I'm a procurement manager at a 180-person engineering firm. I've been managing our additive manufacturing budget — around $95,000 annually across consumables, service contracts, and machine depreciation — for about three and a half years now. Over that stretch I've signed off on 40+ orders tied to our printer fleet.
If you're reading this because you're trying to decide between an Ultimaker S3 and an S5, I'll save you the suspense: there's no single right answer. It depends on how you actually use the machine, not on which one has a better spec sheet. I've watched two teams in our own building buy the "wrong" one for their workflow — in opposite directions.
So instead of a verdict, here's a framework. I'll walk you through three common scenarios, then give you a short self-check at the end to figure out which one you're in.
First, what actually separates these decisions
The obvious differences — build volume, dual extrusion, material compatibility — are real, but they're not the deciding factor for most buyers. The deciding factor is usually how many hours a week the machine runs and how many people touch it.
I use two rough thresholds to sort teams into buckets: effective print hours per week and number of distinct users per month. Those two numbers predict total cost of ownership far better than any spec comparison. If you want to skip my reasoning, jump to the self-check at the bottom. If you want the reasoning, keep reading.
Scenario A: Occasional use, one or two people
This is a design lead or a small R&D function running a handful of prints per week. Maybe 10–15 hours of actual print time. Maybe less. Mostly PLA, PETG, maybe some Tough PLA when you need a functional prototype.
For this team, the S3 is usually the sensible call. It handles the materials you'll actually use, the build volume covers most prototypes under 200mm, and the footprint fits on a bench. From the outside, the S5 looks like the "safer" upgrade — bigger bed, more future-proof. The reality is that bigger bed sits idle 80% of the time in this scenario, and you've paid for it.
That said — and this is where I've seen teams get burned — the S3's single extruder becomes a real limitation the moment you start doing multi-material work with breakaway supports. If you're already printing dissolvable supports for internal geometries, you're not really a Scenario A team. You just haven't admitted it yet.
We didn't have a formal "who actually needs dual extrusion" review before our 2023 purchase. Cost us when the team started prototyping overhangs three months in and had to re-buy.
Scenario B: Steady throughput, a handful of engineers
This is where things get interesting. Roughly 20–50 print hours per week, 3–8 users, mixed materials — nylon, PC blends, dissolvable supports, the occasional CF-reinforced filament. The machine is on most of the day.
The conventional advice here is to buy the S5. Bigger chamber, dual extrusion, higher temp ceiling. That advice is mostly right, but I'd frame it differently: in this scenario, the S5 usually has a lower cost per successful part than the S3 — not because it's cheaper, but because failed prints are more expensive than the machine price difference.
Let me put numbers on that. If your S3 fails one print per week due to chamber temperature issues or support-removal damage, and each failure costs you $8–15 in material plus 4–6 hours of someone's time, you're looking at $600–1,200 per quarter in hidden waste. Over an 18-month horizon, that closes most of the gap to the S5's higher upfront cost.
To be fair, this math only works if your failure rate is actually tied to the machine, not to user error. I've seen teams blame the hardware when the real issue was nobody agreed on print profiles. We fixed that with a shared Cura profile library before blaming the printer.
The S5's dual extrusion also matters more than people expect once you're running production-adjacent work. Water-soluble supports change what geometries are feasible, and that changes what parts your team can actually deliver. It's not a specs argument; it's a capability argument.
Scenario C: Production-adjacent, multi-machine
When you're running north of 60 print hours per week across multiple machines, the single-unit decision stops mattering. What matters is fleet management. This is where Ultimaker's Digital Factory platform and Cura's enterprise features — centralized job monitoring, remote print management, shared material profiles — start to carry real weight.
The counterintuitive advice for this scenario: buy the smaller machines, and buy more of them. Four S3s will often beat two S5s in throughput per dollar, simply because you parallelize jobs, queue redundancy is built in, and a single machine going down doesn't halt production. The S5 wins on single-part size, but if your parts fit in the S3 envelope, the math usually favors numbers over size.
If I remember correctly, when we restructured our fleet in early 2024 we came out roughly 22% ahead on cost-per-part with a distributed small-machine layout. I might be misremembering the exact figure — the point is the direction, not the decimal.
My experience here is based on about 40 machine-level purchases in an engineering environment. If you're running a dedicated print farm that operates 24/7, your cost model will look different — at that scale you're buying uptime contracts and spare-part inventory, not just hardware.
How to tell which scenario you're in
Answer these three questions honestly. Don't round up to the answer you want to hear.
- How many hours did your current printer actually run last month? Pull the job log, don't estimate. If it's under 60 hours total, you're Scenario A. If it's 80–200, you're likely B. If you're printing more than that on a single machine, or running multiple machines, you're C.
- How many people touched the printer in the last 30 days? One or two: A. Three to eight: B. More than eight, or any shared queue: C.
- How many print failures did you have where the fix was "we needed a different material" or "we needed dissolvable supports"? Zero: A. A few: B. Constant: you're probably already C and just haven't bought the second machine yet.
One more thing worth saying: the cheapest option is rarely the cheapest option. I've watched a team save roughly 30% on a machine purchase only to burn it back within a year on failed prints, staff time, and one emergency service call. That's the trap. Buy for how you'll actually run the machine over 24 months, not for what looks like the best deal today.
Prices and specs shift — verify current Ultimaker configurations with an authorized reseller before you decide. The framework, though, tends to hold up.
