A decade ago, Fahz was a novelty with a production problem: a vase whose silhouette traced the profile of a real person's face. People liked it enough to order more than a thousand, and the maker eventually had to stop taking orders. Now the project is back on Kickstarter with a very different pipeline behind it. As 3D Printing Industry reports, Fahz 2.0 generates each vase mesh automatically on a web server from side-profile photos, with no manual CAD step. It then sends the production file by API to the print farm Slant 3D, which prints the vase in PLA and ships it to the customer.
The campaign runs from October 7 to November 2, 2026, with a $10,000 funding target. Rewards start at $30 for a 3-inch vase and $55 for a 5-inch version. Each vase can hold up to six faces.
From Two Printers to an API Call
The interesting part is less the vase than how the work has moved. According to 3D Printing Industry, the original 2015 campaign produced more than 1,000 vases on just two printers, and orders stopped in 2018. The project's own site, fahzface.com, tells the same story from the maker's side: a decade ago Fahz "grew faster than two people and their printers could handle," and after a thousand vases the makers "had to stop taking orders."
Anyone who has run a small print business can guess why. Personalized objects don't scale like a catalog item. Every order is a new part, which means a new file, a new slice, a new print job, and a new chance for something to go wrong. With two machines and a design step that needs a person, demand doesn't have to grow much before the queue outruns the workshop. In the original project, according to 3D Printing Industry, each customer's facial profile still had to be traced by hand in Photoshop, and the maker handled packaging and shipping himself.
Fahz 2.0 attacks both bottlenecks. On the design side, the customer does the input work. The project site says users upload a few side profiles and watch the vase take shape before it is printed and shipped. The geometry is generated on the server, not drawn by hand. On the production side, the finished file goes to Slant 3D programmatically, through an API rather than an email or a manual upload. Slant 3D prints in black or white PLA and handles shipping. 3D Printing Industry puts delivery at roughly two weeks.
That leaves the originator with a different job. Instead of running printers, the designer now maintains a piece of software that turns photos into printable parts reliably enough that nobody needs to check each one before it reaches a print farm. The coverage quotes Slant 3D CEO Gabe Bentz saying the farm's manufacturing capacity lets creators focus on product development rather than logistics, and the designer, Nicholas Desbiens, describes the project as a test of whether a one-person mass-customization business can run without repetitive manual work.
Designing for a Machine That Nobody Watches
Two details in the coverage show what it takes to remove the human from the loop. First, the generated mesh is designed to print on FDM machines without supports. For a farm printing one-off customer parts, that matters a great deal. Supports add material and print time, they need post-processing labor, and they leave marks on surfaces that, in this product, are literally someone's face. A design that prints cleanly without supports can go from the API to the build plate to the shipping box with little handling in between.
That constraint also shapes the generator itself. A face profile is full of features that can create problem overhangs: the underside of a nose, a chin, a brow. Software that produces support-free geometry from arbitrary photos has to stay inside the printer's overhang limits no matter what the input looks like. The coverage doesn't detail how Fahz handles this. But anyone who has modeled lithophanes, vase-mode prints, or other "upload a photo" products knows that the edge cases, not the typical inputs, decide whether an automated pipeline holds up.
Second, each vase gets an engraved serial number. That reads like a collector's touch, and it may be one. In a production setting, though, a unique identifier built into the part also helps match a physical object to a digital order, which counts for a lot once printing, packing, and shipping happen somewhere the designer never sees.
The Campaign and the Storefront
The Kickstarter is not meant to be the only way to buy. The project site says a store will open in November for buyers who don't back the campaign. The crowdfunding run works partly as a launch event and partly as a load test for the pipeline. The $10,000 goal is modest, which fits a product with no printer fleet of its own to buy: production is outsourced to the print farm.
Pricing reflects that. At $30 for a 3-inch vase and $55 for a 5-inch one, Fahz is priced as a personalized gift, not a commodity print. The margin has to cover the print farm's fee, shipping, and the platform's cut, plus the software that makes it all run without a person in the loop.
What It Means for Makers
For most makers, the takeaway isn't the vase. It's the architecture. Fahz 2.0 is a clear, small-scale example of a model that print farms offer: the designer owns the product and the customer relationship, the farm owns the machines, and an API connects the two. The designer never touches a printer, and the farm never touches a CAD file by hand.
That model lowers the cost of entry for parametric and personalized products considerably. A maker who can write software that reliably emits printable geometry no longer needs to buy a fleet, store filament, or run a fulfillment room to sell at volume. The original Fahz hit its ceiling on two printers. This version, at least on paper, has no fixed ceiling on the production side.
The trade-offs are just as real. Material and color choices narrow to what the farm offers, which here means black or white PLA. Quality control becomes a software problem: the generator has to be right every time, because there is no workshop step where a person catches a bad mesh before it prints. And the business depends on a third party's pricing, capacity, and turnaround, with the roughly two-week delivery window set as much by the farm as by the designer.
For makers sketching out their own personalized product, Fahz offers a useful checklist. Design for support-free printing from the start. Make the generator robust to messy real-world inputs. Build traceability into the part. Treat the print farm as infrastructure rather than a vendor you email. Whether Fahz 2.0 reaches its goal will say something about demand for face vases. How smoothly its orders flow from photo to doorstep will say more about whether this kind of hands-off production is ready for small makers.
