Multi-Color Printing: Real Waste Tested
Multi-color 3D printing is the feature every printer manufacturer now markets and almost nobody discusses honestly. The Bambu Lab AMS made four-color printing a $399 add-on and the industry has not stopped pretending that purge waste is a minor inconvenience. It is not a minor inconvenience. On a densely patterned four-color print, you can spend more filament on the purge tower than on the model. I have watched a 30g miniature generate 120g of waste. That is not a rounding error. That is the elephant in the room.
This guide is an honest comparison of every mainstream approach to multi-color in 2026: Bambu AMS, Prusa MMU3, Mosaic Palette 3 Pro, IDEX machines like the Snapmaker J1 and BCN3D Sigma, manual filament swaps, and painting after the fact. I am going to tell you what each method is actually good at, what it wastes, and which workflow fits which use case. At the end, if you only print multi-color twice a year, I will also tell you the answer you probably do not want to hear.
The Purge Problem, Stated Plainly
Every filament-swap system (AMS, MMU, Palette) shares one physical constraint. When the hotend changes from color A to color B, the first few centimeters of extruded color B still contain traces of color A. That cross-contaminated material has to go somewhere. It goes into a purge tower, a purge object, or it bleeds into the print itself and ruins it.
How much waste depends on:
- The hotend melt volume. A standard 0.4mm hotend holds roughly 50-100mm³ of molten plastic that must be flushed.
- The color contrast. Going from black to white needs more purge than going from gray to dark gray. White after black can require 700mm³ of purge per swap to come out clean.
- The number of changes per layer. A print that changes color 12 times per layer wastes 12 times as much as one that changes once.
- The geometry. A flat logo needs fewer swaps than a detailed figure.
Real Waste Numbers
I measured these on my own printers over dozens of multi-color prints, and they match what most community benchmarks report.
| Print type | AMS waste (Bambu) | MMU3 (Prusa) | Palette 3 Pro | Manual swap | IDEX | Painted |
|---|---|---|---|---|---|---|
| 2-color layer swap sign | ~10% | ~8% | ~5% | ~2% | 0% | 0% |
| 4-color logo plate | 40-60% | 25-40% | 15-25% | N/A | 0% | 0% |
| 4-color miniature (high changes/layer) | 200-400% of model weight | 100-200% | 60-120% | N/A | 0% | 0% |
| 2-color functional print | 15-25% | 10-20% | 5-15% | ~2% | 0% | 0% |
Yes, you read that correctly. A densely patterned four-color miniature on a Bambu AMS can generate four times its own weight in purge waste. That is not an exaggeration and it is not a defect. It is the nature of single-hotend filament swapping. The AMS does nothing wrong. Physics does.
If you are budget-sensitive about filament, multi-color via AMS is the most expensive-per-gram type of printing in your entire workflow. Decide if you accept that before you spend money on the hardware.
The Contenders
Bambu Lab AMS and AMS 2 Pro
This is the market leader and for good reason. The AMS is simple, reliable, and the integration with BambuStudio is the best on the market. You load four spools, the slicer shows a visual purge estimate, you hit print, and you walk away. No tuning. No tip-forming nightmares. No calibration dance.
How it works. The AMS holds 4 spools. On a color change, the current filament retracts from the hotend into the AMS buffer, gets cut at the tip, and the next filament feeds forward. The hotend purges the old color into a purge tower, a prime line, or into infill if you enable the waste-into-infill option.
Pricing in 2026.
| System | Price | Colors | Compatible printers |
|---|---|---|---|
| AMS Lite | ~$399 with A1 Combo | 4 | Bambu A1, A1 mini |
| AMS 2 Pro standalone | ~$299 | 4 | P1S, P2S, X1C |
| AMS 2 Pro in combo | ~$650-800 | 4 | P1S Combo, P2S Combo, X1C Combo |
| AMS Hub + 4 AMS units | ~$1200+ | Up to 16 | All compatible Bambu |
The A1 Combo at $399 is the cheapest reliable multi-color setup in the world. It is the reason most hobbyists saying "I want to try multi-color" end up with a Bambu. For a P1S or X1C, the AMS 2 Pro pairs with the core printer at around $650-800 total.
Strengths. Plug-and-play. Reliable tip forming. Full slicer integration. Multi-unit chaining to 16 colors. Works with any 1.75mm filament (Bambu RFID is a convenience, not a requirement). Enclosed P1S and X1C keep the filament dry during long prints.
Weaknesses. The highest waste of any automatic system. No way to reuse purge material. RFID filament reading on third-party filament requires manual setup.
