CNC machining or 3D printing — when to machine, when to print?

The question "CNC machining or 3D printing" comes back with almost every new project — and increasingly it is asked not only by design engineers but also by buyers. Both technologies can make "the same" part, yet they differ in almost everything: material, accuracy, strength and how cost behaves across a batch. Choosing blindly ends either in overpaying or in a part that will not survive its loads.

Below is an honest comparison from the position of a machining company: we show where printing wins outright, where CNC has no competition, and how to combine both technologies in one project instead of picking a religion.

Where 3D printing wins

3D printing builds the part layer by layer, so tool-access restrictions do not apply to it. Its real advantages follow from this:

• geometries impossible to machine — internal channels with curved paths, lattice structures, closed chambers, parts consolidated from several components into one,
• fast prototypes of form and ergonomics — a concept print can be on your desk the next day, with no programming, fixturing or tooling,
• single small plastic parts — handles, covers, cable guides, mock-ups; the start-up cost is close to zero,
• personalisation — every piece can be different at no extra preparation cost,
• no waste from a solid bar — material is used only where the part is.

If the part is plastic, carries no significant loads and you need it quickly in 1-3 pieces — printing is usually the right answer and there is no point pretending a milling machine will do it cheaper.

Where CNC machining wins

Machining starts from solid, homogeneous mill stock — and that is the source of its advantages:

• engineering metals — C45 steel, stainless steels, aluminium alloys or bronzes are machined from certified stock with known, repeatable properties; we covered selection in how to choose the material,
• tolerances — milling and turning routinely hold hundredths of a millimetre, and H7/g6 fits are everyday work; printing operates rather in tenths,
• surface quality — after machining Ra 0.8-1.6 is standard, after printing the layers are visible and finishing is usually needed,
• strength and stiffness — solid material has no anisotropy between layers; an FDM print can be several times weaker in the direction perpendicular to the layers,
• cost in a batch — after one-off preparation the machining cycle time is short, so the unit price drops quickly; in printing the build time of each piece is practically constant.

The mechanism of the price falling with batch size is described in detail in from prototype to production batch — it is what makes machining win on cost at a few dozen pieces, even against cheap plastic printing.

Metal printing — is it competition for machining yet?

Printing in metals (most often laser-melted powder-bed technologies) deserves its own paragraph, because it is what appears in the questions "is CNC still needed". The state of play today, greatly simplified:

• metal printing pays off for geometries that cannot be machined — cooling channels with curved paths, lightweight lattice structures, consolidation of many parts into one,
• the unit cost is high and hardly falls in a batch, so for typical machine parts machining remains clearly cheaper,
• functional surfaces of a metal print still require CNC finishing — fits, threads and bearing seats do not come out of the printer within tolerance,
• material properties depend on process parameters and the build orientation of the part, which in critical applications means additional testing and acceptance.

In short: metal printing is a tool for special tasks, not a replacement for machining in the everyday production of machine parts.

Decision table: six selection criteria

The values in the table are indicative and describe typical cases, not the extreme capabilities of either technology.

If the part lands "on the CNC side" in at least three of the function-related rows (material, tolerances, strength) — printing can serve at most as a prototype, not as the final part.

Beware the single-criterion trap: a part that prints "almost right" dimensionally often loses on strength or on the cost of finishing. Make the decision on the full set of criteria, with the part's function in first place.

Hybrid scenarios: print the prototype, machine the production part

In practice the best projects do not choose one technology but a sequence. Three typical scenarios:

1. Form prototype → machined production. You print the device housing in 2 days, check assembly and ergonomics, apply corrections in CAD, and once the design is frozen you order the aluminium version from the milling machine. The print costs a fraction of a machined prototype, and you catch design errors before spending money on machining.
2. Prints as tooling. Printed grippers, gauges and assembly pads support the production of machined parts — where the tooling does not carry large forces, printing is faster and cheaper.
3. Metal printing + CNC finishing. A part with complex internal geometry is built additively, but the mating surfaces, threads and bearing bores are still finished by machining, because printing does not hold the required tolerances.

A fourth, increasingly common variant is printing a spare part "to survive": the print lets the machine run for a few days while the proper part is machined from metal in the meantime. It is an emergency solution — the print runs at reduced parameters and requires the informed consent of maintenance.

The condition for a smooth print → CNC transition is a good model. Geometry designed "for the printer" (thin ribs, overhangs, pockets without tool access) can be expensive or impossible to machine — check this before freezing the design, ideally at the CAD/CAM design stage. The wider context of choosing a method is in how to choose a CNC machining technology.

How to decide in practice

Instead of starting from the technology, start from three questions about the part:

• What function does it perform? If it carries loads, works with bearings or requires fits — it is almost always machining.
• What must it be made of? A certified engineering metal practically settles it in favour of CNC; a plastic without strength requirements opens the door to printing.
• How many pieces and over what horizon? A single trial piece favours printing, repeat batches favour machining.

Only after these answers compare the prices. A common mistake is comparing the cost of a printed prototype with the cost of a machined prototype and drawing conclusions about the whole project — while in a batch the cost curves of the two technologies look completely different.

Also make sure the offers you compare are equivalent in scope: a printing offer often does not include finishing of the functional surfaces, threaded inserts or material testing, which sit in the price of a machining offer. Compare the cost of a part ready for assembly, not the cost of a raw shape.

Summary

CNC machining or 3D printing is not a choice of the "better" technology, but matching the tool to the project stage. Printing wins where iteration speed, complex geometry and plastics matter. CNC wins where metal, tolerances, surface, strength and cost in a batch matter. Most often the most sensible answer is: print first, then machine.

Have a part on the border of both worlds? Send the model or drawing with a description of its function and quantities via the contact form — we will assess machinability and send back a quote within 48 hours, and if printing is the better option at this stage, we will say so outright.