Last month I did something I've been meaning to do for years. I took five real part drawings — ones we've actually machined for customers — and submitted them to Xometry for a quote. Then I pulled our own pricing and asked two other China direct manufacturers for their numbers on the same parts.
The results were... not what the Xometry marketing page would lead you to expect. But also not as simple as "China is cheaper."
Here's what I found, what surprised me, and a framework for deciding when the platform premium makes sense and when it doesn't.
The five test parts
I picked parts that reflect what actually lands on my desk, not cherry-picked geometries that make one model look good:
Part A — Simple aluminum bracket. 6061-T6, 80mm × 50mm × 12mm. Six holes, two counterbores, no tight tolerances. Qty 10. This is the kind of part any shop with a 3-axis VMC can make in their sleep.
Part B — Stainless manifold. 316L, 120mm × 60mm × 40mm. Internal channels, NPT threads, flatness 0.05mm on the sealing face. Qty 5. Requires 4-axis or 5-axis, careful workholding, and someone who knows how to machine 316 without work-hardening.
Part C — Aluminum enclosure with tight pockets. 6061-T6, 200mm × 120mm × 45mm. Thin walls (1.5mm), tight pocket depth tolerance (±0.02mm), large flat surface that'll warp if you look at it wrong. Qty 20. This is a fixturing and stress-relief challenge.
Part D — PEEK medical device component. Unfilled PEEK, 40mm × 30mm × 15mm. Multiple small features, Ra 0.8 surface finish on sealing surfaces, burr-free requirement. Qty 50. Not geometrically complex, but PEEK is expensive and finicky.
Part E — 7075 aluminum aerospace bracket. 7075-T6, 150mm × 100mm × 25mm. ±0.01mm on bearing bores, weight-reduction pockets, chem film finish. Qty 100. This is the kind of production run where volume should drive per-part cost down — if the shop is set up for it.
The quotes (actual numbers)
Here's what came back. All prices in USD, FOB. Xometry prices are from their instant quoting engine with "standard" lead time selected. China direct quotes are FOB Dongguan or Shenzhen.
Part A — Simple Aluminum Bracket (Qty 10)
| Source | Unit Price | Total | Lead Time |
|---|---|---|---|
| Xometry | $38.20 | $382 | 8 business days |
| YEYUAN CNC | $18.50 | $185 | 7 days |
| China Shop B | $16.00 | $160 | 10 days |
| China Shop C | $21.00 | $210 | 7 days |
Spread: $16.00 – $38.20 per part. On a simple bracket, Xometry is 2.1× the median China direct price. For $382 vs $185, the Xometry premium is about $197 — on one of the cheapest parts you'll ever order.
Part B — Stainless Manifold (Qty 5)
| Source | Unit Price | Total | Lead Time |
|---|---|---|---|
| Xometry | $385.00 | $1,925 | 12 business days |
| YEYUAN CNC | $182.00 | $910 | 12 days |
| China Shop B | $195.00 | $975 | 14 days |
| China Shop C | $210.00 | $1,050 | 10 days |
Spread: $182 – $385 per part. Xometry is 2.1× again — but note the lead times are comparable. The complexity premium scales similarly.
Part C — Aluminum Enclosure (Qty 20)
| Source | Unit Price | Total | Lead Time |
|---|---|---|---|
| Xometry | $127.00 | $2,540 | 10 business days |
| YEYUAN CNC | $62.00 | $1,240 | 12 days |
| China Shop B | $58.00 | $1,160 | 15 days |
| China Shop C | $70.00 | $1,400 | 10 days |
Spread: $58 – $127 per part. Xometry is 2.2×. On 20 pieces, the total difference is $1,160–$1,400 — roughly enough to buy a second batch.
Part D — PEEK Medical Component (Qty 50)
| Source | Unit Price | Total | Lead Time |
|---|---|---|---|
| Xometry | $89.50 | $4,475 | 12 business days |
| YEYUAN CNC | $52.00 | $2,600 | 14 days |
| China Shop B | $55.00 | $2,750 | 15 days |
| China Shop C | $60.00 | $3,000 | 12 days |
Spread: $52 – $89.50 per part. Xometry is 1.7×. The ratio narrows on PEEK — material cost is high ($20+/part in raw stock), so the machining cost difference is diluted. On 50 pieces, the savings from going direct is $1,475–$1,875.
Part E — 7075 Aerospace Bracket (Qty 100)
| Source | Unit Price | Total | Lead Time |
|---|---|---|---|
| Xometry | $94.00 | $9,400 | 15 business days |
| YEYUAN CNC | $46.00 | $4,600 | 18 days |
| China Shop B | $42.00 | $4,200 | 20 days |
| China Shop C | $50.00 | $5,000 | 15 days |
Spread: $42 – $94 per part. Xometry is 2.2× again. On 100 pieces, the difference is $4,800–$5,200. That's real money — enough for tooling, finish, certification charges, or a business-class ticket to Dongguan to visit the factory in person.
Five patterns in the data
1. The Xometry multiplier is surprisingly consistent. Across aluminum, stainless, and PEEK, across quantities from 5 to 100, Xometry's price averaged 2.0× the median China direct price. It was never 1.5×. It was never 3×. It hovered around 2× on everything we quoted.
