This recommendation is a guidance heuristic based on typical process economics — your specific part, material grade and supplier can shift the answer. Use it to narrow the field, then confirm with an mfgiq engineer.
Choosing a manufacturing process is the single decision that locks in five things at once: how much you spend on tooling up front, what each part costs in production, your lead time, the tolerance you can hold, and the geometry you can physically make. Pick the wrong one and you either pay for a mould you'll never amortise, or you machine a part one chip at a time that should have been cast in seconds.
The dominant driver is almost always annual volume, with geometry a close second. Low-volume work — prototypes, pilot runs, spares — favours processes with little or no tooling, where you pay per part but never sink money into a die. That's why our Process Selector ranks CNC machining and 3D printing first when you tell it you need a handful of parts. As volume climbs into the thousands and tens of thousands, the maths flips: a high tooling cost spread across many parts pushes the per-part cost down dramatically, so injection moulding, die casting and investment casting take over. Sheet metal sits comfortably across a wide volume band for flat, bent and enclosure-style metal parts.
Geometry then filters the shortlist. Tall internal channels and organic shapes lean toward 3D printing or investment casting; flat brackets and enclosures suit sheet metal; solid prismatic parts with tight tolerances suit CNC. Material and tolerance act as gates — injection moulding is plastic-only, die casting and sheet metal are metal-only, and very tight tolerances rule out most printing and as-cast surfaces without secondary machining. The tool scores all of these together so you see a ranked field rather than a single guess.
You need five aluminium brackets to validate a fixture before committing to a design. At a quantity of five, any process with tooling is a non-starter — a mould or die would cost far more than the parts and add weeks of lead time for amortisation you'll never see. CNC machining wins because it needs no tooling: you cut directly from billet, hold a tight standard tolerance, and have parts in hand in days. Per-part cost is high relative to mass production, but with only five parts that's irrelevant. In the Process Selector, setting quantity to 1–10, material to Metal and tolerance to Standard ranks CNC first, with 3D printing close behind for even faster, looser prototypes. If you later scale this part, re-run the tool — the answer will change.
You're shipping a consumer product and need 50,000 plastic housings a year with a consistent cosmetic finish. Here the logic inverts: the steel mould is a real upfront investment, but spread across 50,000 parts that tooling cost amortises down to cents per part, and cycle times of seconds make the volume achievable. CNC machining each housing would cost many times more per part and never catch up. Injection moulding is the default for high-volume plastic parts with moderate-to-complex geometry, which is exactly why the tool ranks it first when you select 10k+ quantity and Plastic. The trade-off is the upfront tooling spend and the lead time to cut the mould — only worth it because the volume is there to pay it back. For the borderline cases, compare the crossover directly on CNC vs Injection Molding.
You need a part with conformal cooling channels or an organic, undercut geometry that a cutting tool simply can't reach — and only 20 to 50 of them. Volume is too low to justify a die, and the geometry is too complex for CNC to reach every feature without exotic fixturing. For plastic or where as-printed properties are acceptable, metal or polymer 3D printing wins — it builds internal geometry no other process can, with zero tooling. For a structural metal part needing better material properties and surface finish, investment casting is the stronger pick: it reproduces complex shapes from a printed or low-cost pattern at modest quantities, sitting between printing and high-volume die casting. The Process Selector flags 3D printing as a great fit for Complex 3D, low-volume parts; treat investment casting as the metal-properties upgrade path from there.
| Process | Best volume | Tooling cost | Typical tolerance | Geometry strength | Lead time |
|---|---|---|---|---|---|
| CNC machining | 1 – ~1,000 (prototype to mid) | None to low (fixturing only) | ±0.025 – ±0.05 mm (tight) | Solid prismatic parts, metals and plastics | Days |
| Injection moulding | 10,000+ (high volume) | High (steel/aluminium mould) | ±0.05 – ±0.1 mm | Moderate-to-complex plastic parts, thin walls | Weeks for tooling, then fast |
| Die casting | 10,000+ (high volume) | High (hardened die) | ±0.1 – ±0.2 mm | Complex metal shapes (Al, Zn, Mg) | Weeks for tooling, then fast |
| Sheet metal | ~10 – 10,000+ (wide range) | Low to moderate (tooling/dies) | ±0.1 – ±0.2 mm | Flat, bent, enclosure-style metal | Days to weeks |
| 3D printing | 1 – ~100 (one-off to low) | None | ±0.1 – ±0.3 mm (looser) | Complex / internal geometry, any shape | Hours to days |
| Investment casting | ~100 – 10,000 (low-to-mid) | Low to moderate (patterns/wax) | ±0.1 – ±0.25 mm | Complex metal shapes, fine detail | Weeks |
Tolerances shown are achievable typical ranges before secondary machining; cast and printed parts often add a finishing pass to hit tighter specs. Volume bands are guidance — the crossover point where tooling pays off depends on your part's size and complexity.
The same volume-and-geometry logic applies across hardware-making industries, so the tool fits a broad range of teams. Consumer electronics and hardware startups lean on it to decide when a CNC'd prototype should graduate to an injection-moulded production part. Automotive and industrial-equipment teams use it to weigh die casting against machining for metal components at scale. Medical, aerospace and robotics engineers need it for the tolerance and complex-geometry trade-offs that decide between CNC, investment casting and metal 3D printing. Within those companies, design and manufacturing engineers use it for early DFM direction, product and operations leads use it to sanity-check make decisions and timelines, and procurement uses it to frame supplier conversations before requesting quotes. Wherever a part has to be made in a real volume, the selector gives a defensible starting recommendation.
This selector is a transparent heuristic built on the everyday process economics our engineers work with — it isn't a black box and it isn't a sales funnel. It scores each process against your volume, material, geometry and tolerance, then ranks them so you can see not just the winner but why it won and what came close. The logic is the same one a manufacturing engineer applies on a first pass, made repeatable. Crucially, it's a starting recommendation, not a final answer: it can't see your exact feature sizes, draft angles, material grade or supplier's capacity, all of which can move the result. Treat the top-ranked process as the field to confirm, then validate the design-for-manufacturing details with a real manufacturer before committing tooling.
There's no single number — it depends on part size and complexity — but the principle is the break-even volume where amortised tooling cost drops below the per-part cost of machining. For many small plastic parts that crossover falls somewhere in the low thousands per year; for simple parts it can be higher, for large complex ones lower. The Process Selector ranks moulding and casting ahead of CNC once you select high-volume bands, which is the practical signal to model the crossover and confirm it with a quote.
Because lowest per-part cost only matters if the total cost — tooling plus per-part across your whole run — is lowest, and if the process can actually make your geometry to the required tolerance. A mould may give the cheapest unit cost yet be the worst choice for 50 parts once tooling is included. The selector ranks total fit across volume, geometry, tolerance and material, not unit price alone, so its top pick reflects the real economics for your quantity.
Often, yes. A cast or moulded part frequently gets a secondary CNC operation to hit tight tolerances on critical features, and 3D-printed patterns feed investment casting. The selector recommends the primary forming process; treat tight-tolerance faces, threads and mating surfaces as candidates for a finishing pass, and raise them in a DFM review so the full route is costed before you commit.