Recycling 3D Prints


Despite being “plant based” there are almost no 3D printer filaments that are actually compostable or biodegradable in any meaningful way. PLA needs oxygen, a temperature of 140+°F (60°C), and a variety of organic compounds to break down. You’re not going to find these outside of an industrial composting facility and most of them won’t take PLA anyway because it’s frequently mixed with other things that make it non-compostable.

Fortunately PLA is completely re-usable. You “just” need to melt it down and re-extrude it.

UNfortunately, that’s a lot easier said than done.

This post describes what it would take for a home user convert your old prints into new filament to make new things with. It was written based on the tools available in January 2024.


The first step in re-using the plastic from old PLA is shredding it. To do this you need a bunch of specifically shaped disks made out of hardened metal that will chew your old prints into little pieces. You’ll also need one or two motors with a lot of torque to force the “teeth” of those disks through a block of plastic. The bigger the opening the more cutting disks you need, and the more powerful the motors need to be.

The SHREDII 5s is the most capable thing i’ve seen in a price range achievable by most people. It’s $1,400 BUT you also need 2 identical motors with enough torque and that’s roughly another $1,800. It can only handle something about the size of a soda can. It’s made by the folks on the ActionBox YouTube Channel who have videos of its use, and the prior generations that led to it.

I don’t know of any off-the-shelf shredder that I’d bother buying. They’re either industrial things that are ridiculously expensive, or hobbyist things that can only shred things that are stupidly small. You’ll never be able to shred anything but the tiniest of prints, test strips, and brims.

There’s not a lot the community can do to improve the cost of the shredder. High torque motors are expensive and no-one really wants to build one from scratch. There’s nothing terribly complicated about the shredding part. It’s just some hardened metal disks with vaguely saw-like teeth on two rods that rotate in opposite directions. You can definitely get the cost down significantly if you have the right tools and are willing to put in the effort.

For comparison, here’s a 4 minute overview of the $9,000 Filabot Reclaimer which can’t handle pieces larger than 3" (~7cm) and has feeding issues with many things that are smaller than that, requires 2 stages of shredding, and warms up enough to cause low temperature materials PLA to clump together, exceed it maximum size, and clog within a few seconds.


Cleaning out all remnants of things you’ve shredded before isn’t easy. If you don’t do it, you’ll get small pieces of past filament colors mixed in with the shreds of the current color.


A pre-made desktop extruder is going to cost $3,500 (USD) or more, but there are kits available for homebrew ones that will cost you about $600 to put together and provide very usable results.

At the time of writing the Arteme 3D Mk2 is the best one I’ve seen. [Note: when visiting that site you can change the language at that link by clicking a wee flag at the top.]. Unfortunately there’s no “good” way to measure the thickness of the new filament you’re extruding. It’s possible, but not easy.

This YouTube Video by CNC Kitchen does a good job of showing off the Artme 3D and compares it to a ~$3k machine. The $3K machine definitely seems less hackey and produces more consistent seeming results.

It’s decent enough for a small-scale home setup, but there’s a lot the community can do to improve it. Incorporating something like this cheap (~$5) open source inline filament sensor would be a huge improvement, but would require some programming and electrical engineering to give you a display of the current diameter, or better yet, incorporate it into a speed controller for the extrusion process. Professional extruders run the filament through a water bath to cool it off quickly, and something like that could be incorporated too although it’s probably overkill for this low-speed extruder.

The mixing bin is, in my opinion, completely inadequate for anyone who actually wants mix things to control the color of the resulting filament. In that case you’ll definitely want a separate mixing setup that does a way better job than a human with a big spoon.

Some of you may be thinking “but there are really cheap (~$200) extruders on Ali Express”. Please watch this video from Slant3D going over how amazingly useless the cheap Chinese extruders are. The one they got currently sells for $1344 and was barely functional.



Mixing filament colors is just like mixing paints. Mix enough colors together and eventually you end up with something brown or black. So, you’ll need to keep your filaments separated by color unless you don’t care what color the new roll is.

  • Intentional Mixing

    Unless you’re really careful about keeping your shredded filament colors separately you’re going to be mixing colors. If you’re mixing colors you’re probably going to end up with banding from uneven mixes. It turns out this is a hard problem. The folks over at Protopasta do this for a living, and they still have problems with it. As a result, they mix new colors and put them through one extruder, then shred the results, and re-extrude them in another extruder. Here’s a link to where they talk about this in a factory walk-through.