Best for. Beginners. Hobbyists who print multi-color occasionally and do not want to tune a system. Anyone who cares more about reliability than waste.
Prusa MMU3
The MMU3 mounts on top of a MK4S or MK3.9S/MK3.5S and handles up to 5 filaments. Prusa has been iterating on the MMU concept for years and the MMU3 is the first version that most users report as "reliable most of the time" rather than "reliable if you are a mechanic."
Pricing in 2026.
| Configuration | Price |
|---|---|
| MMU3 Kit | ~$299 |
| MMU3 Assembled | ~$359 |
| MK4S Kit + MMU3 Kit | ~$1,098 |
| MK4S Assembled + MMU3 Assembled | ~$1,458 |
Strengths. Five filament slots instead of four. Meaningfully lower waste than the AMS thanks to Prusa's purge tuning. Open-source hardware and firmware, which means community fixes and customization. Works with essentially any 1.75mm filament without the vendor-lock conversation.
Weaknesses. Much more expensive to enter than an A1 Combo. More calibration to get running well. The MK4S is an open-frame printer, so ABS and ASA multi-color requires an enclosure you add yourself. The reputation damage from MMU2 and MMU2S lingers, fairly or not, and I still see forum posts about specific reliability edge cases.
Best for. Tinkerers. People already on Prusa hardware. Anyone who prioritizes slightly lower waste and 5 colors over the AMS's ease of use.
Mosaic Palette 3 Pro
The Palette is the outsider in this conversation and it deserves more attention than it gets. Instead of switching filaments at the hotend, the Palette 3 Pro splices filament pieces together upstream and feeds a single continuous filament strand to your printer. No tip forming in the hotend. No retraction dance. Less waste per swap.
How it works. The Palette sits next to your printer. It takes 4 input spools, cuts segments, and splices them end-to-end in the correct order. The resulting composite strand feeds into your normal hotend. Waste is limited to small splice transitions and a small startup buffer.
Pricing. Around $799 for the Palette 3 Pro. It is compatible with most printers that have a standard filament path (Bambu, Prusa, Creality, Voron, Ender, many more).
Strengths. The lowest waste of any automatic swap system I have measured. Works with almost any printer. 4 colors (8 with the Palette 3 Pro expanded).
Weaknesses. Significantly more complex setup than an AMS. Requires Canvas software for slicing, which adds a step outside your normal slicer workflow. Splice failures on certain filament combinations. The pre-print calibration is annoying until you learn it. Support is thinner than for Bambu or Prusa.
Best for. People who already own a good single-color printer and want multi-color without buying a new machine. Anyone who cares about minimizing waste more than minimizing complexity.
IDEX: Snapmaker J1, Artisan, BCN3D Sigma
IDEX stands for Independent Dual Extruder. Two hotends, two motion systems, moving independently on the same machine. When you are printing with color A, color B parks at the side. When you swap, color B moves in and color A parks. There is no purge tower because there is no cross-contamination. The two materials never share a hotend.
Hardware examples.
| Machine | Approx price | Build volume | Notes |
|---|---|---|---|
| Snapmaker J1 | ~$1,099 | 300x200x200 | True IDEX, enclosed, reliable |
| Snapmaker Artisan | ~$2,399 | 400x400x400 | IDEX + CNC + laser |
| BCN3D Sigma | ~$3,995+ | 420x300x200 | Professional, production-grade |
| Raise3D E2 | ~$2,499 | 330x240x240 | Educational and prosumer |
Strengths. Zero purge waste. True support-with-dissolvable-material workflows (PLA body with PVA supports, for example). Mirror and duplication modes that let you print two copies of a model simultaneously.
Weaknesses. Expensive. Slower than equivalent single-extruder machines because of parking and calibration overhead. The XY calibration between the two hotends is critical and drift causes alignment issues. Limited to 2 colors unless you accept manual swap on top of IDEX.
Best for. Anyone printing multi-material (PLA + PVA support, or PLA + TPU) rather than pure multi-color. Production shops. People who print multi-color frequently enough that waste matters more than the hardware cost. If you print 4-color miniatures twice a week, an IDEX machine will save its own cost in filament within a year.
Manual Filament Swap
The zero-cost option. You insert an M600 filament change command at specific layers in your slicer, the printer pauses, you swap the filament by hand, the printer resumes.
Strengths. Zero extra hardware. Near-zero waste. Works on any printer that supports a pause command. Perfect for signs, logos, layer-based color changes.
Weaknesses. You must be present for every swap. No within-layer color changes. Two swaps into a four-color print and you are losing the afternoon.
