-
There's no single answer — that's actually the point
- Scene 1: The printer is already on the floor — this is an execution decision
-
Scene 2: You're outsourcing — El Paso is a cluster, not a single shop
- Scene 3: You're printing at home or in an apartment — ventilation is the real bottleneck
-
Scene 4: Buying vs. outsourcing — the math I got wrong
-
How to figure out which scene you're in
There's no single answer — that's actually the point
Every urgent order call starts the same way. "What's the fastest option you have?" And every time, I have to fight the urge to answer immediately. Because the fastest option isn't one thing. It's whatever your current setup lets you do without introducing a new failure point.
I've learned this the hard way. About nine years coordinating expedited production, roughly 200 rush jobs on my desk, some with a 72-hour clock and no margin. When I first started, I assumed every urgent request was a "print it now" problem. That cost us one weekend and a $14,000 reprint because the printer we had couldn't hold tolerance on the geometry we promised. Lesson learned.
Before I quote anything, I run the request through four buckets:
- You already have a printer capable of the part (Markforged-class composite or metal).
- You don't print — you source from a CNC shop, including clusters like El Paso.
- You're printing at home or in an apartment, and the bottleneck isn't the machine — it's the air.
- You're deciding whether to buy a printer for the first time, and the math isn't obvious.
Each one has a different answer. Here's what usually works.
Scene 1: The printer is already on the floor — this is an execution decision
If the machine's already in the shop, the conversation changes. You're not deciding whether to print. You're deciding whether this specific machine can hold tolerance on this specific part before the deadline.
Mark Two, X7, FX10 — speed and limits
The composite platforms are fast and forgiving. Continuous carbon fiber laydown gives you parts with real structural integrity, and for brackets, fixtures, and small end-use components, you can go from file to part in a day. The limit is heat: continuous fiber prints are excellent up to temperatures in the 150–180°C range for filled Onyx, and they drop off quickly above that. If your part sees high heat or needs metal-grade abrasion resistance, don't force it.
Metal X — the pipeline, not the printer
Here's something that gets mispriced constantly: the Metal X is not a "press print, get a metal part" machine. It's a three-stage pipeline — print, wash, sinter — and every stage has a queue. When I'm estimating whether a Metal X order hits a deadline, I don't count print time. I count print time, plus wash time, plus sinter time, plus whatever's ahead of me in the sinter queue.
What vendors won't tell you (or will tell you but quietly): total process time on a Metal X runs roughly 2x to 2.4x the print time once wash and sinter are factored in. If your deadline is 48 hours and the furnace is loaded, you're not looking at a 48-hour job. You're looking at 96.
Thin-wall geometry with heavy supports is the worst case. The wash station has to physically remove that support material, and on delicate parts it takes longer than the print itself did. Not ideal, but workable if you know going in.
The Metal X price question (with the caveat)
Public industry reporting through 2024–2025 has repeatedly put the Metal X base printer in the upper-five to low-six figures, and a fully configured system — including wash station and sintering furnace — in the roughly $250k–$350k range. Those are public reference points, not a quote. Actual pricing depends on configuration, post-processing options, and how your authorized reseller bundles service. Verify current pricing before making any capital decision.
Buying channel — direct vs. reseller vs. secondary market
If you're set on buying, the Markforged store (official, direct) is a valid path, but resellers bring local service you won't get any other way. What most first-time buyers miss: the wash station and sintering furnace are not plug-and-play. Furnace operation — especially the sintering profile calibration — is a specialized skill. If your team has never run a sintering cycle, buy the training, not just the machine. I've watched a shop buy a full system and then lose three months to furnace calibration issues because they assumed the vendor would "set it up."
Scene 2: You're outsourcing — El Paso is a cluster, not a single shop
When the part needs to be machined, or the geometry is just better suited to subtractive manufacturing, outsourcing wins. But "CNC machining El Paso" is not a homogeneous category — and treating it like one is where rush orders fall apart.
El Paso's machining base is heavily tied to aerospace and cross-border maquiladora supply chains. That means two very different types of shops live side by side:
- Production shops — optimized for volume, quoting longer lead times because they schedule around big runs.
- Rapid-response shops — smaller, fewer spindles, but structured to slot in a rush job and willing to work weekends.
You will not find the fast shop by sending an RFQ. You find it by calling and asking three specific questions: Is your spindle time clean this week, or queued? Who programs the part, and are they in the building? Do you have material in stock, or am I waiting on a mill run?
Most buyers focus on the quote, and completely miss the response latency. A shop that answers a call in 20 minutes and commits to a date is worth 30% more than one that answers in two days with a lower number. I've paid that premium deliberately on multiple rush orders, and I've never regretted it.