  • Creating Custom Colors

    Probably the coolest thing about having your own extruder is that you will no longer be limited to the colors you can buy from other manufacturers. You can buy the raw resin pellets in bulk. Then buy whatever pigments you want and mix them together. You can add in sparkles and anything else. The only “gotcha” here is that if you haven’t done so before, you’re going to have to spend some time learning basic color theory as it applies to mixing pigments. At a high level, everything they teach about mixing paints is applicable here, with the caveat that mixing pigments with plastic is probably going to be about as different from paint as mixing watercolors is from mixing oil paints.


Many manufacturers mix things in with their PLA. Because of this we don’t really know what happens when you re-melt and re-extrude them until you try. Additionally different filaments melt at different temperatures and extrude differently. Personally I’d stick to only re-extruding simple PLA and PLA with sparkles until you’ve done some research about the compounds involved.

Note that PLA+ is not the same as PLA and the “plus” could represent the addition of just about anything. There’s no industry standard around the term. So, you can’t mix the two and you can’t - or at least shouldn’t - mix PLA+ from different manufacturers.


If your newly extruded filament is too thick it’ll clog your nozzle. If it’s too thin… well I don’t really know what that does, but i suspect it’d heat too quickly and give you the same result as printing with a nozzle temperature that’s too high. If the thickness is inconsistent you’re just screwed.

I do not like how the Arteme 3D MK2 handles the problem of measuring your filament thickness. I hope the community comes up with a good improvement in the near future. Right now it seems to produce filament that’s “mostly close enough” in the thickness department. I wouldn’t use it for anything where precision extrusion mattered in the final product.


Winding your new spool well is really important. I’ve gotten multiple cheap rolls from cheap Chinese manufactures that are wound poorly, get bound up and snap because they won’t unroll anymore until you get your hands involved. The homebrew winding solutions seem to work reasonably well, but i figure there’s got to be more to it or else even the cheap companies that do this for a living wouldn’t be screwing it up.

I’m not thrilled with the winding solution of the Arteme 3D MK2, but you could either respool it with some of the more elaborate 3D printable solutions, or you could integrate something into a hacked version of the Arteme.


Not technically required, but if you want the best possible results my understanding is that you’ll want to shred, extrude, pelletize, and then re-extrude. Pelletizing is basically just making them into uniform little pellets that’ll help your re-extrusion process.


All in all you’re looking at about $3600 and some kit building to get started. If you already have a CNC that can work with steel you could save about $1400 on the shredder part, and maybe even make a bigger one, but you’re still going to need 1-2 high-torque motors. With a setup similar to the SHREDII you’re going to have to spend time cutting your old prints on a band saw or something to make them fit into your shredder.

Once you’ve got all the pieces you’ll have to be diligent about keeping your old print material separated by color and material type. That means a lot of bins, and that means a much longer time before you’ll have enough waste of a specific color and material to make a new spool. You’ll be much better off if you don’t care about color and just want to reuse material for test prints. In that scenario you only have to keep things separated by material.

You’ll need to clean your shredder after each use, which is probably best achieved by running some clear PLA resin through it. A little bit of that left in the system shouldn’t significantly effect the color of the next filament you shred.

If you don’t want random color banding in your prints you’re going to need to come up with a good mixing solution, and probably extrude your newly mixed color, shred it, and then re-extrude it.

All of this is very fiddly. It takes a lot of time, and it requires energy to shred, mix, and re-melt all your stuff. It requires energy and raw material to make the shredder and motors and extruder and electronics and everything else.

Right now home re-use of filament is doable, but it’s a lot of work. If you’re not generating a lot of PLA waste, I doubt that the small amount of material you keep from a landfill will compensate for the environmental impact of making all the things required to do it. At the same time, we need hobbyists to be willing to put in the time and energy to try this so that we have people iterating on the tools available to the wider community.

I don’t think you should bother with this if you just want to be a little more green. I don’t think that’s realistically achievable with the current tools. Please do try this if you’re a tinkerer who is interested in spending the time to iterate on what we have, and find a way to make it better for others, so that - eventually - something exists that actually does make a difference for everyone.