Production Quantity Economics
Production quantity economics is how the true cost of making a part changes as the order quantity rises, and which process, 3D printing, CNC machining or injection molding, tends to win at a given volume.
At what quantity does 3D printing stop being the cheapest option?
NIST's 2015 report on additive manufacturing costs cites three separate studies that each found a different quantity where 3D printing stopped being the cheapest choice for their own part, from about 42 units for a landing gear part up to about 10,500 units for another. The right quantity depends on the specific part being studied, not a number that carries across parts.
| Study (as named in NIST's report) | Quantity | Part studied | Provenance |
|---|---|---|---|
| Hopkinson and Dickens | 6,000 to 14,000 | their own part, not named in NIST's summary | NIST, Thomas 2015, page 9 |
| Ruffo and coauthors | 9,000 to 10,500 | their own part, not named in NIST's summary | NIST, Thomas 2015, page 9 |
| Atzeni and coauthors | up to about 42 | a landing gear part | NIST, Thomas 2015, page 9 |
What did Hopkinson and Dickens say it cost to print one part?
As reported by NIST, Hopkinson and Dickens' own cost-per-part figures were about 5.20 euro for SLA, about 4.40 euro for FDM, and about 2.20 euro for laser sintering, with machine cost as the dominant driver of cost for SLA and FDM. Those figures come from one paper's own model for one part, not a price for yours.
| Process | Cost per part | Provenance |
|---|---|---|
| SLA | about 5.20 euro | NIST, Thomas 2015, page 9 |
| FDM | about 4.40 euro | NIST, Thomas 2015, page 9 |
| Laser sintering | about 2.20 euro | NIST, Thomas 2015, page 9 |
Do Hopkinson and Dickens and Ruffo's own cost models even agree with each other?
No: Hopkinson and Dickens modeled additive cost as a flat line as quantity rises, while Ruffo and coauthors modeled a declining sawtooth shape, and NIST's 2015 report cites both without resolving the disagreement between them. This page cites NIST's open account of their findings rather than the two paywalled papers themselves.
What simpler rule of thumb do bureaus use for quantity versus process?
Protolabs Network's own rule of thumb states 3D printing is the cheaper process under about 10 units, CNC machining wins from about 250 to 500 units and up, and injection molding wins from about 500 identical units and up. That is general guidance from one bureau, not a calculation for a specific part.
How does our own pricing scale with quantity?
No minimum order. One part or a production run, same flat per-part pricing tiers.
Our own range runs one part to 10,000 parts on the same flat per-part tiers, with no separate tooling quote to negotiate before a production run starts.
What this page does not tell you
- The three NIST-reported studies each modeled one specific part on one specific process. None of them is a crossover a reader can apply to a different part.
- Hopkinson and Dickens' cost-per-part figures blend SLA, FDM and laser sintering into one paper's own model, not separately verified figures for each process.
- Both original papers behind NIST's summary are paywalled, so this page relies on NIST's own account of what they found.
When to print a test coupon instead
Get a real quote for your part at your real quantity before deciding which process wins on cost. The rules of thumb above are general guidance, not a calculation for your specific part. A person reviews your request and follows up by email.
Sources
- Costs, Benefits, and Adoption of Additive Manufacturing: A Supply Chain Perspective, NIST
Thomas, November 2015. Page 9 reports Hopkinson and Dickens (additive cost effective up to 6,000 to 14,000 units for their own part), Ruffo and coauthors (9,000 to 10,500) and Atzeni and coauthors (a landing gear part, up to about 42), and Hopkinson and Dickens' own cost per part (SLA about 5.20 euro, FDM about 4.40, laser sintering 2.20), stating machine cost is the major cost driver for SLA and FDM. States that Hopkinson and Dickens modeled a flat cost line while Ruffo and coauthors modeled a declining sawtooth, an unresolved disagreement between the two papers, both of which sit behind a paywall this report does not.
- 3D Printing vs CNC Machining, Protolabs Network
States a rule of thumb that 3D printing is the cheaper process under about 10 units, CNC machining wins from about 250 to 500 units and up, and injection molding wins from about 500 identical units and up. Also states a CNC minimum wall of about 0.75 mm.
Related
Use cases
Services
Tools
Published 5 September 2026.