How to price FDM prints

Most FDM price calculators are one line: volume × a number. It is not a stupid line - for a family of similar parts it is even roughly right - but it fails in exactly the cases that cost you money, and it fails silently.

Here is what an FDM price is actually made of, in descending order of how much it matters.

1. Machine time, and why volume does not predict it

Two parts of identical volume can differ by a factor of three in print time. A solid 40 mm cube and a thin 40 mm shell of the same mass are the extreme case: the cube is mostly fast infill, the shell is almost entirely slow outer perimeter. The number of layers is set by height, and the work within each layer is set by the cross-section at that height - the perimeter length, the skin area, the infill area.

That is why Quotometry measures a per-layer cross-section profile rather than a bounding box. It is also why it asks for your printer’s actual speeds: outer wall, inner wall, infill, travel and the acceleration you run. A part with a lot of short moves spends most of its time accelerating and decelerating, and a model that ignores acceleration will underquote it badly.

2. Filament, including what you throw away

Material cost is filament mass times price per kilogram, and the mass is not the part’s volume times the material density. It is:

• the walls, at your line width and wall count;
• the top and bottom skins, which are solid;
• the infill, at the density the buyer chose and the pattern factor of the pattern you use - a gyroid does not use the same material as a grid at nominally the same percentage;
• the supports, at the density and style you configure;
• and the waste - purge, priming, the tail end of a spool you cannot use.

Shops usually remember the first three and forget the last two, which is the difference between a healthy margin and none.

3. Labour, which is most of the cost on a small part

Take the plate off, get the part off the plate, pull the supports, inspect it, bag it. On a four-hour print this is noise. On an eleven-minute print it is the entire job, and it is the reason a minimum order value is not greed - it is arithmetic.

Price labour by activity at the rate of whoever does it, not as a flat percentage. The person pulling supports and the person doing final inspection are usually not the same person and not on the same rate.

4. The machine itself

An hourly rate that covers depreciation, the space it stands in, the maintenance it needs and the failed prints it will produce. If your machine rate does not include a failure allowance, your failed prints are coming straight out of your margin, and you will not notice which jobs are doing it.

5. Everything else

Energy, if you care. Nozzles, build sheets, adhesive. Overhead and margin, applied once and visibly.

The number nobody can tell you

Every model above has coefficients, and the honest ones admit that nobody knows yours. How much of your print time is travel rather than extrusion depends on your firmware, your part mix and your slicer settings. How much material a spool really yields depends on your purge routine.

So Quotometry does not hard-code them. They are settings with typical ranges, and the calibration tool fits them from jobs you have actually run: give it three or more parts with their real times and weights, and it searches the coefficients that are allowed to move and reports the error before and after. Typically a shop starts around 20 % out and lands in single digits.

What this means for your quoting

If you quote by volume × a number, you are systematically overcharging thin-walled parts and undercharging dense ones, overcharging long prints and undercharging short ones. Your customers with simple chunky parts go elsewhere; your customers with fiddly ones stay, because you are cheap for them. Over a year that reshapes your order book in a direction you did not choose.

How Quotometry prices FDM, in detail →