Multi-material printing is one of those features that looks magical in a thirty second timelapse and rather different once you have actually lived with it for a month. I have, so here is the honest version, written from a single-nozzle setup with a multi-material upgrade rather than a true multi-head machine. Most of what follows is about that single-nozzle compromise, because that is what most people at home are wrestling with.
The promise is obvious. Two colours, or two materials, in one part. A logo in a contrasting colour. Soluble support that washes away instead of leaving you to snap brittle PLA out of an overhang with a craft knife and a bad temper. Done well it is genuinely brilliant.
The mechanics, and where the waste comes from
On a single-nozzle, single-hotend setup, "multi-material" means the printer swaps filament at the extruder, not the toolhead. Every colour change is a full unload-and-reload cycle: retract the old filament back past the heatbreak, feed the new one in, then purge until the colour coming out the nozzle is actually the new colour and not a muddy in-between.
That purge is the catch. The hotend is a mixing chamber whether you want it to be or not, and you cannot get clean colour two until you have flushed colour one through it. So every change throws away a few centimetres of filament. The slicer manages this with a wipe tower, a little sacrificial pillar printed alongside the real part where the transitions happen.
On a print with frequent colour changes, the wipe tower can use as much filament as the part itself. I had a two-colour keycap set where the tower out-massed the caps by a comfortable margin. That filament is not coming back. You can tune the purge volumes down per material pair, and it is worth doing, but there is a floor below which you get colour bleed and the contrast goes muddy.
The tuning tax
The other cost is time, both wall-clock and yours. Each filament swap adds real seconds, and on a part with hundreds of layers and a change every layer, those seconds compound into a print that takes two or three times as long as the single-colour equivalent. A four hour print becomes an overnight one.
And every new material pairing wants tuning. Purge volume, swap temperatures, the ramming and cooling moves the firmware uses to shape the filament tip so it feeds cleanly next time. Get the tip shaping wrong and you get a blob that jams on reload, which on a single nozzle means a failed print and a cold-pull to clear it. I keep a little table:
pair purge(mm³) notes
PLA->PLA 70 same brand, fine
PLA->PETG 140 bleed if lower; let it flush
PLA->PVA 90 soluble support, slow swaps
That table is hard-won. Nobody hands it to you; you print test swatches and dial it in.
So is it worth it?
For soluble support, unreservedly yes. Washing PVA away to reveal a clean overhang that would otherwise be a destructive nightmare to remove is the single best argument for the whole setup, and on functional parts it changes what geometries are even sensible to print.
For decorative multi-colour, it depends entirely on how much you value the result versus the waste and the time. A logo or a two-tone case, occasionally, lovely. A rainbow of frequent changes across a tall part, you are essentially printing a second object out of waste filament to get the first one, and a paint pen or a printed insert glued in might get you there with a tenth of the bother.
What I have settled on: I keep the multi-material upgrade fitted because the soluble-support trick alone earns its keep, and I reach for colour changes sparingly, on parts where the swaps are few and the wipe tower is a rounding error rather than the main event. Treated that way it is a quietly useful tool. Treated as the default it is a filament shredder with a print attached.