AnkerMake M5 3D printer review: speed is not enough

By Sean Hollister, a senior editor and founding member of The Verge who covers gadgets, games, and toys. He spent 15 years editing the likes of CNET, Gizmodo, and Engadget.

If you buy something from a Verge link, Vox Media may earn a commission. See our ethics statement.

Last April, I introduced you to a seemingly game-changing new 3D printer: the AnkerMake M5. "Printing So Smart, It's Easy", the company's tagline read.

3D printing has never been exactly what I’d call "easy," but Anker really turned my head with its multi-part pitch:

One year later, how did Anker do? Personally, I’m experiencing an awful lot of whiplash.

I’ve now spent several months with two AnkerMake M5 printers, burning through multiple spools of filament to produce dozens of parts, and I want to be clear: you can genuinely get decent functional parts out of an AnkerMake M5, at remarkably fast speeds, even if you’re a 3D printing beginner.

Recently, I nailed the hilt of a Legend of Zelda sword and a print-in-place tank with moving treads on my very first try. I made a bouncing ball out of TPU and printed see-through shapes out of transparent PETG without having to tweak a single setting — I simply dropped a model into the company's PC software, picked the right filament in "Easy" mode, and waited for a smartphone notification to let me know my print was done.

But it took me a lot longer than 15 minutes to get that far. Sure, that was enough time to assemble the printer's core components, but it took longer before I realized Anker didn't properly tighten the belts (and grossly overtightened the wheels and some screws) before the printer arrived at my door. Neither Anker's printed instructions or the LCD screen told me anything about fixing these issues, or even how to properly load filament. (Anker's head of marketing told me a year ago that the printer would offer one-button filament loading; the feature still doesn't exist.)

Then, I had to wait months for Anker to fix the printer's firmware, which refused to properly auto-level the bed, would forget key parameters when you shut it down for the evening, scraped prints with its nozzle and left strange cavities and lumps on every single print I tried. That's mostly fixed as of a March update — I can now cover the entire bed with a single sheet of thin plastic of mostly uniform consistency. (3D prints live or die on their first layer, so you always want to get off on the right foot.)

But the quality, at least with my review units, still isn't what Anker promised.

Above, you’ll see pictures of a case I printed for my DJI Mini 2 drone three different ways: at the top is one I printed on my old Ender 3 Pro at 50 millimeters per second, then one on the AnkerMake M5 slowed down to the same 50mm/s, and finally one at the AnkerMake M5's 250mm/s default speed on the bottom.

You don't need to zoom in much to see that my old Ender 3 Pro did a better job, with nice clean lines all the way up. The AnkerMake surface just doesn't have that smooth consistency the company promised, regardless of whether I speed the printer up or slow it down, tighten my belts and wheels, or even adjust the Z-block tension. The 3D printing community calls these lines "ringing" or "ghosting," and it's typically blamed on a printer's high-speed vibrations affecting the print quality. I see this effect on almost every part I’ve tried, and I’m not the only one.

In almost every other way, the AnkerMake's print quality is good! I really like my Legend of Zelda sword, and I was impressed by the M5's results on the Autodesk Kickstarter Geometry Test; it's a little weak at overhangs, but with good bridging, dimensional accuracy, and very few excess strings of plastic spiderweb hanging off its pointy little spires. Yet AnkerMake claims it got a 25.5/30 on that test with a perfect score on vibration, and that's not what I saw: my printers only managed a 21/30 using Anker's own pre-sliced model and a brand-new roll of the specified filament.

As you’d expect, those vibrations can get worse if you run the printer in the new "500mm/s" fast mode that Anker introduced this month. Here's a couple 3DBenchys so you can see what that looked like for me:

Surface quality isn't the only disappointment. I was looking forward to keeping this supposedly quiet printer in my house, but I quickly had to move it to the garage because of the constant fan noise even when idle — not to mention how the printer inexplicably performs its homing maneuvers by noisily smacking its parts around.

I also haven't had a single timelapse video worth sharing. Here's the promise vs. the reality:

Anker's timelapse feature is not smart enough to do the bare minimum: It doesn't even wait until the bed is in the same position before snapping each shot, so what you see is a print jerking around. (It’d also be real nice if it briefly turned on the printer's built-in light, so you could see the object I’m printing is blue — not white.)

But for me, Anker's biggest broken promise is its "AI camera," which has not worked even a single time in my months of testing.

