Autres blogs
Price = total cost times a markup. That's the whole thing. The trick is that most people only count the filament, and filament is usually the smallest number on the list. Here's the full cost stack I use, then a worked example you can copy.
Cost = Material + Electricity + Machine depreciation + Failure buffer + Labor. Then Price = Cost times your markup, or equivalently Cost divided by (1 minus your target margin). Everything below is just filling in those five numbers honestly.
If you'd rather not do the arithmetic by hand, the 3D Print Cost Calculator does the per-print version and the 3D Printing Business Calculator does the whole-shop version. This post explains what those numbers mean so the output isn't a black box.
The one everyone gets right. Take the grams of filament the part uses and multiply by your price per gram.
Material = (grams / 1000) times spool price. A $20/kg spool is $0.02 per gram. Get the grams from your slicer's estimate, or weigh the finished part on a kitchen scale, supports and purge included. Weighing is more honest because the slicer doesn't always count the skirt, the purge line, or a failed first layer you reprinted.
Small, but real, and almost every calculator skips it. Electricity = power (kW) times time (h) times your rate ($/kWh).
A desktop FDM printer draws somewhere around 50 to 150 W while running, with brief higher spikes during heat-up. Prusa's own published figures land near 80 W average for a PLA print and around 120 W for ABS, because the heated bed does most of the work and ABS runs a hotter bed. So a 5-hour PLA print at $0.15/kWh is roughly 0.08 kW times 5 h times $0.15, about 6 cents. Not nothing over a year, but not the line that makes or breaks a quote.
Your printer wears out. Spread its cost over the hours it'll run.
Depreciation = printer cost / expected lifetime hours, times print time. Desktop machines commonly get quoted at 3,000 to 10,000 hours of useful life; a lot of people assume around 5,000 as a middle figure. An $800 printer over 5,000 hours is $0.16 per print hour. Then add a bit for the consumables that aren't the printer itself: nozzles, belts, PTFE tube, the occasional bed sheet.
Prints fail. If you only charge for the good ones, the bad ones come out of your margin.
Add something like 5 to 15% on top of your material and time to cover failed prints, clogs, and tangled spools. Five percent is a well-tuned machine you babysit; fifteen is a busy shop running long unattended jobs. The right number is the one you actually measure, so track your real failure rate and use that.
The one that quietly dominates small jobs. Labor = (setup + post-processing hours) times your rate.
Split it in two. Unattended machine time, where the printer runs and you're doing something else, is cheap to price, maybe $5 to $10 an hour if you bill it at all. Hands-on time, slicing, plate prep, popping the part off, removing supports, sanding, packing, is your actual labor and should be billed higher. Fifteen minutes of hands-on work at $20 an hour is $5, and on a small print that's often more than material, electricity, and depreciation combined.
Once you have the cost, you mark it up. Common markups run from about 1.2x (a 20% bump) up to 2x (doubling). A frequently cited target is roughly a 40% margin, which means dividing your cost by 0.6 rather than multiplying.
Markup and margin aren't the same math. A 2x markup is a 50% margin. Dividing by 0.6 is a 1.67x markup. Pick one and be consistent.
If you sell online, add the platform's cut. Etsy, PayPal, Stripe and friends take roughly 5 to 15% of the sale plus a fixed fee around $0.25 to $0.50 per order. Bake that in before you set the sticker price, not after, or the platform eats your margin.
Here's the whole thing on one 50 g part that prints in 5 hours. Assumptions: filament $25/kg, printer pulls 100 W, electricity $0.15/kWh, $800 printer amortized over 5,000 hours, 10% failure buffer, 15 minutes of hands-on labor at $20/hour, and a 2x markup.
| Line | Math | Cost |
|---|---|---|
| Material | 50 g times $0.025/g | $1.25 |
| Electricity | 0.1 kW times 5 h times $0.15 | $0.075 |
| Depreciation | $800 / 5,000 h times 5 h | $0.80 |
| Failure buffer | +10% of the above | $0.21 |
| Labor | 0.25 h times $20/h | $5.00 |
| Total cost | ≈ $7.34 | |
| Price (2x) | $7.34 times 2 | ≈ $14.68 |
Look at where the money is. Labor is $5.00 of a $7.34 cost. Material is $1.25. Electricity is under a dime. If you'd priced this part on filament weight alone, you'd have quoted a couple of dollars and lost money on your own time.
That's the whole lesson: for small, quick prints, labor dominates, and pricing on grams badly undercharges. Electricity and depreciation are each small, but they're the two lines most calculators drop, and dropping them means every quote is quietly a little short.
Treat the specific numbers here as conventions and typical ranges, not standards. The 100 W draw, the 5,000-hour lifetime, the 10% failure rate, and the markup multiples all vary a lot by machine, region, material, and business model. There's no official body that sets them. The dollar totals in the example are illustrative, meant to show the shape of the stack, not a quote you should hand a customer. Plug in your own measured numbers and the shape holds even when the digits change.
For the full side-hustle picture, spreadsheets and all, see how to make money with a 3D printer.
If designing the part is where your hours go, that's the line I'd attack first. GrandpaCAD turns a chat prompt into parametric CAD, so a one-off bracket or replacement clip is minutes of describing instead of an afternoon in CAD, which keeps your labor line small enough to actually make the margin work.