Does infill percentage change print cost?

Yes — and not by a small amount. A 60 mm cube at 0.2 mm layers costs $2.83 at 20% infill and $10.50 at 100%. Nearly four times, for the same object.

Infill is the one setting that moves two cost terms in the same direction at once, which is why it hits harder than any other slider in the profile.

All four rows: PLA at $22 a kilogram, 100 W, 30 cents an hour of machine time, US electricity rate.

Why the jump from 0% is so steep

Note that 0% infill is not free — $0.92, not nothing. A hollow part still prints its perimeters and its solid top and bottom skins, and on a small part those are most of the plastic. The floor is the shell.

That is also why the relationship is not proportional. The first step, from 0% to 20%, roughly triples the bill. The last step, from 80% to 100%, adds a far smaller share of it: by then the interior is nearly solid and the walls have not moved at all.

The setting is rarely worth what it costs

Above roughly 40%, extra infill buys very little stiffness on most geometries — wall count does that job far more efficiently. Four walls at 20% infill will out-perform two walls at 50% infill on nearly every load case, and it costs less.

So the useful question is not "what infill saves money" but "what is actually carrying the load". If the answer is the shell, the infill number is a rounding term you are paying four figures a year for at print-farm scale.

Where infill genuinely matters

• Compression through the part. Foot pads, spacers and anything that gets squeezed need interior material.
• Screw and insert bosses. Local solid regions, not a global percentage.
• Anything printed in a soft filament. With TPU, the infill number is effectively a hardness dial for the finished object.

Model your own part in the print time estimator, then price it in the print cost calculator. The mechanics of which pattern to use are in the infill patterns guide.