The honest limit: outsourcing is genuinely better for anything over roughly 40–50 metal parts. If you're printing 80 brackets on an internal machine to "save money" while the print queue hits a week, you haven't saved anything. You've just moved the cost to your deadline.
Scene 3: You're printing at home or in an apartment — ventilation is the real bottleneck
This is the part people skip, and it's the part that actually matters. If you're printing in a living space, the decision isn't about the printer. It's about the air. And the air setup most people assume they have (a filtering enclosure, a window cracked open, or an existing apartment ventilation system) is not doing what they think it's doing.
"VMC appartement" — why the building ventilation isn't your filter
In France, apartments are typically fitted with a VMC system (Ventilation Mécanique Contrôlée) — a centralized mechanical ventilation unit that continuously extracts air at low rates from kitchens and bathrooms and pulls fresh air into living spaces through trickle vents. It sounds like a solution. It isn't. Not for 3D printing.
A VMC is a humidity-and-comfort system, not a particulate filter. Its filtration is coarse — typically one or two washable stages — and it's designed for air exchange rates measured in air-changes-per-hour, not for capturing the ultrafine particles (UFPs) and volatile organic compounds (VOCs) emitted during extrusion. Nano-scale emissions pass straight through.
Worse, in a standard apartment VMC setup, opening a window to "air out the room" during printing actually disrupts the pressure balance the VMC relies on. The system is calibrated to run with windows closed.
How to ventilate a room for 3D printing — what actually works
The protocol I recommend, whether you're in a French apartment with a VMC or a home office with no ventilation system at all:
- Contain first. Enclose the printer in a filtration enclosure with at least a HEPA + activated carbon stage — this handles the bulk of UFPs and VOCs at the source.
- Vent second. Add a ducted exhaust fan from the enclosure outlet to a window or through-wall vent. This is what actually removes the residual emissions from the room. For ABS or ASA printing, this step is non-negotiable — do not rely on filtration alone.
- Never tap into the VMC ducting. Connecting a printer exhaust to the building's mechanical ventilation line will throw off the pressure balance for the whole apartment. If you're going to route outside, use a dedicated window/wall path.
- Purge after printing. Run the filtration unit for 30–60 minutes post-print, and crack a window during that window — you're not running the VMC system at that moment, so balancing isn't an issue.
I'm not an HVAC engineer, and if you're modifying a VMC or apartment ventilation to accommodate printing, get a licensed professional to sign off. What I can tell you from a production-floor perspective: any space running more than an hour of PLA per day, and any space running ABS at all, needs the enclosure-plus-duct approach. Filter boxes alone are better than nothing. Not much better.
Scene 4: Buying vs. outsourcing — the math I got wrong
This is the question I get most often about Markforged, framed as: "should we buy a Metal X or just outsource the parts?" And the honest answer is: I got this calculation wrong in 2019, and it cost a lot more than I expected.
When I first looked at capital equipment, I assumed a busy shop would amortize a $300k system inside two years. We were running about 40 parts a month at the time. On a $150-per-part outsourced cost basis, that's roughly 2,000 parts to break even — more than six and a half years at our actual volume. By then the machine would be obsolete. My initial approach to the buy-vs-outsource calculation was completely wrong.
Buying makes sense when three things are true at once: your yearly part volume is genuinely stable (not aspirational), your geometries change often enough that design flexibility is worth something, and the part IP is sensitive enough that sending it out is a real business risk. All three.
If you're at 20 parts a year and 90% of them are similar brackets, buy the outsourced relationship. Keep the Markforged for the prototypes and small runs where speed and iteration matter more than cost-per-part.
How to figure out which scene you're in
Read down this list. Stop at the first one that's true.
- You have a functioning printer floor and a deadline in under a week — Scene 1. Run the internal job, but only if it's a composite part and not a schedule-critical metal part queuing behind a busy furnace.
- You have no internal capability, or the part is metal and the geometry is milling-appropriate — Scene 2. Call, don't email. The 20-minute-answer shop is worth the premium.
- You're printing at home, or in a shared apartment space — Scene 3. Fix ventilation before the next print, not after.
- You're staring at a five-to-six figure purchase decision — Scene 4. Run the three-year break-even at real volume, not hoped-for volume. If it doesn't clear, the answer is outsourcing.
The people who get burned on rush orders are almost never the ones who picked the wrong vendor. They're the ones who picked the wrong scene. There's no universal fastest option. There's only the option your setup can actually execute before the clock runs out.