Content Menu
● The Real Cost Structure of a CNC Job
● Fixed Costs versus Variable Costs in Practice
● Where the Volume Thresholds Actually Sit in 2025
● Real Examples from the Shop Floor
● Factors That Move the Thresholds
● Practical Rules Most Shops Use Today
● Frequently Asked Questions (FAQ)
Most shops live or die by the numbers on the quote sheet. When a customer asks for 20 pieces of a new pump housing in 17-4 stainless, the first question in every machinist’s head is the same: “Is this job going to make money, or are we just burning hours?” The honest answer almost always comes down to volume. Run ten pieces and the price per part can look painful. Run two hundred and suddenly the same job turns respectable. Run two thousand and it becomes the kind of work everyone wants on the floor.
That sharp drop in unit cost as quantity rises is the whole reason CNC machining still dominates so many industries in 2025. It is also the reason engineers spend half their time trying to figure out where the line sits — the exact quantity where a job stops being a prototype run and starts being real production. Cross that line too early and you eat the setup costs. Cross it too late and you leave money on the table or lose the order to injection molding or casting.
This article digs into the real numbers behind those decisions. We will walk through every cost element that actually shows up on a shop invoice, look at how fixed and variable costs behave, and then lay out the volume ranges that most experienced shops use as rules of thumb — and back them with data from published work instead of just shop-floor folklore.
A typical CNC quote is built from four buckets that never change, no matter who is doing the machining.
When the run is small, setup and programming dominate. When the run is large, machine time and material take over. The switch happens faster than most people expect.
Every shop keeps a rough spreadsheet that looks like this for a mid-complexity aluminum manifold, 150 mm × 100 mm × 80 mm, 16 pockets and drilled holes:
| Quantity | 10 pcs | 100 pcs | 500 pcs | 2000 pcs |
|---|---|---|---|---|
| Programming + fixtures | $2,800 | $2,800 | $2,800 | $2,800 |
| Material per part | $42 | $38 | $35 | $32 |
| Machine time per part | $68 | $58 | $52 | $48 |
| Overhead @ 40 % | $44 | $39 | $35 | $32 |
| Total per part | $392 | $106 | $62 | $51 |
| Fixed cost per part | $280 | $28 | $5.60 | $1.40 |
At ten pieces the fixed portion is 71 % of the price. At five hundred pieces it has fallen to 9 %. That single shift explains why shops quote what they quote.
After looking at hundreds of real quotes and the studies listed below, three broad bands keep showing up.
Low volume: 1–100 pieces Setup still rules. Per-part price is usually 4–10× the raw material cost. This is prototype territory or spare-parts runs. Anything under roughly 70 pieces for aluminum or 40 pieces for stainless is rarely profitable unless the customer pays a clear premium.
Mid volume: 100–1,000 pieces The sweet spot for most job shops. Fixed costs have dropped below 20–25 % of the total and the shop can still change over to the next job without crying. For 7075 aluminum parts of average complexity, the unit price typically lands between $45 and $90 once you cross 150 pieces.
High volume: 1,000 pieces and up Material and cycle-time reductions take over. Bulk material discounts kick in, tools are bought in lots, and operators get very good at loading the fixture blindfolded. This is where CNC starts to compete head-to-head with die casting or cold forging on pure dollars — but only if the part design is frozen.
These numbers move, of course. Titanium pushes everything to the right (thresholds often start at 300–400 pieces). Brass or free-machining steel pulls them left (sometimes profitable at 50–80 pieces).
Example 1 – Hydraulic manifold in 6061-T6 Setup + programming: $3,200 Cycle time: 28 minutes on a 4-axis horizontal At 80 pieces the shop quoted $184 each. At 250 pieces the same job dropped to $96. The customer split the difference and ordered 180 — right at the edge where the shop still made margin.
Example 2 – Titanium 6Al-4V bone plate for orthopedic use Five-axis work, cycle time 42 minutes, $4,800 in fixtures and first-article inspection. The medical device company originally wanted 120 pieces. Quote came in at $620 each. They raised the order to 420 pieces and the price fell to $218. That single change turned a money-losing job into the shop’s best margin of the quarter.
Example 3 – Brass plumbing connector, 5 operations on a lathe + mill Setup only $1,100 because the shop already had the collets and live tooling. Material $4.80 per piece. The threshold was under 60 pieces — anything above that and the job printed money. They run 8,000 a year now and the per-piece cost is $6.20 all-in.

Volume thresholds are not mysterious — they are simple arithmetic once every cost element is on the table. The numbers shift with material, complexity, and shop rates, but the shape of the curve never does. Fixed costs fall as 1/N, variable costs stay flat, and the unit price follows a path that drops fast at first and then flattens.
In daily work this means two things. First, never accept a low-volume job without understanding exactly where the fixed costs land. Second, always show the customer what the price would be at the next logical volume break — 100, 250, 500, 1,000 — because the savings are usually large enough to change the order size.
Master those two habits and the question stops being “Can we make this part?” and becomes “How many should we make to make everyone happy?”
Q1: At what quantity does CNC usually beat 3D printing on cost for metal parts?
A: For 316L stainless or 7075 aluminum, the crossover is typically 60–120 pieces depending on part size.
Q2: How much does stainless steel push the threshold compared to aluminum?
A: Roughly 1.8–2.2× higher volume is needed to reach the same unit price because of slower speeds and faster tool wear.
Q3: Is there a quick way to estimate the break-even point without full CAM programming?
A: Yes — take total expected setup + programming hours × shop rate, divide by (target price – material cost). That gives the minimum quantity to break even on fixed costs.
Q4: When does injection molding finally kill CNC for plastic parts?
A: Usually around 8,000–15,000 pieces for commodity resins once mold cost is amortized, sometimes earlier for high-value engineering polymers.
Q5: How much effect do modern trochoidal toolpaths have on the threshold?
A: They routinely cut cycle time 25–40 %, which can pull the mid-volume threshold down by 80–150 pieces on pocket-heavy parts.