Best for. Occasional multi-color users. Anyone doing 1-2 layer swap projects per month. A cheap way to learn if you even need a multi-color system before buying one.
Painting After Printing
The forgotten answer. Print in one neutral color, sand, prime, and paint the details with a brush. Professional miniature painters and serious cosplayers almost never use multi-filament systems. They paint, because paint is infinite colors, supports gradients, fixes layer lines with primer, and generates zero filament waste.
Head-to-Head: Which Method Wins What
This is the table I wish someone had given me three years ago.
| Use case | Best method | Why |
|---|---|---|
| Multi-color signs, logos, two-tone nameplates | Manual swap | Fewest changes, lowest waste, no hardware |
| 4-color flat prints (keychains, coasters) | Bambu AMS | Fastest path from "idea" to "done" |
| Detailed multi-color figures (miniatures, busts) | Painting on a single-color print | Resin printer + brush beats any AMS for quality |
| Multi-material (rigid + flex, PLA + TPU) | IDEX (Snapmaker J1) | No cross-contamination, different materials |
| Dissolvable supports (PLA + PVA) | IDEX with PVA | Only way to get clean dissolved supports |
| Production multi-color, many colors | AMS Hub (16 colors) or IDEX depending on color count | Depends on palette size |
| Budget entry, experimenting | A1 Combo at ~$399 | Cheapest reliable AMS |
| Existing printer, want multi-color without replacing it | Palette 3 Pro | Works with most printers, lower waste than AMS |
| Occasional multi-color, hate waste | Manual swap or painting | No ongoing cost |
What Most Multi-Color Guides Get Wrong
Three patterns show up in almost every article I see on this topic.
They hide the purge number. Every guide lists AMS pros and mentions waste as a small bullet point. "Some filament is wasted during color changes." That sentence hides the truth that purge can exceed the model weight by 2-3x on complex prints. If a guide does not give you concrete waste numbers, it is not a guide. It is marketing.
They skip Palette and IDEX. Every guide is Bambu AMS versus Prusa MMU3 and nothing else. Palette 3 Pro and IDEX are fundamentally different approaches to the same problem and both are better than filament-swap systems for specific use cases. Leaving them out makes the comparison dishonest.
They ignore painting. Miniature painters, cosplayers, and anyone who cares about finish quality paint their prints. Any guide that treats multi-color printing and painting as unrelated topics is written by someone who has not finished a detailed piece by hand. For figures and cosplay, paint is the answer 80% of the time.
The Waste Mitigation Playbook
If you already own an AMS or MMU and want to reduce the waste you generate, this is what actually helps.
Print the Purge Into Infill
BambuStudio and OrcaSlicer both support "flush into infill" and "flush into object" options. Instead of a standalone waste tower, the purge fills the interior of your current model. Zero visible waste, zero net extra filament. Only works when the model has enough infill volume to absorb the purge, which is usually true for medium to large prints.
Print the Purge Into a Secondary Object
You can load a second object onto the build plate (a calibration cube, a filament sample, a simple bracket) and tell the slicer to flush into that object instead of a tower. You now get two prints for the price of the purge on one. I run this on every multi-color print that does not have enough infill to absorb the flush.
Tune Flush Volumes by Color Pair
The default flush volumes in BambuStudio are conservative because Bambu cannot predict every filament combination. For specific color pairs you use often, run a test print and lower the flush volume until you see contamination, then bump it back up a small amount. Some color pairs need only 30-40% of the default flush. Others need 110%. There is no universal number.
Use Fewer Colors, Smarter
Most 4-color prints would look almost identical as 2-color prints. Plan the model around contrast instead of palette. A 2-color print generates a fraction of the waste of a 4-color print because each layer has fewer changes. The most efficient multi-color print is one where each color appears once.
Accept That Some Prints Should Be Painted
If your model has 20 color changes per layer across 200 layers, you are about to run a 40g model with 200g of waste. Print it in gray and paint the details with acrylic. You will save an hour of print time, a roll of filament, and the result will look better.
Manual Swap Setup, In Detail
Because this is the cheapest option and worth taking seriously.
In BambuStudio or OrcaSlicer
- Slice the model normally with a single filament.
- In the preview view, scrub the layer slider to the layer where you want to change color.
- Right-click the layer and select "Add color change" or "Add filament change (M600)."
- Repeat for each change layer.
- Re-slice. The printer will pause at each change and beep.
In Cura
- Slice the model.
- Open Extensions → Post Processing → Modify G-Code.
- Add the "Filament Change" script.
- Set the layer number for the change. Cura will insert an M600 command.
- Save the sliced file and print.
In PrusaSlicer
- Slice the model.
