3D Printing vs Injection Molding: Which Costs Less at 100, 500 or 1,000 Parts?

The Short Answer

Injection molding needs a mold before the first part exists, then makes each part cheaply. 3D printing needs no tool, and its price per part falls only slowly as the quantity rises. At 100 parts each molded part carries one hundredth of the mold's cost, at 500 one five-hundredth, and at 1,000 one thousandth, so the answer depends on how large the mold quote is next to the printed price. There is no universal break-even quantity: peer-reviewed studies put the crossover anywhere from a few hundred parts to tens of thousands, and one study found a tenfold difference between two parts. Price both routes at your own quantities, then let design stability, time to a first part and the part's geometry weigh as much as the arithmetic.

Two Different Cost Structures

The two processes spend their money at different times. Hopkinson and Dickens, who costed molding from real quotes for tooling plus a price per molding, found that the cost of the tool far outweighs the unit cost of each additional part. In their model of printing, reviewed by NIST, the machine cost per part stays constant whatever the volume, while the molded cost per part falls as the mold is spread over more parts. On our quote page, quantity pricing tiers are applied automatically, so the per-part price you see already reflects the quantity you enter.

Spreading a Mold Over 100, 500 and 1,000 Parts

Call the mold quote T and the molder's piece price p. Each molded part then costs T divided by the quantity, plus p. The mold's share is plain arithmetic, and it is the line that moves most between 100 and 1,000 parts.

Where the Two Lines Cross

Going from 100 to 1,000 parts cuts the mold's share of each part tenfold, while the printed price per part (P) changes only by its quantity tier. The two routes cost the same at roughly T ÷ (P − p) parts. Because P steps down at each tier, work out the crossing with the tiered price at each quantity rather than with one number.

• Hypothetical example, with invented ratios that show the arithmetic only: suppose the mold quote equals the price of 300 printed parts (T = 300P) and the molded piece price is one tenth of the printed price (p = 0.1P).
• At 100 parts, each molded part costs 3P + 0.1P = 3.1P, about three times the printed part.
• At 500 parts, it costs 0.6P + 0.1P = 0.7P.
• At 1,000 parts, it costs 0.3P + 0.1P = 0.4P.
• The crossing falls near 300P ÷ 0.9P, about 333 parts. Make the mold quote equal to 3,000 printed parts instead and the crossing moves to about 3,333.
• These ratios describe no real part, mold or quote. The size of the mold quote relative to the printed price is what sets the answer, which is why only your own quotes can settle it.

What Published Studies Found

Cost studies agree on the shape of the curves and disagree on where they cross. That disagreement is the most useful finding they offer.

Why the Studies Disagree

• Hopkinson and Dickens (Loughborough University) assumed a machine building the same part for a full year with every build filled, and that the printed parts' material properties, surface finish and accuracy were not an issue. They noted that parts designed for molding may need redesign to work as printed parts. The cover's crossover was about one tenth of the lever's, which they read as printing suiting smaller parts.
• NIST's review of the cost literature (Thomas and Gilbert, SP 1176) concludes that printing is cost effective for small batches, and that in many instances its cost exceeds that of traditional methods. It reports the mold as a large share of molding cost in the Atzeni study, which compared laser sintering with molding.
• Bezzina and Refalo (University of Malta) modeled a farm of desktop FFF printers against multi-cavity steel molds. Below about 20,000 compacts a year printing was cheaper because of the molds' cost; above it, molding was. They note that break-even points reported by researchers vary between 200 and 20,000 parts, depending on the size and complexity of the printed part.
• None of these is a quote. Each assumes its own equipment, utilization and costs, on machines of its time, and the Atzeni figures are for laser sintering, a process we do not run. Use them for the shape of the curve and your own quotes for the number.

Worksheet: Your Own Numbers

Fill this in from a printed quote and a real mold quote. On the quote page, enter 100, then 500, then 1,000 with the material you would actually use; the tier applies automatically, and the page can be downloaded as a PDF for your records. Ask the molder for the tool price, the piece price at each quantity, and what both include.

Lines People Leave Out

• Sampling and first-article costs on the mold side, and the price of modifying the tool if the design changes after it is cut.
• Cavities and tool life: how many parts each cycle makes and how many the molder expects the mold to make, if you will reorder.
• Secondary work on either side, such as inserts, painting or assembly. Post-processing on printed parts is quoted on request.
• The quantity you will buy over the product's life, not only the first order. A mold bought for one order of 500 is spread over 500 parts.

Factors That Usually Decide

When the arithmetic is close, these decide. Hopkinson and Dickens list printing's advantages plainly: no tooling freezes the design, so geometry changes avoid the time and cost of new tooling; the lead times imposed by tooling disappear; and printing allows variable wall thickness and zero draft, which injection molding will not tolerate.

Molding Has Design Rules Printing Does Not

MIT's course notes on injection molding (2.008, Design and Manufacturing II) set out rules a printed part never had to follow:

• Uniform walls. Keep the part as thin and uniform as possible; that shortens the cycle, improves dimensional stability and avoids surface defects. Thick sections cause sink marks, and uneven thickness is one cause of warpage.
• Draft. Faces need draft to release from the mold, typically 1–2° depending on the material, the dimension and the surface texture.
• Parting line. The line where the mold halves meet goes on a flat or simply curved surface, so it does not flash.
• Undercuts. A feature that locks the part in the mold needs a moving slide core in the tool.
• Shrinkage. Molded parts shrink as they cool, by amounts that differ several-fold between resins and differ along and across the direction of flow.

Designing for Both

A part designed around printing's freedoms may need redesign before it can be molded. If you expect to mold later, design to molding rules from the start and print the prototypes to those rules, so the printed parts test the design you will tool. Our design tips cover the printing side, and 3D modeling services can rework a model, quoted per project with a fixed price and schedule before work starts.

Material, Finish and Tolerance

• Material. A printed part and a molded part in the same polymer family are not the same part. FDM parts are weakest across their layers (print orientation and strength), and Hopkinson and Dickens note that printed materials often fail to match their molded counterparts. If printed parts stand in for molded ones in a test, read specifying printed parts for validation testing.
• Surface. A molded part copies its tool. In Bezzina and Refalo's comparison the molded compacts were smooth and glossy, while the FFF compacts showed layer stepping and stringing, and removing layer lines would have added cost. Compare our SLA and FDM surfaces before you decide, and see the post-processing guide for finishing options.
• Function. In the same study the printed compacts missed their dimensional limits because of thin walls, and their thin hinges failed tests the molded compacts passed. Test thin, flexing features first.
• Tolerance. The typical tolerances we publish are FDM ±0.2 mm and SLA ±0.1 mm, general guidance that depends on geometry, not a guarantee (tolerances and accuracy). A molded part's tolerance depends on the tool and on shrinkage, so ask the molder what it will hold on your drawing.

Printing and Molding in Sequence

The two processes are often used one after the other rather than as rivals: print prototypes until the design stops changing, print first units or variants while the mold is designed, built and sampled, and keep printing low-volume variants after molding starts. Urethane casting from an SLA master is another route for small runs; see printed molds vs mold masters. We print FDM and SLA parts; we do not run injection molding. For printed runs, see low-volume production. For the batch side of the decision (first articles, revision control and inspection), read small-batch production, and for the launch side, launching a product with 3D printing.

Before You Decide

• Fill in the worksheet from your own two quotes. Do not borrow a number from a study or from another part.
• Ask how likely the design is to change, and what one change would cost on each route.
• Check the part against molding rules, and the printed material, surface and tolerance against the requirement.
• If you print the run, ask for the free printability review. An engineer checks walls, tolerances, orientation and material choice at no cost; it is a printability review, not engineering validation.