What to charge for a 3D print

Most shops arrive at a price the same way: guess, feel uneasy, take the job anyway. It works until the month you realise the machines ran constantly and the bank account did not move.

Here is the method. It fits on the back of an envelope, and the only hard part is being honest in three places where almost everybody rounds in the customer's favour.

Add up what it actually costs you

Three things go into a print, and almost every shop counts one of them.

The plastic

Not the weight of the part. The weight off the spool.

Purge towers, brims, skirts, the priming line, and the print that failed at hour nine all came off the same spool as the part you sold. If you price by the weight your slicer reports for the model, you are giving away everything else.

A five per cent allowance is a careful shop with a single-material machine and a low failure rate. A multi-colour printer that purges on every filament change is nowhere near five per cent - on a four-colour print the purge can outweigh the part. And a shop running unattended overnight has a failure rate it has probably never measured.

Measure it rather than assuming it. Weigh a fresh spool, run it dry, and add up the part weights your slicer reported over its life. The gap is your real waste rate, and it is the number most likely to be quietly wrong on every job you have ever quoted.

Then price the spool properly. A spool costs what you paid for it plus what it cost to get it to you: shipping, customs, and the drying you do before it goes on the machine. The filament cost calculator turns a spool price into a cost per gram and a cost for the job, which is the form you actually need it in.

The machine

A printer wears out, draws power, and needs parts.

Spread its purchase price over the hours you expect it to print before you replace it. Add the electricity it draws while it runs - measure it with a plug meter for one print rather than reading the nameplate, because the nameplate is the peak and the heated bed is only on hard for the first few minutes. Add the servicing it needs in a year, divided by the hours it runs in a year: nozzles, belts, a hotend, the build plate you eventually scratch through.

That is your machine cost per hour. On a desktop machine it usually lands between ten and thirty cents. Small per hour, and not small at all across a twelve-hour print - and a twelve-hour print is an ordinary Tuesday.

The hourly rate calculator works it out from what the printer cost, what you expect to get out of it, and what you pay for power.

Two things people get wrong here. The first is spreading the machine's cost over its calendar life rather than its printing life - a printer that sits idle four days a week costs the same to own and earns nothing while it sits. The second is forgetting that a farm has costs no single machine has: the shelving, the enclosure, the dry box, the spare parts drawer, the rent on the corner of the room. Put those somewhere, even if it is a flat number added to every job.

Your time

Levelling, removing supports, sanding, gluing, painting, packing, answering the email, writing the quote, driving to the post office.

This is the one everybody forgets, and it is the one that decides whether the business is a business or an expensive hobby. It is also the one that does not scale: a second printer doubles your output and does nothing at all for the twenty minutes of finishing each part needs.

Put a real hourly rate on it. If you would not do somebody else's job for that rate, it is too low. And time the finishing on a few real jobs rather than estimating it, because the estimate is always the memory of the one that went well.

Then multiply

Cost is not price.

The multiplier on top covers everything that never appears on any single job: the print that failed and was reprinted for free, the quote you wrote and never won, the customer who vanished after you had bought the filament, the hour you spent on the phone explaining why the model as drawn cannot be printed. It also covers the thing a price is actually for, which is profit.

Most shops land between 1.5x and 3x on cost.

Below 1.5x you are running a charity with a power bill. Above 3x you had better be quick, or unusual, or both - and if you are, charge it, because the shops competing at 1.2x are going out of business and will not be there next year.

Two adjustments worth making by hand:

Rush work costs more. Not because it costs you more to print, but because it costs you something to reorder the queue, and because a customer who needs it Thursday has already decided they are paying for Thursday.

Design time is not print time. If you are drawing the model, that is a separate line at a separate rate, quoted before you start. Bundling it into a per-gram price is how a two-hour CAD job gets paid at the rate of forty grams of PLA.

Sanity-check the number against the market

Cost-plus gives you a floor, not an answer. Once you have the floor, look at what the job is worth to the person buying it.

A jig that saves a factory an hour a day is not priced on its twelve grams of PETG. A cosplay helmet competes with every other cosplay helmet on the internet. The same printer, the same plastic, and two completely different right answers.

Where your cost-plus price is far below the market, take the market price. Where it is far above, either say no or find out why - usually the print is being run in a way that suits the machine rather than the job.

Do it once, properly, and then stop doing it

The print cost calculator does the sums for one job: filament, machine time, labour, waste, multiplier, price. The same calculator sits on the front page if you want to try it before reading anything else, and the link it produces remembers your numbers, so you can send a quote to somebody without retyping any of it.

That is enough for a quote. It is not enough for a business, because a calculator only knows what you type into it, and the thing you get wrong is not the arithmetic - it is that the spool price you typed in was last year's, or that you forgot the job failed once, or that the printer you costed at thirty cents an hour has been running twelve hours a day since March and has long since paid for itself.

That is the part worth automating: your real spool prices, drawn down as jobs consume them; your real printers with their real hours; and a number at the end of the month that tells you which jobs made money rather than which ones felt busy.

The one thing worth doing tonight, before any of that: measure your actual waste rate. Five per cent is what everybody assumes, almost nobody checks, and it is wrong on every single job until somebody does.