Small Quantities: Buying 1-10 Parts With No MOQ

Buying one to about ten parts is a normal corner of custom manufacturing. The processes that serve it best share one trait: no per-part tooling to pay for. CNC machining, sheet metal cutting and bending, and 3D printing all qualify. The catch is that some costs refuse to shrink with quantity. Programming, fixturing, setup, and queue time are close to fixed for a job. At single-digit quantities they land on very few parts, and the per-part price can look absurd even though nothing about it is unfair. This page explains why minimums exist, where the floors sit, and how to buy one to ten parts well. For tens of parts and beyond, the companion guide is low-volume manufacturing.

Why minimums exist at all

A supplier declining a five-part order is usually not guarding a fat margin. Nearly every custom job carries a block of work that happens once, before the first good part exists: reviewing the drawing, programming tool paths, sourcing material, staging cutters, arranging workholding, and proving out the first piece. Paperwork happens at both ends. That block takes roughly the same hours whether the job makes one part or one hundred. Spread across a thousand parts, it disappears. Landed on two parts, it dominates, and a shop that quotes it honestly will quote a unit price that startles the buyer.

Machinist community discussions describe this plainly. Shop owners report that even a trivial-seeming one-off consumes a couple of hours from purchase order to shipment, spent on programming, material pickup, and packing. The fixed block varies by process, but its shape is consistent:

• programming and tool-path or nest preparation
• workholding: soft jaws, fixtures, or vise setups, built per design
• tool staging and first-piece proveout
• material sourcing, including any stock bought in lengths that exceed the job
• inspection planning and any first-article check
• order overhead: paperwork, shipping, invoicing

Minimums come in two shapes. A piece-count minimum says the supplier will not quote below some number of parts, and it is common where jobs carry heavy fixed work or dedicated tooling. A job minimum is a dollar floor on any order, however few parts it has. Small shops report setting them anywhere from the low hundreds of dollars to well over a thousand, while others set no floor and bill by the hour. Tooling-based processes sit at the far end of the spectrum: injection molding and die casting tooling often reaches the tens of thousands of dollars, which is why their minimums are quoted in the thousands of parts. Neither shape is a scam. A minimum is an amortization fact stated as a policy.

Processes that serve single-digit quantities with no real minimum

The pattern to look for is simple. There is no per-part tooling, and the fixed work is small enough that a single part can carry it. Three process families fit.

CNC machining

CNC machining needs no mold or die. A program is written once, the part is fixtured once, and the machine can make one piece as readily as fifty. Feasibility is almost never the limit at quantity one. Economics is. A simple aluminum bracket quotes happily at one piece. A multi-operation part in an exotic alloy is still makeable at one piece, but the setup and material risk stacked on a single part can push the price past what makes sense. Common alloys such as aluminum 6061 and stainless 304 are the friend of a small order: the stock is easy to buy in small amounts, and the cutting data is familiar in any shop.

Sheet metal cutting and bending

Laser, waterjet, and plasma cutting start from standard sheet and plate, and press brakes form standard bend radii with tooling the shop already owns. Nothing in that route is dedicated to the buyer’s part, so a single bracket or enclosure from standard-gauge stock carries no tooling bill. The costs that do bite are programming the flat pattern, nesting it onto a sheet, and setting up the brake. Each of those is fixed per design. Custom form tooling would reintroduce a tooling cost, but that belongs to volume production, not single-digit buying.

3D printing

Additive manufacturing is inherently a one-off process. There is no fixture and no per-part program beyond file preparation and build layout, so fixed per-job work is small and price scales mostly with part volume and machine time. The one-to-ten penalty is flatter than in machining: the second part costs close to the first. The trade is in materials and properties, which do not always match machined metal. Printing wins the single-digit question most clearly for plastic parts, complex geometry, and internal features machining cannot reach.

Where the floor really sits

For CNC, the floor is setup and programming hours, which makes quantity one easy to buy and easy to overpay for. In sheet metal it is flat-pattern programming and brake setup, and in printing it is file preparation and machine occupancy. A supplier in any of these processes can and often does quote exactly one part. The floor the buyer collides with is the job-level floor, not the process.

How unit price behaves from one to ten

Per-part price at small quantities is fixed cost divided by quantity, plus a marginal cost per part. The division term makes the curve. Going from one to two halves the fixed burden per part; going from nine to ten changes it by about a tenth. As a hedged, industry-typical illustration drawn from supplier pricing examples, a single part often prices at several times its own ten-piece per-part price. Moving from one to five can cut the per-part price close to half. Beyond ten to roughly fifty the curve flattens: material and machine time, which scale with each part, take over from setup. The cost comparison by process page maps that flattening across processes.

Why pricing one, five, and ten together works

The steepness of the curve depends on how much fixed work a specific design carries, so a buyer cannot know where it flattens in advance. Asking a supplier to price one, five, and ten of the same part in a single pass reveals it. The three numbers show the setup load. They also show the point where extras become cheap, which feeds the spares decision below.

