CNC Machining vs 3D Printing vs Injection Molding: Which Process for Your Part?

Start with what the part has to do

The first mistake in process selection is to start from the machine. The part should dictate the route, not the other way round. Begin by writing down what the component actually has to achieve: the environment it lives in, the tolerances that matter, the material it must be made from, and how many you will make. Those four facts eliminate most options immediately, and the remaining choice is usually between two processes rather than three.

Volume is the loudest signal but not the only one. A bracket made ten times a year is a different animal from the same bracket made ten thousand times a month, and the right process flips between them. But geometry and material can override volume: a part with internal channels or a lattice that no cutter can reach will point at 3D printing regardless of quantity, and a part in a food-grade polymer with tight cosmetics will point at molding even at modest volume. The job is to let the part's own requirements do the elimination, then argue only about the last two candidates.

CNC machining: precise, quick to start, costly per part

CNC machining removes material from a solid block. Its great strength is that it starts almost immediately - there is no tool to design and cut first - and it holds tight tolerances on familiar, well-characterised metals and plastics. For a prototype, a fixture, a bracket or a low-volume production run, machining is usually the default because the setup cost is low and the lead time is short, and the part you receive is made from a solid with known material properties.

The weakness is the unit economics. Because every part is carved from stock, the per-part cost does not fall much as quantity rises, and large or complex parts can be expensive because they consume machine time and material that ends up as swarf. Machining also struggles with certain geometries - deep internal cavities, conformal cooling channels, and anything undercut that a standard cutter cannot reach - where the part either cannot be made or needs multiple setups that erode accuracy. For production volumes above a few thousand identical parts, machining usually loses to molding on price, and it was never the tool for a hollow, organic shape.

3D printing: freedom of shape, slow and weak at scale

Additive manufacturing builds a part layer by layer from a digital model, which removes almost every geometric constraint that machining and molding fight with. Internal channels, organic lattices, conformal shapes and one-off customisation are where 3D printing is unbeatable, and because there is no tool or cutter path to design, a new geometry is often only a file change away. That makes it the natural choice for prototypes, jigs, fixtures, and parts where the shape itself is the point rather than a by-product of the process.

The limits are throughput, material and consistency. Printed parts are slow to make one at a time, so the per-part time does not collapse with volume the way molding does; most printed materials are not yet the engineering grades you would specify for a load-bearing production component; and mechanical properties can be anisotropic, stronger in one build direction than another. For a few units, or a part whose shape cannot be made any other way, 3D printing wins outright. For ten thousand identical production parts, it usually does not, and the part it produces cannot stand in for the moulded part's isotropic behaviour.

Injection molding: cheap per part, expensive to begin

Injection molding forces molten plastic into a steel mould at high pressure, then ejects a cooled, solid part. The economics are the mirror image of machining: a high upfront cost - designing and cutting the mould, which can be a substantial sum and weeks of lead time - followed by a very low per-part cost once the tool is running. At volume, nothing else comes close for a plastic component, and the surface finish can be moulded in rather than applied later.

Molding also imposes its own rules. The part needs draft so it can eject from the tool, reasonably uniform wall thickness to avoid sink and warpage, and a tool that can be parted and, where necessary, carry side actions for undercuts. The geometry therefore has to be designed for molding from the start, which is why a moulded part and a machined part that look similar on a drawing can have very different designs underneath. A supplier that owns the whole chain - mould design, tool manufacture, molding and assembly - is worth finding, because the questions about draft, gating and cosmetic finish all land in one place. Manufacturers such as DAYIN, a plastic injection molding OEM/ODM with 30+ years and plants in both China and Vietnam, run that one-stop model, so the tooling and the production volume are planned together rather than handed between separate vendors.

A three-way comparison you can actually use

The table below is the short version of how the three routes split. It is a planning guide, not a verdict - the exact numbers for a specific part come from pricing the routes against the drawing.

Where the decision is really made: volume, geometry, material

Volume sets the broad band, but geometry and material decide within it. If the part has internal passages, a lattice, or a shape no mould can be parted from, 3D printing is the only route regardless of quantity. If the part is a tight-tolerance metal component, machining wins because it uses the real material and does not depend on a melt. If the part is a cosmetic plastic component needed by the thousand, molding wins on both unit cost and finish.

Consumer products are the clearest illustration. A robot vacuum cleaner housing, a pet feeder shell or a kitchen utensil is molded by the million because the unit cost and the surface finish both matter, and the geometry can be designed for the tool from the outset. A maker of such goods, for instance DAYIN's robot vacuum cleaner range, is effectively a molding operation first and a product company second - the economics only work because the tool amortises across volume, which is exactly the condition molding needs to be the right answer.

Lead time and iteration: the hidden axis

The brochure comparison ignores time, and time is often the deciding factor. Machining and printing both deliver a first part in days, which is why they dominate prototyping. Molding delivers nothing until the tool exists, which is weeks. The common pattern is to machine or print the prototype, validate the design in the field, and only then commit to a mould - so the molding decision is frequently deferred rather than made up front, and the deferral is usually correct.

Iteration cost also differs sharply. Changing a machined part is a new program; changing a printed part is a new file; changing a molded part after the tool is cut can mean re-cutting steel. That asymmetry is why the design should be settled before tooling, and why a molding partner who runs a mold trial or a prototype in parallel earns their fee many times over. A change caught before the tool steel is cut is an afternoon; the same change caught after is a rebuild.

A simple decision matrix

If the prose is too much, the matrix below settles most cases. Read down the scenario column until one fits, then read why.

Working with a molding partner

If molding is the answer, the early conversation matters more than the mould steel. A capable partner will review the design for moldability before cutting metal - flagging draft, wall thickness, gating and undercuts while changes are still free, which is the cheapest possible moment to find a problem. The scope to look for spans R&D design, mold manufacturing, injection molding, secondary processing and electronic assembly, so the tooling and the production volume are planned as one programme. DAYIN's OEM/ODM services page describes exactly that remit, with an ISO 9001-certified laboratory carrying the quality control step, which means the inspection plan travels with the production lot rather than being an afterthought. Arriving with a machined or printed prototype that has already passed field trials is a far stronger position than arriving with a concept, because the partner can quote a production-intent tool against a part that has been validated.

What to send is short: a 3D model, the material grade, the expected volume and forecast, the critical tolerances, and the finish the part needs. Volume and forecast decide whether a mould is justified at all, and the critical dimensions decide where the tooling and inspection effort should go. State the datums so the partner and the inspector are working from the same reference, and the quote that comes back will describe its assumptions rather than hiding them.

The bottom line

There is no universally right process, only a right process for a specific part at a specific volume. Machining and printing are the fast, low-commitment routes that suit prototypes and short runs; molding is the volume route whose savings only appear once a tool is paid for. Start from what the part has to do, let volume and geometry do the elimination, and treat the molding decision as one to be earned by volume rather than assumed. The expensive mistakes are not in picking the wrong machine - they are in committing to a mould for a design that was not ready, or machining a part by the thousand that should have been molded.

Frequently asked