How much does a printed metal part really cost: the five costs you don’t see

Almost everyone asks the price per kilogram, and that's the wrong question. In metal 3D printing, the cost of a part depends primarily on how many hours the machine takes and how much space it frees up for other parts, not on its weight. Two parts of the same weight can cost twice as much, and rotating a part ninety degrees can halve its price.

What does the price consist of?

Item	Indicative Weight	What it Depends On
Machine Depreciation	30–45%	Build Hours. This is the dominant item
Material	10–25%	The volume of the part plus its supports
Thermal Post-processing	10–15%	One furnace cycle is divided among the parts that fit
Support Separation and Removal	10–20%	Manual labor. Depends on access, not size
Machining and Finishing	0–30%	Only surfaces with tolerances
Quality Control	5–15%	Increases significantly if documentary evidence is required
Preparation and Engineering	5–10%	Distributed among the series. In a single part, it's significant

The first line is the one that matters, and that's why it's important to understand exactly what makes the hours grow.

Why height matters more than weight

An LPBF machine works layer by layer. Each layer has two times: the time it takes the coating machine to spread the powder, which is the same for the entire build plate whether there is one part or forty, and the time it takes the laser to melt, which does depend on how much surface area needs to be covered.

From this arise two consequences that defy the intuition of anyone coming from machining:

Height is expensive. A 200 mm high part needs twice as many layers as a 100 mm high part, and therefore twice as many coats of primer, even though it has half the material. Laying a part flat is usually the most cost-saving decision of the entire project.

Company is cheap. If the build plate is half empty, the coating hours are paid for by your part alone. If it's full, they're shared. That's why a batch of twenty costs much less per unit than a single part, without the process having changed at all.

Put the two together and you have the practical rule: what's being priced isn't the item itself, it's the space and time it occupies on the shelf.

The material is deceiving twice

Powder is expensive to buy, yes, but powder that doesn't melt isn't wasted: it's vacuumed, sifted, and returned to the system. What the part actually consumes is its volume plus that of its supports.

Where the material cost really hurts is elsewhere: changing the alloy. Draining the circuit, thoroughly cleaning it, and refilling it with a different powder takes hours of work and carries the risk of cross-contamination. That's why a part made with an uncommon alloy can cost significantly more than the price of its powder suggests. It's not the weight of the powder itself: it's the cost of the change.

The five costs that do not appear in the budget

These are the things that turn a profitable project into one that wasn't.

1. Iteration. Almost no part comes out right the first time. Between the first build and the good one, there are usually one or two rounds of adjustments to orientation, supports, and parameters. If the budget only allows for one build, the budget is wrong.

2. File Preparation. Orienting, generating supports, slicing, and checking takes one to several hours for a skilled person. In a series, this time is less frequent; in a single part, it can be the largest part of the entire process.

3. Removing supports. This is manual work, and its cost is not related to the size of the part; it depends on whether or not the tool can reach it. A support in an inside corner can cost more than the entire build.

4. The Evidence. If the client requests a material certificate, process report, tests on specimens manufactured during the same build, or documented dimensional control, that takes time and involves testing. It can add a very large fraction to the total, and that's perfectly legitimate: what's not legitimate is discovering it at the end.

5. The scrap rate. Defective parts are still paid for. In a mature and well-parameterized process, the scrap rate is low, but not zero, and for the first part of a new geometry, it is high by definition.

What really drives the price of your part

Ordered by the weight of each lever:

Orientation. Changes height, support surface, and quality simultaneously. It's free and has the greatest impact.
How many fit together. A batch that fills the build plate spreads out the fixed hours.
The supports required by the design. A chamfer instead of a horizontal edge eliminates an hour of manual labor.
How many surfaces need tolerance. Each one adds a machining operation.
The level of evidence. There's a big jump from none to full traceability.
The alloy. It matters, but less than people think.

Note that the first four are design decisions, not manufacturing decisions. By the time the file reaches the shop floor, most of the cost has already been determined.

An order of magnitude, with all due caution

To give a reference and not a magic number: a small functional part in stainless steel, the size of a fist, with no documentation requirements and made in a short run sharing a build plate, is in the order of hundreds of euros per unit. The same part on its own, with the build plate to itself and a process report, can multiply that figure several times over.

That price range is so wide because the variables above have more influence than the part itself. Anyone giving a price per kilo without having seen the geometry is just guessing.

How we calculate it

That's what SamyQuote is for: it estimates build time, build plate utilization, and cost per part based on geometry, material, and quantity, using the same process parameters we use in manufacturing. It's not a price list: it's the same calculation we do for ourselves.

And if instead of the number you want the part, AdditiveDelivery will manufacture it without you having to buy anything. The price comes from that same calculation plus an agreed-upon margin and shipping, not from a bid.

—

The cost of setting up the entire installation, in the real CAPEX of entering additive manufacturing. How to design to require fewer supports, and therefore fewer removal hours, in supports in 3D metal printing. And when this technology beats machining and when it doesn't, in printing or machining.

And the comparison with the other two ways for a complex part, in printing, casting or sintering.