How Much Does CNC Machining Cost? A Complete Pricing Guide

Key Takeaways

• CNC machining costs $15–200+ per part at prototype quantities, depending on material, complexity, and tolerances. Simple aluminum brackets start around $15–30; complex 5-axis titanium aerospace parts can exceed $500.
• The six cost drivers, ranked by impact: cycle time (30–50% of total) → setup time (10–25%) → material (10–20%) → post-processing (5–15%) → inspection (5–15%) → programming (5–10%). Optimize in that order.
• Hourly rates range from $15–35/hr (China) to $75–150/hr (US). But the hourly rate alone is misleading — a Chinese shop running 2 shifts may quote a lower rate but a US shop with faster machines and better tooling may win on cycle time for complex parts.
• Quantity is the biggest cost lever. A part that costs $85 at qty 1 drops to $22 at qty 100 and $8 at qty 1,000. Setup and programming are amortized; material and cycle time dominate at volume.
• For US buyers, landed cost = part price + tariff (28–78%) + freight. A $50 aluminum part from China lands at ~$95 after tariff and shipping — still often 30–40% below US domestic. Non-metal parts (28% tariff) are the strongest value proposition.

Introduction: Why “How Much?” Is the Wrong First Question

Every day, machine shops receive RFQs that consist of a STEP file and one line: “How much?” The answer depends on at least six variables — and the difference between an optimized design and an unoptimized one can be 3× on the same part. This guide breaks down exactly what drives CNC machining costs, gives you realistic price ranges for common part types, and shows you how to cut your quote by 20–40% before you even send the RFQ.

We write this as a custom CNC shop that quotes parts every day. The numbers here are real, not theoretical. They come from actual production data across aluminum, steel, stainless, and plastic parts in quantities from 1 to 10,000.

Note: These ranges are for parts machined in China. US domestic prices are typically 2–3× higher at prototype quantities and 1.5–2× at production quantities. Add tariff for landed cost comparison (see section below).

Cost Driver #1: Cycle Time (30–50% of Total)

Cycle time — the minutes the machine spends actually cutting — is the single largest cost component. It is driven by material machinability and feature complexity:

Practical impact: Switching from 316L stainless to 6061 aluminum cuts the machining portion of your cost by roughly 60%. If your application can tolerate aluminum (with anodizing for corrosion), the material + cycle time savings compound dramatically. If you need stainless, 303 (free-machining) costs ~20% less to machine than 304 — ask your shop if it meets the corrosion requirement.

Best for low cycle time: 6061-T6 aluminum, brass C360, Delrin — materials that cut fast and are forgiving on tool wear.
Avoid when: Specifying titanium for non-aerospace parts — 5× slower than aluminum and 4× more expensive raw material. Unless weight or biocompatibility is non-negotiable, find an alternative.

Cost Driver #2: Setup Time (10–25% of Total)

Every time the part is removed from the machine and re-fixtured to machine a new face, you pay for setup. A part machined in one setup (all features accessible from one direction) costs significantly less than the same part requiring four setups. This is why 5-axis CNC machining — despite a higher hourly rate — can be cheaper for complex parts: it machines multiple faces in one setup, eliminating 2–3 re-fixturing operations.

Setup cost by process:

• 1 setup: Baseline cost. The part is loaded once. Single-op turned parts and simple 3-axis prismatic parts fall here.
• 2–3 setups: +40–80% over baseline. Each additional setup requires a new fixture or soft jaws, tool loading, and work offset setting — typically 30–60 minutes of skilled labor.
• 4+ setups: +100–200% over baseline. At this point, ask your shop: “Can we do this on a 5-axis machine in fewer setups?” The higher machine rate may be offset by setup elimination.

Best for low setup cost: Design parts so that all critical features are accessible from one or two directions. For multi-sided parts, consider 5-axis to reduce setups.
Avoid when: Your design has tight-tolerance features on 6 faces of a cube — each face needs its own setup, and the tolerance stack-up between setups multiplies the scrap risk. Redesign or accept the cost.

Cost Driver #3: Material (10–20% of Total)

Raw material cost varies dramatically — from ~$4/kg for mild steel to ~$80–120/kg for PEEK. But material cost is only part of the story. Material choice also affects cycle time (see above), tool wear (stainless eats end mills 3× faster than aluminum), and — for US importers — the tariff rate:

The tariff asymmetry is real: A $50 aluminum part lands at ~$95 (part + $39 tariff + $6 freight). A $50 Delrin part lands at ~$70 (part + $14 tariff + $6 freight). If Delrin meets your functional requirements, the 50% landed cost difference is impossible to ignore.

Cost Driver #4: Tolerances — The Precision Tax

Every dimension with a ±0.01 mm tolerance costs more than a ±0.1 mm tolerance. The cost curve is not linear — it jumps at specific thresholds:

The tolerance rule of thumb: For every tight-tolerance dimension on your drawing, ask: “What happens if this is ±0.1 mm instead?” If the answer is “it still works,” delete the tight tolerance. One unnecessary ±0.01 mm callout can add $10–30 per part. Ten of them can double your quote.

