The published floor across named China suppliers is 5,000 to 10,000 pieces per item. Dalica publishes 5,000 per product; Topfeel Pack, UKPACK, Lisson and Vesla each publish 10,000 for custom moulds; Gidea PAC describes a typical industry range of 5,000–20,000; JingSourcing puts unique-shape cosmetic bottles at around 50,000 per style. Numbers below that are achievable, but only by changing what is being made — an existing mould with custom decoration instead of a new mould. That is the basis on which Vella publishes roughly 500 for undecorated stock and 2,000 for stock plus screen printing. And one figure nobody escapes: electroplated closures sit at about 10,000 whoever you ask.
This comparison is published by Vella, which appears in it. Every competitor figure below is quoted from that company's own public page and checked on 30 August 2026 — none is inferred, and none comes from a third-party listing. Where a company does not publish a company-wide figure we say so rather than estimate one. Published minimums are starting points, not commitments: treat this as a way to calibrate expectations before you brief, not as a substitute for a written, SKU-specific quote.
What nine suppliers publish, in their own words
| Supplier | Published minimum | Where it is published |
|---|---|---|
| Dalica China manufacturer |
“MOQ typically starts from 5000 pieces per product” | FAQ on dalicapkg.com |
| Boyu Packaging China manufacturer |
Plastic from 5,000; glass from 10,000; lead time 20–35 days | Its own supplier roundup, 20 Aug 2026 |
| Gidea PAC China manufacturer |
“a typical MOQ ranges from 5,000 to 20,000 units” | Its MOQ explainer page |
| Topfeel Pack China manufacturer |
“For most items, the MOQ starts at 10,000 pieces” | FAQ on topfeelpack.com |
| UKPACK China manufacturer |
Custom airless jars: “standard MOQ is 10,000 pieces per design”. Samples in 7 working days, mass production in 35 days. | Its airless jars product page |
| Lisson / Function Packaging China manufacturer |
Custom cosmetic tubes: “MOQ typically starts from 10,000 pieces” | Its tube customisation page |
| Vesla China glass manufacturer |
“MOQ: 5,000 pcs on stock molds; custom molds from 10,000 pcs”. Production 7–15 days on stock molds, 25–30 days custom. | Capabilities section, veslapack.com homepage |
| JingSourcing general sourcing agent |
Items requiring mould development: “MOQs typically start at around 10,000 pcs”. Unique-shape cosmetic bottles: “around 50,000 pcs per style”. | Its MOQ negotiation guide |
| Vella beauty packaging sourcing partner |
Undecorated stock from about 500; stock plus screen print from 2,000; sprayed custom colour 2,000–5,000; electroplated closures 10,000; new mould 3,000–10,000 plus RMB 25,000–70,000 tooling; digital labels and sleeves about 100 | Published across this site |
All competitor figures quoted from the named company's own public page, checked 30 August 2026. Boyu's figures are as stated in its own published roundup rather than a product page. Figures change; re-check before relying on any of them.
Read that table correctly, or it will mislead you
Two suppliers can publish the same number and mean completely different things. Before comparing anyone's MOQ with anyone else's, three distinctions do most of the work.
1. A minimum is per component geometry, not per finished unit
This is the single most common misreading. A pump bottle is not one item — it is a bottle, a pump and usually an overcap, frequently produced in three different plants, each carrying its own minimum. A published figure of 5,000 generally means 5,000 of each part. A brand that reads it as "5,000 finished bottles" and budgets accordingly gets an unpleasant surprise at quotation.
The same logic bites on colour. Spray minimums attach to the shape being sprayed, not to the colour. Four different geometries in one colour is four spray set-ups, not one — the shapes do not consolidate into a single run. This is why a matching six-piece collection is a harder brief than it sounds, and why the sensible first question is which items already exist in the finish you want.
2. Stock mould versus new mould is the real dividing line
Almost every dramatic difference in published minimums comes down to this one question, not to how generous a supplier is feeling.
- An existing mould already exists and has been paid for. The factory is only being asked to run it in your colour with your decoration. Minimums here are genuinely low.
- A new mould has to be cut before anything is made — typically RMB 25,000 to 70,000 for a cosmetic component — and the factory will want 3,000 to 10,000 pieces to justify setting it up. JingSourcing's published figure of around 50,000 for a unique-shape bottle is the upper end of the same principle.
