Subscribe to Make Magazine Today!

Make: 42 Annual 3D Printing Guide


From scrappy laser-cut wooden kits only a few years ago to today’s sleek-looking machines, desktop 3D printing continues to evolve into evermore consumer ready machines. Since there are so many uses and user types for 3D printers, from schools and teachers to product prototypes and engineers, our guiding question was never “which printer is best?” but was instead “which printer is right for you?” We tested 26 printers and judged performance through a series of different prints and tests. Check out our findings and find the best printer for your needs.

Then, get tips on printing movable parts in a single print, make sure to stick your first print layer from the start, and get the skinny on Local Motor’s fully 3D-printed car.

Plus, add glowing cyberpunk spikes to your favorite futuristic outfit, 3D print a pair of custom eyeglasses, build a adjustable-height modeling table on the cheap, and learn to pickle delicious beets and grapes. All this and more in Make: Volume 42!






Universal Translator

Universal Translator

Turn a $35 mini-computer into a feature-rich language translator that not only supports voice recognition and native speaker playback, but is also capable of translating between thousands of language pairs.


Categories: Electronics Raspberry Pi


  • jstults

    This is great; thanks for sharing!

    I was excited when I read the part about you using a continuous measure (height at failure) instead of a binary response (fail / no fail). But then I went to the evaluation protocol page, and you have folks “bin” their responses into categories, e.g. “Assign the print a “1” if average deviations in X or Y are greater than 0.4mm.” Why do this when they’ve already made a measurement, and can just report the more informative continuous response?

    I think more quantitative and standardized tests for these consumer grade devices is a great thing, so minor methodology nitpicks aside: Bravo!

    • Luis E. Rodriguez

      We would really rather test these long term and will, but that would mean we would never go to print! We feel like we have a solid testing platform thanks to Andreas’ hard work!

    • andreasbastian

      When I was developed the test geometries, I calibrated the dimensions of the positive fine feature test to reliably stop printing on one of the two machines I tested on. It turns out the machine that kept failing was at the low end of the distribution of machines that can pull off that geometry, so the distribution was weighed more heavily than intended towards completion. With the tight timeframe of the shoot out, we decided to judge more on deposition control than completion rather than re-calibrate the probe and re-run the prints. The negative space test geometry is my new standard for minimal ambiguity in reading the test result and I am designing testing suite 2.0 to be far more “auto-diagnostic”.

  • Michael Overstreet

    A lot of your tests seem to be design to see how well the printer prints without support material? Why is that?

    I think you should have more tests to see how well a printer prints with support materiel not without it. In the future how well a printer prints with support material will become more and more important.

    • Anna Kaziunas France

      We felt that the current crop of machines should be able to handle bridging and minor overhangs without support. Most could. As Andreas mentions, we tested to failure so the tests are intentionally overkill.

      We ran one long overnight sculptural print with support material on all the printers (just to get a rough idea) – but did not quantify those prints. I can tell you that most software currently does a terrible job of creating easily removable supports and surface pockmarks after removal were rampant. At this juncture, I would recommend creating custom supports in MeshMixer over any automatically generated support.

      Overstreet, you print lots of robot parts. I’ve noticed that the DARwin Mini parts often require support. Are you creating custom supports or auto-generating? Are you still using an Afinia H Series? Which one?

      • Michael Overstreet

        I have a UP Plus Version 1.2 and I am thankful that the software generates the support material for me and that the printer does a good job of printing it out!

        It is too bad that other printers can not do the same and based on the reviews it looks like few printers can print in ABS anymore too. It is going to make it difficult for people to print out usable robot brackets for the ROBOTIS-MINI, ROBOTIS-OP and my BIOLOID-3D robot.

        It actually feels like 3d printers have less capability then last year and the year before that?

        • Luis E. Rodriguez

          Looks like only 7 of the 21 FDM printers specified PLA or Nylon only. In most cases that is due to a lack of heated bed which is an easy upgrade for open-srouce printers, and available from the manufacturer. We ran these as stock.

          • Hal9000

            That’s quite true, I know my Ultimaker Original can print ABS, PLA, Nylon, etc with the heated bed upgrade ( it takes a bit to put it in (I had some help) but it’s so worth it. I learn a bit more about printing on a regular basis and I’ve only seen the technology improve.

            I will grant you that there are more lower end models entering the market and that may lend the appearance of an incline however that is truly not in proportion to the incredible advances that continue to be made

        • andreasbastian

          The Up/Afinia machines are some of the only machines on the market that can reliably implement single-material support. No other open source slicers seem to have the solid interface support that the Up/Afinia software has. Most machines tend to compensate for that shortcoming with active cooling of the extrudate.

          • Michael Overstreet

            In many ways it prints like the higher end Stratasys printers that my friends have. We can send each other .STL files back and forth with little modification. Most of are talk is how to make the parts better not how to get our printers to print.

          • Luis E. Rodriguez

            You are correct. Those are all closed-sourced printers you’re printing your open-source robots on. Those printers that Tomio and Shibata use are very expensive, as is the proprietary chipped filament cartridges and required support cleaning station. The Fortus 250mc is is the range of $44k-$54K. They use a heated chamber (patented) and disposable build platforms. The printers we reviewed are affordable desktop alternatives to those with every possible setting exposed for the user to tweak, on the open printers at least. Thankfully, STL and the smaller OBJ format are open for us to use!

      • Chuck Hellebuyck

        The $500 Davinci 1.0 3D printer with included software will automatically include support material as a clickable option in three different density settings. Breaks-away easily. And includes a heated bed for ABS. PLA coming soon (though some have hacked it and have it printing with PLA now).

        So if I run these same tests on my Davinci but include support material does that still meet the test guidelines?

        I ran the original Make 3D Torture Test on my Davinci with supports and it printed fine. See it here:

    • Luis E. Rodriguez

      Most of these printers are open-source so you can choose to have supports on or off as well as the density and style. Also if you look at the review issue, it very clearly states what materials are supported for each reviewed printer. ABS is easy to work with with the right care. We were printing with it in 2009 on cupcakes. Open-source printers mainly use Repetier host/Slic3r or Cura. but all put too much out and waste material, including the UP/Afinia’s. Mesh mixer does an exponentially better job. Also the reason your parts require so much support is because they were originally intended for plastic mold injection, I’ve seen more support than model on some of those parts. Try reorienting them or modifying them for 3d printing to save plastic. Also there are so many better materials than ABS to use. Try Nylon or PET+ if you are able to. Refer to the e-Nable project, they do a lot of materials testing. Good luck!

      • Michael Overstreet

        I already know this, I was just curious why it was not in the magazine? As it would help people new to 3d printing out.

        • Anna Kaziunas France

          There are only so many pages in a single magazine. We’d love to publish more, but we crammed in as much as we could. ;-)

          • Michael Overstreet

            Sounds like my work here is done then, Cool! Now back to more important work!

        • Luis E. Rodriguez

          That is actually a good idea for a blog post followup! Thanks Overstreet!

  • 3dhacker
  • Roberto TM

    Why the Cube 3 and Cube Pro were left out of this review?