The short answer
If you're sourcing Velo3D for production metal parts in 2025, treat it as an aerospace-tier capital purchase—not a manufacturing upgrade. It doesn't replace a fiber laser for metal cutting, and it doesn't replace a VMC. What it does is print geometrically complex metal parts that would otherwise require multi-piece assembly. If your shop is already paying for that assembly, the math might work. If it isn't, it probably won't.
For most companies evaluating this: don't buy. Contract the parts out. The economics only make sense once you're producing hundreds of complex metal parts a year—or when the design simply can't be made inside a traditional process window.
Why you can trust this (and what I don't know)
Let me get the positioning out of the way first. I'm not a metallurgist, and I've never run a metal printer. Our team processes roughly 70–80 vendor orders a year, and only a handful of those are capital equipment. I'm the person who has to read the spec sheets and translate between engineering and finance—two groups that never agree on what "worth it" means.
So take everything I say about laser parameters, scan strategies, and grain structure as secondhand. It's what engineers tell me, filtered through what I can actually verify.
What I do have firsthand experience with is the economics of this kind of purchase. And honestly, the price on the quote and the total cost over three years are almost never the same number.
What Velo3D actually is—and what it isn't
Velo3D makes laser powder bed fusion systems. The Sapphire line is the flagship platform. The pitch is that their technology lets you print complex metal parts with minimal or no support structures—which matters because supports are the messy part of metal AM. They're expensive to remove, they leave surface defects, and they limit what geometry you can actually build.
The customer list is public: SpaceX, Anduril, and others in aerospace and defense. That's a real signal. You don't get into those supply chains without passing qualification. But—and this is important—it's also a signal about who the machine is actually built for. These are companies with dedicated AM engineering teams, quality systems, and a tolerance for a long learning curve. If you don't have any of that, the customer logo list is not a promise about your experience.
What people often miss: the client roster tells you who can afford the ecosystem, not who can afford the machine. Those are different budgets.
What it competes with (and what it doesn't)
Three things get compared against each other in these conversations, and they shouldn't:
- Fiber laser cutting — slices flat stock into 2D profiles. Fast, cheap per part, limited by thickness and geometry.
- VMC (vertical machining center) — subtracts material from a solid block. High precision, wide material range, constrained by tool access.
- Metal AM (Velo3D and peers) — builds parts layer by layer. Highest geometric freedom, highest cost per part, lowest throughput.
If you're putting these on the same comparison sheet, you don't have a problem yet—you have a decision to make about which problem you're solving. Laser and VMC are the cost benchmarks you have to beat. AM is what you reach for when the design can't escape the traditional process window without turning into six parts and a weld fixture.
What most people don't realize is that the cost conversation in metal AM isn't really about the machine. It's about everything around the machine.
The part nobody puts in the demo
Here's something vendors won't tell you up front: post-processing is where the budget goes. Support removal, heat treat, hot isostatic pressing, surface finishing, machining interfaces, and inspection—each one is a line item, and each one has its own queue time.
Velo3D's no-support capability does cut into that cost. It doesn't eliminate it. Most production parts still need some combination of the above. When you build your TCO model, assume the printer is maybe half the number. The rest is downstream capacity you either already have or have to build.
I keep coming back to that when I look at quotes. When I compared our outsourced complex-part spend against a rough print-and-finish estimate side by side—same parts, same quantities—I finally understood why engineering kept saying the machine 'pays for itself' while finance kept saying it didn't. They were measuring different halves of the same number.
Where the 2025 SpaceX contract chatter matters—and where it doesn't
You've probably seen the Velo3D–SpaceX relationship referenced in supplier discussions. Publicly reported coverage confirms Velo3D hardware is used in that supply chain. I don't have verified data on the specific contract volume or terms as of early 2025, and I'm not gonna pretend otherwise.
What I can say is this: for a procurement context, the relationship is useful as a qualification flag, not as a capacity claim. It means the technology can survive aerospace quality gates. It doesn't mean it fits your part mix, your volumes, or your operator skill level. Those are separate questions and they get answered with a print trial, not a press release.
When you shouldn't buy one
Straight list, from experience:
- You have fewer than a few hundred complex parts per year. Outsource. The overhead will eat the savings.
- You don't already have a quality engineer or the budget for one. Aerospace-qualified metal AM is not a 'load it and go' operation.
- Your parts are simple enough to machine. If a VMC can make it inside tolerance at a reasonable cycle time, the AM premium isn't justified.
- You're looking for a cheaper laser cutting replacement. It isn't one. Different process, different output.
- Your leadership can't tolerate a 12–24 month ramp. The hardware installs fast. The process capability doesn't.
Where I'd push back on my own take: if you're in a defense or space program with a locked design intent and a growing need for consolidation of assemblies, the calculus flips. In that scenario, the printer isn't a cost center—it's a design enabler. That's a strategic buy, not a procurement buy, and it should be justified that way.
Otherwise the honest recommendation hasn't changed. Quote the parts. Verify the supplier's post-processing chain before you sign anything. The machine's logo on the case matters less than the paperwork behind it.
Note: pricing, lead times, and program details reference publicly available information as of early 2025. Verify current specifications and contract terms directly with the vendor before making capital commitments.