Prusa has just pushed a firmware update that cuts MMU3 filament change time enough to shave hours off real multi-color prints.
The upgrade targets the Original Prusa MMU3, the company’s five-filament, single-nozzle multi-material unit used on machines like the MK4S and CORE One. In the new release, Prusa says a typical change drops from 52 seconds to about 42 seconds, a roughly 20% improvement, with up to nine seconds saved per swap depending on setup. That sounds incremental until you remember that complex models can trigger hundreds or even thousands of swaps.
Single-nozzle multi-material is still the most common path for hobbyists and many pros because it avoids the cost and mechanical complexity of toolchanging. While simple in concept, unloading, loading, purge routines, and wipe tower moves can soak up a lot of print time. Prusa claims that on a complex five-color model, up to 60% of total time can be spent on non-printing tasks, even with an MMU system that is already considered relatively efficient. That definitely rings true for anyone who’s done multicolor 3D prints.
Prusa Research demonstrates the impact with straightforward arithmetic. For example, they describe a “moderately complex” five-color print might require 400 filament swaps, which at nine seconds saved per swap is about an hour gained. At the high end, a model with 2,200 changes could save around five and a half hours, and the company cites an example with 3,777 changes translating into about 9.5 hours saved. For small businesses, the same improvement applies to batches: a run of 25 parts with 150 changes each is suddenly a lot less painful to schedule. This could mean a notable increase in throughput, and that is critically important for, say, print farms that intend on running the equipment 24/7.
What Changed In The MMU3 Swap Cycle
Prusa’s explanation is quite interesting because it is not a new sensor or a new piece of hardware; it is time recovery through more careful motion planning and sequencing. The company says unload/load movements plus “cooling moves” on the wipe tower represent about 60% of swap time for PLA and 70% for PETG. That means there is a lot of potential to shave of small waits and dead time.
As an example, one of the many improvements is “parallel idler disengage.” Previously, once the extruder had the filament, it still waited for the MMU to fully disengage an idler mechanism, costing about two seconds. Now the extruder starts pushing filament toward the nozzle while the MMU disengages in the background. Another optimization is “predictive idler staging,” where the idler is moved toward a ready position during unloading so the next engagement does not start from a fully parked state.
Prusa also mentions a smoother unload ramp-up, effectively a refined pressure relief step ahead of retraction. The goal is to avoid extruder motor stalls while still increasing retraction speed. They don’t explain all of the improvements, but you get the idea here: cut out any part of the workflow that is causing a wait, and do things in parallel if possible.
Taken together, these optimizations trim hundreds of milliseconds from multiple steps without really affecting any of the overall workflow steps.
Compatibility, Setup, And A CORE One L Detour
According to Prusa, supported printers benefit immediately, including MK4S, MK3.9, MK3.5, and CORE One/One+ systems. Users need both updated printer firmware and updated MMU3 firmware, listed as version 6.5.3 for the printer and 3.0.4 for the MMU3. Prusa also points out that PrusaSlicer profiles should be refreshed via the Configuration Wizard, and that front PTFE length settings matter, because tube length can add small delays to each load.
There are quality-of-life changes as well, including a “Preload All” option intended to reduce menu gymnastics when loading multiple spools of the same type. Prusa also advises against using high flow (CHT) nozzle styles for MMU printing, saying the real-world speed gain is minimal while standard nozzles better support clean unloads and consistent tip forming. That is definitely the case, as a job can easily fail if the MMU is unable to load the filament at any point; reliability takes priority here.
Perhaps the most consequential news is strategic: Prusa says it will not develop an official MMU3 kit for the larger CORE One L. The reason given is the upcoming INDX system, which Prusa positions as the next step, with higher speed and up to eight toolheads. Instead, Prusa will publish adaptation guidelines for users who want to modify a CORE One+ MMU3 setup to work on the CORE One L, including printed conversion parts and separately sourced larger buffer plates, with STEP/DXF files provided.
They also say that this upgrade to the MMU system will be the last, confirming their strategy to move everything (at some point) to INDX-style extrusion.
The big question is how many users will bother adopting the faster MMU3 workflow versus waiting for INDX, especially if multimaterial 3D printing becomes more popular. On the other hand, this improvement is mainly a firmware upgrade and some tube tweaking.
In single-nozzle multi-material 3D printing, nine seconds on each layer is a huge amount of wasted time, and now it can be recovered.
Via Prusa Research
