At what quantity does urethane casting beat 3D printing?

Usually somewhere between 25 and 100 identical parts, though the crossover depends on part size and finish requirements more than on the number itself. Below about 20 units, printing almost always wins because the silicone tool never pays back. Above roughly 500, the tool count gets awkward — a silicone mold yields around 20 to 25 usable parts, so 500 parts means 20-plus molds and the economics start pointing at rapid aluminium tooling instead. Run your own numbers: divide the quoted tooling cost by the per-part saving to get the quantity where casting overtakes printing.

Do I still need 3D printing if I choose urethane casting?

Yes — the master pattern is almost always 3D printed. A casting shop needs a physical part to build the silicone tool around, and that master is typically printed on SLA or a fine-layer FDM machine, then sanded, filled and polished to the finish you want the cast parts to carry. The mold copies the master exactly, including its flaws, so the printing and finishing work is the quality ceiling for the whole run. That is why the two processes are usually bought together rather than chosen between.

How accurate are cast urethane parts?

Shops typically quote around ±0.15% of nominal dimension with a floor near ±0.25mm on small features, which is looser than a well-controlled printing process and much looser than machining. Two effects drive it: the polyurethane shrinks slightly as it cures, and the silicone tool relaxes a little with every pull, so part 20 is not dimensionally identical to part 1. For cosmetic housings and enclosures that is invisible. For mating features, press fits or anything that has to hold a tolerance across the whole batch, say so on the quote and expect the shop to tool around it.

Can urethane casting produce clear or rubber-like parts?

This is where it beats printing outright. Cast polyurethanes cover roughly 20 Shore A to 80 Shore D in one process family, so a soft gasket, an overmoulded grip and a rigid housing can all come from the same shop in the same week. Optically clear polyurethane polishes to genuine transparency, which no current printing process matches without heavy post-processing. There are also filled grades that mimic the stiffness and heat resistance of ABS, PP and polycarbonate closely enough for functional testing and trade-show units.

How long does a urethane casting run take?

Plan on one to two weeks for a first batch, against days for printed parts. The sequence is: print and finish the master (one to three days), build and cure the silicone tool (one to two days), then cast, demould and post-cure the parts (roughly one cycle per mold per day). Adding molds buys throughput but not a shorter start, because the tool has to exist before any part does. If your deadline is inside a week, print the batch and treat casting as the follow-on run.

What files does a casting shop need from me?

The same files a printing service needs — a watertight STL or, better, a STEP file with the original CAD geometry — plus one thing printing does not require: a decision about parting lines and gate locations. The shop will pick them if you don't, and the parting line ends up wherever it is easiest to split the tool, which may be across a visible face. Mark the cosmetic surfaces on the drawing. Our guide at /blog/how-to-prepare-files-for-a-3d-printing-quote covers the file side; the parting-line conversation is a short email on top of it.

Urethane Casting vs 3D Printing for Low-Volume Parts

If you need 25 to 500 identical plastic parts, you are in the one volume band where the process choice is genuinely close. Urethane casting vs 3D printing is the decision, and it turns on what the parts are for far more than on the quantity — a cosmetic housing for a trade show and a functional bracket for a test rig at the same quantity go to different processes. When you're ready to quote either one, the 3D Prototyping Hub directory lists shops that run both.

This guide covers where the cost sits in each process, what the lead times really look like, and the two or three part characteristics that settle the question outright.

What Urethane Casting Actually Is

Urethane casting — also sold as vacuum casting, cast urethane or RTV molding — is a soft-tooling process, and the tool is silicone rather than steel.

The sequence: you supply a master pattern, which is almost always 3D printed and then sanded, filled and polished. The shop suspends that master in a frame, pours liquid silicone around it and cures it, then cuts the cured block along a parting line and removes the master. That leaves a cavity that is an exact negative of your part — including every scratch and layer line you left on the master. Liquid polyurethane is then degassed, poured or vacuum-drawn into the cavity, cured, demoulded and post-cured.

Two consequences follow from that, and they explain almost everything else on this page. First, the master is 3D printed, so this is rarely a choice between printing and casting — it is a choice about whether printing makes the parts or makes the pattern. Second, the tool is consumable. A silicone mold yields roughly 20 to 25 good parts before the cavity degrades, so a 200-part order means the shop builds and manages around ten tools.

