We keep having the wrong conversation about Velo3D.
Everyone wants to talk about the SpaceX contract, the Anduril work, the 2025 roadmap. Is Velo3D a viable supplier? Will they survive the consolidation? Is the technology ready for real production?
After 5 years of managing procurement for a mid-sized advanced manufacturing group, I've come to believe we're missing the point. The real story isn't about who prints rocket parts. It's about how the industry's perception of additive manufacturing is finally catching up to reality.
And that changes what we should be asking our own vendors—regardless of whether we're buying a $2M Sapphire system or a $200k batch of machined brackets.
What Was Best Practice in 2020 May Not Apply in 2025
I remember the pitch decks from 2021. „Metal 3D printing will replace CNC.“ „No supports needed for everything.“ „Complexity is free.“
That was nonsense. And the people who believed it got burned.
But here's the thing: the industry has changed more in the last 18 months than in the previous five years. And Velo3D is a useful lens for that shift—if you look at the right signals.
When I see the news about Velo3D's SpaceX relationship (they've been a qualified supplier for years), I don't see a hype story. I see a company that survived the valley of death by focusing on a narrow, brutally difficult niche: critical, complex metal parts for customers who cannot afford failure.
That's not a commodity play. That's a specialty supplier play. And it's a model more procurement teams should understand.
The Argument We're Stuck In
Most industry commentary frames Velo3D as either:
- A revolutionary pioneer unfairly dragged down by market conditions, or
- An overhyped also-ran that couldn't compete with EOS or SLM Solutions.
Both narratives are lazy. And both miss the operational reality that actually matters for buyers like us.
Here's what I actually track when evaluating a supplier like Velo3D—or any critical-component vendor, for that matter:
- Qualification cycles. How long does it take to get a part approved? Months? Years?
- Process consistency. If I order the same part twice, will I get the same metallurgy? (Spoiler: this is where metal AM still struggles.)
- Cost per good part. Not cost per build. Cost per good part. That changes everything.
- Who's actually using it in production—not just in R&D.
On those criteria, Velo3D's work with SpaceX and Anduril is genuinely significant. These aren't companies that tolerate defects. They're not running lab experiments. They're putting parts on flight vehicles and in defense systems.
That doesn't mean Velo3D is right for everyone. It doesn't mean the technology is mature enough for high-volume automotive. But it means the „can't be trusted for real applications“ argument is stale.
And that's the evolution I've seen: metal AM is no longer a prototyping curiosity or a low-volume bridge. It's a legitimate production technology for a specific class of problems.
The Nuance The Headlines Miss
Here's the part that frustrates me as a buyer. The same people who mocked Velo3D for „overpromising“ now overcorrect in the other direction—dismissing the entire additive sector because of one company's stock price or a delayed earnings report.
Neither extreme is useful for making purchasing decisions.
I learned this the hard way. A few years ago, we evaluated a metal AM system for a prototype-to-production program. The vendor pitched „complexity for free.“ We tested it with our own parts. We found that design freedom is real, but cost-per-part is still highly variable—especially when you factor in support removal, post-processing, and qualification testing.
The issue wasn't the technology. It was our expectation that it would behave like traditional manufacturing. It doesn't. Yet.
Velo3D's specific bet—that you can eliminate supports for many geometries—is compelling if it holds up in production. The fact that SpaceX and Anduril keep coming back suggests it does, at least for their use cases. As of my last review in Q4 2024, their publicly announced customers and contract wins pointed in that direction. Things may have evolved since then—defense supply chains change fast—so verify current partnerships before making any strategic decision.
Red Light Therapy After CO2 Laser? The Question Nobody's Asking
Now, here's where I could digress into laser cutting or ball mill components. But I won't, because the meta-issue is the same.
When I search for information about tempered glass laser cutting machines or supply ball mill feeding end caps, I get the same pattern: people asking „can I do X with Y?“ and getting answers from vendors who have never run a production floor.
Can you use red light therapy after a CO2 laser? Depends on the context. Are you talking about skin treatment or material processing? The question itself reveals a fundamental misunderstanding—which is fine if you're new, dangerous if you're making procurement decisions.
The same logic applies to Velo3D. „Can their printer make complex parts?“ Yes. „Can it do everything, cheaply, overnight?“ No. The answer is it depends—on your qualification needs, your volume, your acceptance criteria.
Anyone who gives you a one-word answer doesn't understand the problem.
What I'd Ask A Velo3D (Or Any AM Supplier) Before Buying
If you're actually considering a system, don't start with „is it better than EOS?“ Start with operational questions:
- What is your process capability index (Cpk) for critical dimensions? If they don't know, walk away.
- Can you show me parts that have been in service for 10,000+ hours? Not test coupons. Real parts.
- What's your support removal rate for complex internal channels? This is where the „no supports“ claim gets tested.
- How many of your installed systems are running 24/7 production shifts? Not demo units. Production.
- What happens when a build fails at hour 80? Who absorbs the cost? How predictable is your rework?
I've asked these questions to at least five AM vendors over the past two years, including some very well-known names. The answers are... inconsistent. Which tells me the industry is still maturing.
The Industry Has Evolved. The Conversation Hasn't.
Here's my point, and I'll state it clearly: We are still debating whether metal AM is „real manufacturing“ when the evidence is already in.
SpaceX doesn't use Velo3D to print table lamps. Anduril doesn't rely on unproven tech for defense systems. These companies have done the risk assessment, and they've decided the risk of falling behind is greater than the risk of adopting AM.
That doesn't make Velo3D a safe stock pick or a one-size-fits-all solution. It makes them a legitimate option for a narrow, high-stakes slice of the market.
The fundamentals of procurement haven't changed: quality, delivery, cost, service. But the context has transformed. Five years ago, „metal 3D printing for flight hardware“ was a fantasy. Today, it's a qualification pathway.
If your mental model of Velo3D or metal AM is still stuck in 2020, you're not being skeptical. You're being left behind.
As for whether Velo3D is a „good“ supplier? That's the wrong question. The right question is: Is Velo3D a good supplier for your specific requirements?
For critical, low-volume, complex metal parts where failure is non-negotiable: they're one of the few genuinely proven players.
For anything else? Keep looking. And that's fine. It's called procurement, not magic.
Simple.