- In preview, drag the filament change marker on the vertical layer slider to the desired layer.
- Re-slice. A pause command is inserted automatically.
During the Print
When the printer pauses, the hotend stays hot and the current filament retracts slightly. Remove the filament manually, feed the new color in, purge a small amount with the extruder control until you see clean color, then resume. Do not skip the purge step. If you do, the first few millimeters of the next layer will be cross-contaminated.
Finding Multi-Color Models
Multi-color prints need files designed with multi-color in mind. A single-body STL will not work. You need:
- Multi-part STL sets where each color is a separate file, imported into the slicer as aligned objects
- 3MF files with color information embedded
- Model pages explicitly tagged "multi-color," "AMS," "MMU"
Search 3DSearch with "AMS" or "multicolor" added to your query. For printer-specific profiles, the A1 mini settings guide and the X1C vs P1S comparison cover multi-color print tuning for Bambu hardware. If you are picking filament brands for multi-color work, the best filament brands for 2026 shortlist is where I start.
Common Mistakes That Waste Filament and Time
Starting with a 6-color model on your first AMS print. Start with 2-color, get comfortable with the flush behavior, then graduate. Your first multi-color print should be a layer-change sign, not a figurine.
Ignoring wet filament. Wet PETG or PLA makes cross-contamination worse because the moisture causes pops and fluctuations at the hotend. Dry your filament before any multi-color print where finish matters.
Leaving the default flush volumes untouched for a month. The defaults are conservative. Tune them for the color pairs you print most. Ten minutes of tuning saves grams on every print for the life of the printer.
Printing multi-color miniatures on an AMS instead of painting them. This is the single biggest mistake I see. Miniatures have tiny, intricate color boundaries that generate hundreds of color changes per print. You will waste more filament than the model weighs. Print it in gray and paint it.
Using open-spool PETG in the AMS for 24+ hour prints. PETG absorbs moisture. A long multi-color print with PETG in an open AMS chamber degrades the filament over the course of the print. Use a dry box feed or the enclosed AMS 2 Pro with a desiccant.
Not planning flush into infill. Every BambuStudio multi-color print should have "flush into infill" enabled by default unless the model has zero infill. It is a free waste reduction.
Forgetting that the printer is slower multi-color. A 2-hour single-color print can be a 5-6 hour multi-color print because of swap overhead. Do not plan a multi-color job if you need the printer back in 3 hours.
Which Should You Actually Buy?
The honest answer, without brand loyalty.
If you have no printer and want to try multi-color: Bambu A1 Combo at $399. It is the cheapest reliable entry and it will not fight you. Accept the waste and move on.
If you have no printer and want multi-material (PLA + TPU, or PLA + PVA supports): Snapmaker J1 at ~$1,099. It is the only category where IDEX is the clear correct answer.
If you already own a single-color printer and want multi-color without replacing it: Mosaic Palette 3 Pro at ~$799. It works with almost any printer and has the lowest waste of any swap system.
If you are already on Prusa: MMU3 at ~$299 kit. It pairs cleanly with your existing MK4S workflow. Yes, the internal link anchor is intentional as a reference point for settings workflows.
If you print multi-color once every few months: Do not buy anything. Use manual M600 swaps. Save the money for filament.
If you print detailed figures and care about finish: Do not buy a multi-color system. Buy a $200 resin printer for the figures and a brush. Paint is still the best multi-color tool for minis and cosplay.
The Honest Close
Multi-color 3D printing in 2026 is simultaneously the most accessible it has ever been and the most wasteful type of printing you can do. Both things are true. The A1 Combo at $399 makes four colors available to anyone with a desk and a power outlet, and that is legitimately impressive. The fact that the same machine can generate triple its own model weight in purge on a detailed print is also true and nobody in the industry wants to talk about it.
Match the method to the job and you will avoid 90% of the regret. Layer swaps for signs. AMS or MMU for flat multi-color prints and functional parts. Palette for low-waste convenience on existing printers. IDEX for multi-material and dissolvable supports. Painting for detailed figures and cosplay. Anyone who tells you one tool solves every multi-color case is selling something.
When you are ready to find models that actually print well in multi-color, search 3DSearch for files specifically tagged for AMS or MMU. For related reading, the best filament brands for 2026 list is where I buy multi-color sets without RFID lock-in, and the stringing fix guide is relevant because multi-color prints surface stringing problems faster than single-color prints due to the extra retractions per layer.
Search for related 3D models
Find 3D models related to this article
Search across 6 platforms including Printables, Thingiverse, and MakerWorld in one place. Get AI-powered slicer settings tailored to your printer.