Where the Cost Lives

Printing has no tooling cost. You pay per part, and that per-part price is roughly flat whether you order 5 or 500 — each part takes its own machine time and its own material. How much 3D printing costs breaks down the drivers.

Casting front-loads cost into the silicone tool, then charges a lower per-part price on top. Tooling for a small-to-medium part typically quotes in the high hundreds to low thousands of dollars, and per-part prices commonly land in the tens of dollars. The exact figures move with part size, undercut complexity, the number of tools your quantity requires and the material grade.

The arithmetic is the same one that governs injection molding vs 3D printing, just with a much smaller and much softer tool:

Breakeven quantity = tooling cost ÷ (printed cost per part − cast cost per part)

Two things make the casting side of that equation behave differently from a steel mold, and both cut the same way. The tool is cheap, so the breakeven lands low — often in the 25-to-100 range rather than the 500-to-2,000 range injection molding needs. But the tool also wears out, so tooling cost is not genuinely one-time above about 25 parts; it steps up with every additional mold. That step is what makes casting stop winning again somewhere in the high hundreds, where rapid tooling in aluminium takes over.

Lead Time: Days vs Weeks

	3D printing	Urethane casting
Time to first part	1–3 days	1–2 weeks
Tooling	None	Silicone tool, ~20–25 pulls
Typical economic band	1–50 parts	25–500 parts
Per-part cost curve	Flat	Falls, then steps up per mold
Design changes mid-run	Free	New master, new tool
Best at	Geometry, iteration, speed	Finish, soft materials, clear parts

The lead-time gap is structural, not a queue problem. The tool has to exist before the first part does, and no amount of paying for rush service removes the master-print, mold-cure and post-cure steps. Adding molds buys throughput once the run has started; it does not move the start.

That is also the practical answer when a deadline collides with a quantity. Print the units you need this week, and start the casting tool in parallel for the units you need next month — the master you print is work you were going to do anyway.

Where Casting Wins Outright

Volume is not usually what decides it. These three characteristics are:

Soft and rubber-like parts. Cast polyurethanes span roughly 20 Shore A to 80 Shore D. Printed TPU is capable but slow, limited in durometer range, and rarely matches a cast gasket for surface quality.

Optical clarity. Cast clear polyurethane polishes to real transparency. Printed "clear" parts are translucent at best and need extensive post-processing to approach it.

Cosmetic surface across a whole batch. Every cast part in a tool carries the same finish because it came from the same cavity. Fifty printed parts carry fifty slightly different surfaces, and bringing them all to a paint-ready state is fifty separate post-processing and finishing jobs — which is often where the printed option quietly loses on cost.

Where Printing Wins Outright

The design isn't frozen. A revision costs a new master and a new tool in casting. In printing it costs a re-slice.

The geometry has undercuts, internal channels or consolidated assemblies. Silicone tools tolerate far more undercut than steel, but a part with trapped internal geometry still can't be demoulded. Printing does not care.

The parts have to perform mechanically. Cast polyurethanes mimic engineering thermoplastics; they are not those thermoplastics, and their heat and creep behaviour reflects that. For load-bearing or elevated-temperature use, a printed nylon or a filled engineering grade is usually the more honest answer — see functional and end-use 3D printed parts.

The quantity is genuinely small. Under about 20 units, the tool never pays back. Low-volume 3D printing services covers how to get that batch priced properly.

How to Spec a Casting Job

Three things belong on the request that a printing quote does not need:

Parting line and gate location. Say which faces are cosmetic. If you don't, the shop splits the tool where it is easiest, and the seam lands where it lands.
Material by property, not by name. "Rigid, ABS-like, 40% glass-filled equivalent, white" gets you a better match than a trade name the shop may not stock.
Tolerance on the features that matter. Casting quotes hold to roughly ±0.15% with a floor near ±0.25mm. Call out mating features individually rather than applying a blanket tolerance the process cannot hold.

Ask any shop you're considering two questions: who prints and finishes the master, and how many parts they pull from a tool before they rebuild it. The first tells you whether the quality ceiling is in-house or subcontracted. The second tells you whether the price you were quoted covers the whole quantity or only the first mold.

Getting Both Quoted Together

The efficient move at this volume is to quote the printed batch and the cast batch from the same conversation, because the printed master is the first step of the casting path either way. A shop that runs both will tell you which side of the line your part falls on faster than a spreadsheet will.

Find a 3D printing and low-volume production provider →

Hero image by Tom Claes.