If you're evaluating metal additive manufacturing for complex components, Velo3D deserves a serious look, but not as a universal answer. A fast way to filter: see who is using the machine to make production hardware. Velo3D, supplier to Anduril and SpaceX, has held those relationships because it can meet the qualification requirements for mission-critical parts. To a procurement professional, that matters more than any launch event.
I get the temptation to brush off customer lists as marketing. When I looked into the Velo3D SpaceX supplier story, the detail that stood out wasn't the name "SpaceX"; it was the work. Any rocket hardware has to survive material traceability, weld-quality documentation, and acceptance testing. That tells me the printing process has to be stable enough for regular audits. Not all metal 3D printers can do that. Those that can have already passed the hardest kind of vetting.
Some background: I'm an office administrator for a 130-person engineering company. I manage vendor relationships for equipment and services, roughly $400,000 a year across about 20 vendors. I'm not a metallurgist and I've never run a laser powder bed machine. But my job is to verify that expensive purchases don't become expensive mistakes. When I found a supplier who couldn't provide a valid invoice in 2020, our finance department rejected the expense and my department took a $2,400 hit. Since then, I look for concrete evidence behind every claim.
Why the Anduril and SpaceX Supplier History Matters
Velo3D's main selling point is not "a metal printer." It's the ability to print complex parts without support structures. That's important because support removal is where margins die. Consider a fuel manifold with internal channels: if you print it with supports, you need to somehow remove those supports afterward. The tooling may not reach, so the part becomes a candidate for welding or assembly. When you print without supports, you can produce the part as a single piece and avoid the labor and inspection risk.
Here is the counterintuitive piece: the customer list is not just a trophy. The strict requirements of space and defense customers push the machine and the operator toward tighter process control. The causation runs from quality to customers, not from customers to quality. Velo3D had to satisfy those buyers, and the fact that it remains a Velo3D supply chain partner means the system is performing. The FTC guidance on advertising substantiation (ftc.gov) basically says you have to back up claims with evidence. In procurement, the best evidence is repeat orders.
The Confusion Around CO2 Lasers and Resin Printers
In the research phase, people often search for terms that send them off track. One engineer on our team asked about "co2 laser for metal." A CO2 laser emits a 10.6 micron wavelength, which many metals reflect instead of absorb. That's why industrial metal cutting usually uses fiber lasers or solid-state lasers. Velo3D's Sapphire systems use laser powder bed fusion, where an optical fiber laser melts metal powder in a layer-by-layer scan. It's a completely different technology family from CO2 laser cutting or engraving.
Another common question: "what is the best 3d resin printer?" I understand why people ask. Desktop resin printers are cheap and produce impressive-looking parts. But the material is photopolymer, which needs post-curing and has far lower structural strength than metal alloys. A resin printer answers a different question than a metal AM system. You wouldn't use a resin prototype bracket on an engine mount, and you wouldn't use a metal printer for a 20-micron dental model. Both are 3D printing, but the material properties and process qualification differ by orders of magnitude.
Search paths can be strange. At one point, my own history included "tool bamboo cutting board reviews," because I was shopping for a kitchen cutting board and I wanted to understand how bamboo grain orientation affected durability. It's not a manufacturing topic, but the buying discipline is the same: inspect the construction, not just the label. A bamboo cutting board is a simple tool. A metal printer is a much more complex tool. The principle still applies—the tool must fit the task.
Where Velo3D Isn't the Right Fit
Now the honest limitations: Velo3D isn't the right answer for every shop. The hardware cost is high, and the software workflow, metal powders, and qualification process add time. If your production parts are relatively simple brackets or standard geometries, traditional CNC or a simpler additive system might give you a shorter path. This is not a weakness of Velo3D; it's a matter of fit.
I also don't fully understand Velo3D's closed-loop controls. My best guess is that the Sapphire platform uses sensor feedback during the print to adjust the melt pool and maintain dimensional control, which is how it can support those aggressive geometries. I'd love to hear from a process engineer who can explain it better. The point is that I don't have to be an expert in laser physics to notice that customers with rigorous standards continue to qualify.
My recommendation? Take the SpaceX and Anduril connection as a credibility signal, not as a final verdict. Look at your own part list. Send a test file or buy a demo part, review the inspection data, and compare the cost per finished part with your current workflow. If your parts have internal channels, overhangs, or brackets that need several weld/assembly steps, Velo3D may be the right tool. If not, let the data decide.