So "can I get a low MOQ?" is usually the wrong question. The useful one is: does a stock mould exist that is close enough to what I want? If it does, low minimums follow automatically. If it does not, no supplier can conjure them.
3. Some minimums cannot be moved by anyone
It is worth knowing in advance which parts of a brief are negotiable and which are physics.
| What you want | Realistic minimum | Movable? |
|---|---|---|
| Digital label or shrink sleeve | ~100, effectively none | Yes — very flexible |
| Undecorated stock component | ~500 | Yes |
| Stock component + screen printing | ~2,000 | Some |
| Sprayed custom colour | 2,000–5,000 per geometry | Limited |
| Hot foil stamping | ~2,000 | Some |
| New mould | 3,000–10,000 + tooling cost | No — tooling is fixed |
| Electroplated / metallised closure | ~10,000 | No |
Electroplating deserves its own warning because it ruins more launch plans than any other single decision. The plating line has a batch size that does not scale down. A brand ordering 2,000 units of five items has 10,000 gold caps and clears the bar; drop to four items and it has 8,000, and the gold simply stops being available — not at a surcharge, at all. If a metallic finish is central to your identity, count the pieces before you cut the range.
Why a sourcing partner can publish a lower number — and where it cannot
Vella publishes 500 and 2,000 against an industry that publishes 5,000 and 10,000. That difference deserves an honest mechanical explanation rather than a claim, because the mechanism is also the limit.
Vella is not a factory. A factory quotes the minimum that makes its own line worth running, and it is right to. A sourcing partner is doing something structurally different: routing each component to whichever factory is already tooled and already set up for small runs, and substituting an existing mould with custom colour and print for a newly cut one.
The saving comes from avoiding the tooling, not from negotiating a factory down. Where no small-run route exists, a sourcing partner cannot beat the factory minimum either. Electroplated closures are still around 10,000 through Vella. A genuinely new mould still costs RMB 25,000–70,000 through Vella. Anyone promising you a custom mould at 500 pieces is either absorbing a loss they will recover elsewhere, or has not told you it is a stock shape.
When negotiation actually works
The most useful published guidance on this comes from JingSourcing, a general sourcing agent rather than a packaging supplier, and it matches what happens in practice: negotiation succeeds when the gap is small. Its own framing is that where a factory's minimum is 2,000 and you need 1,500, there is usually room. Where it is 10,000 and you need 1,000, there is not — and pressing on it wastes the goodwill you will need later on lead times and quality.
For a large gap, the productive move is to change the specification rather than the number:
- Existing mould instead of a new one — the single biggest lever available.
- Label or sleeve instead of direct decoration — drops the decoration minimum from thousands to roughly a hundred.
- Screen print instead of sprayed colour — a coloured print on a clear or white stock component reads as brand identity at a fraction of the minimum.
- Fewer geometries, more sizes of the same geometry — three sizes of one shape share tooling and spray set-up; three different shapes do not.
- Anodised aluminium instead of sprayed metallic — more durable on a twisted lid and avoids the plating batch problem entirely.
What a low minimum costs you
It costs more per unit, and any supplier telling you otherwise is not being straight. Set-up costs — tooling, print screens, machine changeover, inspection — are fixed, and a smaller run spreads them across fewer pieces. Expect roughly 20 to 40 per cent more per unit than at volume.
That premium is usually the right thing to pay on a first run. You are buying the option not to commit cash to inventory of a product that has not sold yet. The mistake is not paying it — the mistake is paying it on a second run, after the product has proven itself and the quantity should have gone up.
Vella is a sourcing partner, not a factory, working with vetted ISO 22716 / GMP partner factories. That model is strongest in one specific situation: you want a custom look across several formats, at quantities a factory would refuse, with quality inspected and one shipment consolidated for you.
It is the wrong choice if you want a single unit of a stock jar shipped tomorrow — a domestic stock shop beats us on that, and we will say so. It is also the wrong choice once your volumes are genuinely high and a format is proven, at which point going direct to the factory that already makes it will cost you less. See how low-MOQ sourcing works → or send a brief.
Frequently asked
Tell us the quantity you
actually want to order.
Send your formats and target volume. We come back within 24 hours with which parts of the brief fit a low minimum, which do not, and what it would cost either way — including when the honest answer is that someone else suits you better.
Send a brief →