Anker advertises that its camera should be able to detect three distinct types of issues:

In order to detect any of these, you currently need to use Anker's own slicer to create an AI model that it supposedly passes along to the printer, so it can — theoretically — constantly check whether the image it's getting from the camera looks like the right shape.

To put it mildly, the camera did not stop my prints when they slipped off the bed, nor when pieces broke off mid-print. I literally printed spaghetti on purpose and the camera did not detect it, to say nothing of the time a print accidentally became plastic pasta.

And of the four times my filament stopped coming out of the extruder (one of which was a jam; three of which were because the filament got caught on the reel, which unfortunately can't trip a printer's filament runout sensor), the AnkerMake M5 spent all four times merrily printing nothing in the middle of the air. The camera never noticed anything was wrong.

The only time error detection stopped my prints, it was for false positives, like when my black TPU ball's first layer was perhaps not what the camera expected. So it doesn't surprise me a bit that one of Anker's firmware updates turned off timelapse video and error detection by default.

And I could live with that, but for one nagging fear — that because of some poorly designed or manufactured part or some new firmware update, I will one day wake up to a printer that failed so catastrophically it’ll need to be repaired.

I haven't had that happen yet, but there's some reason to worry. AnkerMake's subreddit and Discord groups contain numerous horror images of failed prints exploding into a mushroom cloud of plastic that penetrates the entire print head, some right up to the circuit board. While some have luck melting it off with careful application of a hair dryer, a few find the hot plastic has melted critical components and it's time for an entire replacement extruder.

When customers report hardware issues, they attest in the AnkerMake Discord servers and subreddit, they’re sometimes expected to spend considerable time proving the problem exists before Anker agrees to ship them replacement parts, which they then have to install themselves.

Not everyone is having huge problems! I lurked in those AnkerMake communities for months, and I saw plenty of people say it's printing like a dream. (Tom's Hardware reviewer Denise Bertacchi, who tests 3D printers for a living, gave this machine four stars.) But every Discord moderator I spoke to agreed: Anker has a quality control issue. Not all machines are equal.

I wasn't able to get Anker PR to meaningfully speak to any of these alleged issues, or acknowledge the extruder at all. "The reports I have received from our customer service team and product managers show the M5 hardware issues are all within normal tolerance levels," Anker global PR head Eric Villenes told me in February.

For instance, he suggested that most V-wheel issues can mostly be solved "by simply moving the V-wheel back and forth a few times" and that Anker will step in if they’re actually damaged. He also said Anker's working to replace improperly installed USB-C cables on a case by case basis. The only component with a known issue: there was a batch of failing touchscreens that the company will replace for any affected user.

Otherwise, says Villenes, the company's focus is on software, and I do have to give Anker some credit there. In my first draft of this review, I was ready to write off the printer entirely, giving it one of the lowest scores in the history of The Verge. Back then, every single one of my prints had gaps and bulges, the machine couldn't stay connected to Wi-Fi, dripped filament where it shouldn't, the screen occasionally flipped upside down, and the slicer was an utter mess. Things have considerably improved since then, the company's added must-have features like Vase Mode and the ability to pause a print via Gcode (to, say, change filament colors), and I’m finally getting a bunch of prints I like.

I just hope it’ll only get better from here on out, because Anker isn't done changing things up. In late February, it announced it plans to switch its entire printer slicer software over to PrusaSlicer, and some firmware updates have broken things even as they’ve fixed others — like the one time the print head started shimmying whenever you preheated it, making it hard to load filament, or the current issue where the bed will sometimes refuse to heat up if it's under a certain temperature.

Anker has now had an entire year to get the AnkerMake M5's software right, but it still feels like a beta. And I have a really hard time recommending a product whose manufacturer is so obviously figuring it out as they go — particularly when the company's advertising it like a finished product and speedy rivals have arrived.

It's one thing if you’re catering to an audience of Kickstarter fans who are backing your idea at a substantial discount while admitting it needs serious work. It's something entirely different to sell that product at Amazon, B&H and Best Buy, all while promising it should work beautifully and intelligently and automatically protect you from failures, just fifteen minutes after you open the box.

Update, 4:21PM ET: Rephrased a line to avoid confusion; while you do need to use Anker's slicer to create the AI image, you can start with Gcode from other slicers like Cura and Prusa.

/ Sign up for Verge Deals to get deals on products we've tested sent to your inbox daily.