10 foreign 3D printing startups worth paying attention to in 2023

Most of the well-funded foreign 3D printing startups in 2023 sit in metal powder bed fusion. The pitch is consistent: internal channels, lattice structures, and consolidated assemblies that would otherwise need brazing or welding. For a hydraulic manifold or a heat exchanger core, that is a real gain. The part comes out as one piece, and you skip a stack of joints.

The limits show up early. Production metal systems commonly build within a 250 to 400 mm envelope, so a 4,000 mm frame is out of reach. Surface finish off the machine usually lands around Ra 8 to 15 μm, which means any sealing face or bearing bore still needs machining. Support removal inside narrow channels is manual work, and it is easy to miss a strut.

Material choice is narrower than the datasheets suggest. Ti-6Al-4V and 316L print well and are qualified in many shops. Copper alloys and some aluminium grades are harder because of laser reflectivity. If your drawing calls for 7075 or 17-4PH in a specific temper, check whether the printed version reaches the same mechanical properties before you commit.

The right way to use these startups is as a geometry test. Print one manifold, pressure test it, and confirm the channel layout works. Then move to machining when the design freezes.

The polymer side of the list splits into two groups. Material extrusion startups target fixtures, jigs, and low-load brackets. Resin and powder-based startups target visual models, housings, and small production runs of non-structural parts. Both are useful, and both have tolerances that engineers tend to overestimate.

A desktop extrusion machine might hold ±0.3 mm on a good day, and less on a tall thin wall. Resin systems do better on fine features but shrink during cure, so a 100 mm part can move 0.2 to 0.5 mm depending on geometry. If your assembly has a press fit, do not print it and expect it to drop in. Print it, measure it, and machine the mating part to match.

For enclosures and covers, printing is often the fastest route to a fit check. For anything that carries load, sees heat above 80 °C, or needs a thread that will be tightened more than a few times, a printed polymer part is usually the wrong call.

One practical note: printed parts are anisotropic. A bracket loaded across the layer lines can fail at a fraction of the datasheet strength. Orient the part so the load runs along the layers, or accept a lower working load.

A few startups on the list use bound metal or binder jetting. Metal powder is held in a polymer binder, printed, then debound and sintered. The appeal is lower capital cost than laser powder bed, and the ability to run larger batches. The catch is shrinkage. Sintering pulls the part down by roughly 15 to 20 percent, and the compensation is handled in software, not by the operator.

That shrinkage makes tight tolerances hard. A sintered part might hold ±0.5 percent of dimension, which on a 100 mm feature is ±0.5 mm. You then machine the critical faces. This is a reasonable path for small complex parts in moderate volume, but it is not a replacement for a machining center on a tight-tolerance feature.

Debinding and sintering also take time. A batch cycle can run several days, so the process is not a same-week answer. If your project needs parts in three to five days, printing these geometries is not the route.

Where it does earn its place: small manifolds, impellers, and brackets where the printed shape is close enough and only two or three faces need finishing.

The clearest signal is a tolerance callout. If a drawing asks for ±0.005 mm, or a surface finish of Ra 0.8 μm, printing will not hold it without secondary machining. At that point you are paying for two processes, so it is worth checking whether machining the whole part is cheaper.

The second signal is size. A printed part that spans 800 mm is possible on some systems, but flatness and warp become a problem. A 5-axis machining center handles a 4,000 mm part in one setup, and the geometry is defined by the cutter, not by thermal history.

The third signal is material. If the part needs 7075-T6, 17-4PH in condition H900, or a specific temper, machining from certified stock is the shorter path. You get a mill certificate and a known heat treat, not a sintered approximation.

None of this makes printing a bad choice. It makes it a choice with a boundary. Draw the boundary at the tolerance, the size, and the material, and the decision usually makes itself.