How to buy small quantities efficiently

The levers all attack the same fixed block:

• Add copies of the same part. Ten identical parts are one job with one setup, not ten jobs, and each copy past the first carries mostly material and machine time.
• Batch different parts into one order. Parts sharing a material and thickness share stock, nesting, and shipment, and a consolidated order can clear a shop’s job minimum on its own. That bundling is the standard workaround for dollar floors.
• Standardize on stocked sizes. Common alloys in standard bar, sheet, and plate sizes avoid minimum mill quantities and remnant waste, both of which land hard on small orders.
• Relax tolerances that do not matter. Put general tolerances in the title block, in the style of ISO 2768, and reserve tight bands for the few features that need them. Every unnecessary one adds slower passes and inspection time.
• Send complete files and drawings up front. A STEP model with units stated, plus a drawing carrying material, finish, critical dimensions, and tolerances, lets a supplier price without assumptions.
• Accept standard finishes. As-machined, standard anodize, and passivate carry predictable costs. Specialty finishes bring lot charges that weigh heavily on a small batch.

The hidden cost: engineering back-and-forth

At these quantities the engineering attention around an order can rival the machine time. Clarifying questions, revised quotes, and missing-dimension delays are fixed costs per order. Shops answer an unclear package in one of two ways: they pad the price to cover uncertainty, or they stall the order with questions. Buyers report quoting processes stretching over months on incomplete packages. A complete first transmission is the cheapest lever here, and it is free.

Where to buy: three channel shapes

Small orders reach suppliers through three generic channel shapes, and each trades price, attention, schedule control, and engineering help differently.

Prototype and short-run shops

These shops are built around small orders. Quoting, fixturing, and scheduling assume runs from one piece into the hundreds, so a single-digit order is normal work, not a favor. Attention, engineering help, and schedule behavior tend to be good. The trade is unit price: convenience at small quantities is priced in, and the per-part cost at ten pieces can sit above what a production shop would charge at a hundred.

General job shops

A general job shop takes small orders but fits them around production work. Price at five to twenty-five pieces can beat a prototype shop’s, and access to real engineering conversation is often stronger. What it costs is schedule control: small jobs queue behind larger contracts, ship as gap fillers, or wait while machines are tied up. A buyer with a hard date needs to say so early.

Online marketplaces

Marketplaces aggregate many small orders into steady demand and standardize the quoting path. The reach is wide: many processes and materials in one place. The trades: engineering help runs thinner for an unusual part, there is less visibility into which shop runs the work, and unit prices include a platform margin. For a standard geometry in a common material the aggregation helps the buyer. For a demanding part it can remove exactly the conversation the part needs.

No channel is universally better. A simple aluminum part fits a marketplace, a complex multi-operation part wants a job shop with engineering support, and an urgent one-off belongs at a prototype shop.

Quality at one to ten parts

Process capability does not drop with quantity. A machining center holds the same tolerance on part one as on part one thousand, and a laser cuts the same edge on a one-piece nest as on a full sheet. What changes is verification economics. First-article inspection, dimensional reports, and material certifications are per-order costs, so at three parts they weigh far more per part than at three hundred. Statistical sampling means little on a lot that small.

Aerospace work formalizes the first-article check in standards such as SAE AS9102, which requires documented verification that every design requirement is understood and met on a representative part. At single-digit quantities that rigor can be a meaningful share of the order, so it should be chosen deliberately. The questions that settle the matter are few:

• What does the quoted price include: a first-article check, a dimensional report, a material certificate, or nothing beyond a visual check?
• Which dimensions will actually be measured, and with what equipment?
• Is a full dimensional report available as an option, and what does it add?
• Will the material certificate cover the lot, and does it cost extra at this quantity?

Asked once, in writing, these align the order with what the part needs.

When the no-minimum frame is the wrong frame

Buying the smallest possible quantity is not always right. Two situations flip the answer.

If reorders are coming

When more of the same part will be needed, a slightly larger first batch is usually the cheapest insurance. Once setup is paid, each additional part costs close to its material and machine time. Buying five when three are needed and treating the extras as spares costs far less than paying setup twice. This spares-and-kitting logic is the strongest lever here, and the replacement and spare parts page develops it for maintenance and repair work.

If the design is still changing

A design that may still change makes every order a prototype, and paying full setup for parts a revision will strand is waste. The right page then is rapid prototyping; the question becomes how to learn fastest rather than how to buy cheapest. When quantities grow past the single digits, the question moves to the low-volume manufacturing guide, where tooling trade-offs take over from setup arithmetic.

Checklist for buying single-digit quantities

• The process is a no-tooling family: CNC machining, sheet metal, or 3D printing.
• One, five, and ten are priced together so the setup load is visible.
• Copies of the same part are batched into one order, and parts share material where possible.
• Tolerances are tight only where function requires; general tolerances cover the rest.
• Files, drawings, units, material, and finish are complete in the first transmission.
• Inspection expectations are stated and priced, and the design will not be stranded by a revision.

Common mistakes at small quantities

• Comparing a one-part price to a hundred-part price and calling the supplier unreasonable, when the difference is setup arithmetic.
• Ordering exactly the number needed, then paying setup a second time for the reorder.
• Tightening tolerances across the whole drawing instead of on the two features that matter.
• Sending a file without units, material, or finish, then absorbing the padded price or the delay.
• Choosing a channel by price alone and losing the schedule control or engineering help the part needed.
• Skipping the inspection conversation, and paying for a report nobody reads or none at all.