That's useful as a mental shortcut. Take your Xometry instant quote. Divide by two. That's roughly what a competent China direct manufacturer would charge. The actual number will move around depending on material, complexity, and quantity, but it's a better starting assumption than guessing blind.
2. The premium isn't in the material — it's in the platform overhead. Look at Part A ($16 vs $38 on a part where material is $3) and Part D ($52 vs $89 on a part where material is $20). The absolute premium is similar ($19.70 vs $37.50), but the percentage premium shrinks on expensive materials because raw stock cost is the same either way. The platform markup is on machining and overhead, not material.
What this means: if your parts use expensive materials (PEEK, titanium, 17-4 PH), the percentage savings from going direct is smaller. If your parts use cheap material (aluminum, mild steel), the percentage savings is larger.
3. Lead times were comparable. Xometry quoted 8–15 business days. China direct quoted 7–20 days. Nobody was dramatically faster. The real difference is in shipping — Xometry ships from US-based manufacturing partners so domestic shipping is 1–3 days. China direct requires international shipment (3–7 days express, 14–30 days economy). If you're not in a rush, the landed timeline is similar. If you need parts on your desk Friday, Xometry's domestic advantage matters.
4. China quotes had more variance. The three direct manufacturers ranged from $16 to $21 on Part A — a 31% spread. Xometry has one price. The China market is competitive but not transparent; you have to quote with multiple shops to find the market-clearing price. Xometry gives you that price instantly.
5. Volume discounting was weaker on Xometry. Going from qty 5 to qty 100, the China direct price on Part E dropped from about $58 to $42 — a 28% reduction. Xometry dropped from about $112 to $94 — a 16% reduction. Platform pricing is less sensitive to volume because the manufacturing partner still needs their margin and Xometry still needs theirs. Direct manufacturers can spread setup and programming cost across more parts.
Where the Xometry premium is worth it
This isn't a "Xometry bad" article. I've used them. I understand why people use them. The premium buys specific things:
When you need one or two parts tomorrow. Xometry's instant quoting with domestic manufacturing means you can have parts in 3–5 days without an international shipment. For urgent prototyping, a 2× premium to save a week of downtime is often a no-brainer.
When you don't have a supplier relationship. Qualifying a new Chinese manufacturer takes time. You need to assess communication quality, verify certifications, test with a small order, and build trust. Xometry removes that friction. If you order CNC parts twice a year and don't want to manage a supplier relationship, the 2× premium buys you convenience.
When quality consistency across many part numbers matters more than per-part cost. Xometry's quality system means you get consistent results without having to audit individual shops. For programs with dozens of different part numbers, the administrative cost of managing multiple direct suppliers can exceed the machining cost savings.
When your finance team hates international wires. Xometry invoices in USD with a US entity. China direct suppliers typically invoice via wire transfer (SWIFT) or Alibaba Trade Assurance (2–3% fee). If your AP department charges $200 in internal processing per international wire, that eats into the savings fast.
When ITAR or export control applies. Xometry has ITAR-registered partners. Most China direct shops don't — and shouldn't — touch ITAR parts. That's a hard line, not a cost decision.
Where direct China sourcing wins
When you're ordering production quantities. At 100+ pieces, the 2× gap translates to thousands of dollars. The savings fund your supplier qualification effort many times over.
When you want a long-term manufacturing partner, not a transaction. A direct supplier learns your parts, your standards, your packaging preferences. By the third order, you're not explaining from scratch. Xometry routes each order to whichever partner has capacity — you're starting fresh on every PO.
When your parts need process optimization. A direct manufacturer can suggest: "If we add a dovetail here, we can eliminate this secondary operation and save 15%." Xometry's platform generally makes what you send them — they're not paid to optimize your design.
When you need capabilities that don't fit the platform mold. Xometry's instant quoting has geometry limitations. Overhangs beyond certain angles, deep pockets with small corner radii, polished surface finishes — these get flagged for manual review or rejected. A direct shop can evaluate the whole part and say "we can do this, here's how."
The decision framework
Here's how I think about it. Run your project through these questions in order:
1. Is it ITAR or export-controlled? → Xometry (or another domestic source). China is not an option.
2. Quantity under 10 and timeline under 5 days? → Xometry. The speed premium is worth it.
3. Quantity over 100 with recurring orders? → Direct China. The savings compound.
4. Complex geometry that needs DFM iteration? → Direct China. You want a manufacturing partner who will optimize with you.
5. One-off project, no existing supplier relationships? → Xometry. The convenience premium is justified.
6. Annual spend over $20,000 on CNC parts? → Direct China. At that volume, the savings from even a 1.5× reduction pay for the qualification effort within a few orders.
This isn't theoretical. I've seen companies move from Xometry to direct sourcing and cut their CNC spend by 40% without touching quality. I've also seen companies try to save money by going direct, burn weeks qualifying a supplier who turned out to be a trading company with no machines, and go back to Xometry with relief.
Both paths work. The trick is knowing which path your project is on.
If you're sourcing CNC machined parts and trying to figure out whether direct China manufacturing makes sense for your volumes, send me your drawings. I'll give you an honest assessment — including whether I think you'd actually save money or whether you're better off staying with a platform. No pitch, just real numbers from someone who does this every day.