3D Printers - Print Accuracy Prusa MK4S vs Anycubic K3M

Last Update: 02.05.2025

During printing of test objects for DnD board games I found that the Prusa MK4S and the Anycubic Kobra 3 Max (K3M) have a little different print accuracy. So, I was curious to investigate and note my findings on my website.

As test print I used the hole test model I created in the past, as the differences were notable for hole structures. Thus, I printed only 5mm height of the model to save time and material.

Outer dimensions

The prints I got were measured using a caliper and I got the following info for the outer dimensions (all values are in mm):

Printer Setting Width Length Height Avg. Deviation
  3D Model 79.00 23.00 5.00  
K3M Standard 78.91 22.88 4.87 0.11
K3M Standard+LineWidth (*) 78.90 22.92 4.85 0.11
K3M Standard+HoleComp (**) 78.96 22.92 4.89 0.08
MK4S 0.2mm structural 78.85 22.88 5.04 0.08
MK4S 0.2mm speed 78.88 22.87 5.03 0.07

(*) 0.2mm Standard profile with 0.4mm nozzle, but Line width Default, Outer wall, etc. increased to 0.45mm (from 0.42mm), as the PrusaSlicer uses such widths for the MK4S

(**) 0.2mm Standard profile with 0.4mm nozzle, but the “X-Y hole compensation” set to 0.15mm instead of the default 0.05mm

So, in total really nice results and within the accuracy (not resolution!) of the caliper.

Hole dimensions

Going further to the hole dimensions I got the following measurements (all values are in mm):

Printer Setting D2 D3 D4 D5 D6 D7 D8 D9 D10 Avg. Deviation
K3M Standard 1.72 2.84 3.80 4.81 5.80 6.86 7.80 8.87 9.88 0.18
K3M Standard+LineWidth (*) 1.78 2.79 3.77 4.82 5.64 6.72 7.75 8.69 9.73 0.26
K3M Standard+HoleComp (**) 1.97 2.98 3.94 5.02 6.00 7.10 8.04 9.03 10.01 -0.01
MK4S 0.2mm structural 1.83 2.80 3.93 4.92 5.92 6.90 7.93 8.96 9.91 0.10
MK4S 0.2mm speed 1.91 2.87 3.95 4.93 5.92 6.95 7.95 8.98 9.97 0.06

So the X-Y hole compensation setting was very usful for the K3M. I was surprised that the speed profile gave a better average deviation than the structural profile on the MK4S, but with only one print per setting it is hard to say if it was just a lucky print or not.

Measurement setup

Hardware used:

Software used: