There are print request sites where people will put up a print model and filament requirements and you can agree to provide them by X date for Y dollars.

Sometimes people need one offs and dont want to buy a printer, so they pay $50 for $5 worth of plastic/electricity. Sometimes other folk need 100 of something and pay $5/each for something like a green rectangle. With solar panels or cheap electricity, as long as you are making a profit after buying plastic and have the process tuned in, you basically have machines making $1-3/hr just running 24hr/day.

Manufacturers will pay quite a bit for tooling if FDM plastic material properties fit their requirements. Especially if you are able to do CAD from a specification. Like if you're willing and able to 1-day print and ship some very specific green rectangles, you may be pleasantly surprised at the margins on those.

I run a small online store and make between $400 and $1,000 a month in profit.

I sell specialty phone cases that connects Razer jungle cat controllers like a switch. Also Galaxy fold cases that hold the new s pen. Other stuff too

Frogcase.store

As inefficient as it is, for smaller quantities it does still to be quite a bit cheaper and faster than older methods. There's obviously a breaking point, but if you don't know you'll hit that point, I imagine that it's pretty easy to just keep going with FDM prints beyond the line where you could have saved money if you'd known ahead of time.

That said, I don't do that. I sell articulated toys at the local market. So far I've always made way more than enough to cover my costs, and I figure I'm about 1/3 of the way to paying for all my printers, including the 1 I actively use, the 1 I rarely use, and the 2 that I never use.

I figure that upgrading to a Bambu would give me multi-color capability, and remove a fair bit of stress from my printing. So I'm pretty tempted, but I'm trying to wait to see what they release next.

It is a HUGE business in mechanical design spaces.

Low-quantity prototyping is prime realestate since injection molds cost 1500€ and rapid prototyping is not a thing without a bunch of money behind you.

My company does mostly medical devices and some consumer design doesn't have the equipment for high-quality 3D prints. We use a company called materialize that does extremely high quality sand-blasted or waterproof impregnated 3d prints as an example for low quantities. They made one prototype for us that literally sounded and felt like ceramic while being waterproof plastic.

It's good business. A run of 10 prints or so is also like 800€ or something, so they make decent money from it if they can automate the process as much as possible.

Getting a part injection molded is a pretty involved process. You have to find a manufacturer who will work with you, have a prototype mold made, do a test run, correct the mold/material/process, have a new mold made (iterate as necessary), then do a full production run. Making the molds is usually pretty expensive by itself. It's all fine if you're making thousands of units, but if you only need like 100 of something it's usually not worth the time or money.

It's a lot faster and easier to work through prototypes by 3D printing them (it is what they were originally made for), and if you need a relatively small batch it might be more efficient.