Is lifetime volume under a few thousand units? If yes, printing almost certainly wins on cost, and you avoid betting tooling money on a forecast.

If you are in the ambiguous middle, send us the part with your expected quantities. We will quote the printed version within one business day, and if the honest answer is that your project belongs in a mold, we will say exactly that.