Cost Driver #5: Quantity — The Learning Curve and the Setup Amortization

Quantity is the most powerful cost lever in CNC machining — and the relationship is not linear:

• 1–10 parts (prototype): You pay full setup + programming cost. These fixed costs can be 40–60% of the total. A $200 programming fee on 5 parts = $40/part in programming alone. On 500 parts = $0.40/part.
• 50–200 parts (small production): Setup and programming are largely amortized. Cycle time and material dominate. Unit cost drops 60–80% from prototype pricing.
• 500–5,000 parts (production): Fixturing is optimized (dedicated soft jaws, multi-part fixtures). Cycle time is tuned. Material is purchased at volume pricing. Unit cost drops another 30–50%.
• 5,000+ parts (high volume): You approach the material + machine time floor. Further savings come from design optimization (reducing cycle time) and material sourcing — not from amortizing fixed costs, which are already near zero per part.

Best for quantity savings: Order 100+ parts. The setup/programming amortization cliff is between 50–100 parts — below that, fixed costs dominate. Above that, you are paying for material and machine time.
Avoid when: Ordering 5,000 identical parts from a shop optimized for prototypes. A production shop with multi-part fixtures and automated loading will quote 30–50% less than a prototype shop — even at the same hourly rate.

Cost Driver #6: Surface Finishes and Post-Processing

Every process that happens after the part comes off the machine adds cost. Some are essential (deburring). Some are cosmetic. Some are both. Know the difference:

The US Buyer’s Landed Cost Equation

If you are sourcing CNC machining from China, the quoted part price is only the starting point. Your actual cost is:

Landed Cost = (Part Price × [1 + Tariff Rate]) + Freight + Customs Broker Fee

Worked example for a $50 aluminum bracket, qty 100, shipped from China by air:

Now compare with US domestic quoting: the same bracket from a US shop typically costs $85–150/part. The China landed cost of $95.50 is at the low end of that range. For a 100-part order, the savings are $0–5,500 — real but not transformative. The tariff math narrows the gap considerably on metal parts. For a detailed comparison of custom shops vs platforms, see our custom vs standard CNC machining analysis.

But switch to Delrin (plastic): China quote is similar (~$45, plastic machines even faster than aluminum). Tariff drops to 28% ($12.60). Landed cost = ~$64/part. US domestic for Delrin CNC: $75–120/part. Now the savings are $1,100–5,600 on 100 parts — and the gap widens with quantity.

Best for landed cost: Non-metal CNC parts (Delrin, PEEK, nylon) — 28% tariff. Aluminum and steel parts are still competitive but with a narrower margin after 78% tariff.
Avoid when: High-value aluminum parts where the tariff alone ($39 on $50) erases most of the China cost advantage. Compare with US domestic or Mexico (USMCA duty-free) alternatives.

How to Cut Your Quote Before You Send the RFQ

Five DFM optimizations that reduce CNC machining cost by 20–40% — applied before the shop ever sees your drawing:

1. Relax non-critical tolerances. Every ±0.01 mm on your drawing is a cost multiplier. Go through dimension by dimension: which ones actually need to be tight? Change the rest to ±0.1 mm or leave them to the general tolerance standard (ISO 2768-mK).
2. Add internal corner radii ≥ 1 mm. Sharp inside corners require tiny end mills that break easily and take many passes. A 1 mm radius allows a standard Ø2 mm end mill to cut the corner in one pass. Cycle time drops 20–40% on pocket-heavy parts.
3. Reduce setups by orienting features to one or two faces. If you have holes on 5 faces of a part, ask: can any of these be moved to a face that already has features? Each eliminated setup saves 30–60 minutes of labor.
4. Choose free-machining material grades. 303 stainless instead of 304. 6061-T6 instead of 7075. C360 brass instead of C260. The machinability difference is 20–50% faster cycle time with longer tool life.
5. Avoid deep pockets with small corner radii. A pocket 50 mm deep with R1 mm corners requires a long-reach Ø2 mm end mill that deflects, chatters, and must move slowly. Widen the corner radius to R3 mm (Ø6 mm tool) or reduce the pocket depth. Cycle time can drop by 50% or more.

Conclusion: The Cheapest Part Is the One Designed for the Process

CNC machining cost is not mysterious. It is the sum of material, machine time, setup labor, and post-processing — all driven by design decisions made before the RFQ is sent. A part optimized for manufacturability costs half what an unoptimized one costs, even from the same shop on the same machine. The best time to reduce cost is not during supplier negotiation. It is during design. Run through the five DFM optimizations above. Then send the RFQ. The quote that comes back will reflect the engineering you put in — not the engineering you left out.