De Minimis Is Gone — But Formal Entry Isn't Mandatory. The $2,500 Line US Buyers Should Plan Around

What actually changed

On 24 June 2026, US Customs and Border Protection issued interim final rules indefinitely suspending the de minimis administrative exemption for shipments valued at $800 or less, across all modes of transport. This followed the earlier suspensions — China and Hong Kong on 2 May 2025, then all other origins on 29 August 2025.

The practical effect: there is no longer a value threshold below which a commercial shipment enters the United States duty-free and paperwork-free. Every commercial shipment now requires an entry filed in CBP's Automated Commercial Environment, with duties paid.

That much you have probably read. Here is the part most coverage gets wrong.

Formal entry is not required on every shipment

A great deal of writing on this topic collapses "de minimis is gone" into "formal entry on everything." Those are different claims, and the second one is not true.

Informal entry — Entry Type 11 — remains available for commercial shipments valued at $2,500 or less. It is a lighter, cheaper process than formal entry. It exists, it was not repealed, and it is now the main pathway for the small orders that used to travel under de minimis.

Formal entry (Type 01) is required when a shipment exceeds $2,500 in value, is subject to a quota, or is subject to antidumping or countervailing duties.

There is also a third channel that behaves differently from both: the international postal network. For postal items from China, the duty collected by the carrier is either 120% ad valorem or $200 per shipment, in lieu of other duties. A new postal informal entry process for mail shipments valued at $2,500 or less took effect 24 July 2026, and CBP has announced a voluntary electronic test, Entry Type 13, for mail shipments at or under $2,500 beginning 22 September 2026.

So the correct mental model is three channels, not one:

If you take one thing from this article: the number to plan around is $2,500, not $800.

The fee with a floor — and why it reaches further than you think

The Merchandise Processing Fee is charged at 0.3464% of entered value. But for formal entries it carries a statutory minimum of $33.58 and a maximum of $651.50 for fiscal year 2026 (1 October 2025 – 30 September 2026), plus a $4.03 surcharge if the entry is filed manually.

The minimum is the interesting part, and it is widely misunderstood as something that only bites on very small orders. Run the arithmetic on where the floor stops binding: $33.58 ÷ 0.3464% = $9,694.

Below roughly $9,694 in entered value, you pay the $33.58 floor regardless of how small the shipment is. A $500 order and a $9,000 order pay the same MPF. Only above that point does the fee become genuinely proportional.

Ocean shipments also carry the Harbor Maintenance Fee at 0.125% of value. That one is genuinely proportional — no floor — so it does not distort small orders the way MPF does.

The full fixed-cost stack per entry

MPF is only one line. The charge that usually dominates is the customs broker's fee, and here the honest answer is a range rather than a figure: brokers price differently by entry complexity, line-item count, and relationship. Industry-published ranges for 2026 put a standard formal entry at roughly $150–$400, with simpler entries lower and multi-line or agency-flagged entries higher. Treat those as indicative ranges published by brokers and freight platforms, not as a settled market price — get a quote for your own volume.

Single-entry bonds are published in the $50–$250 range per entry, again varying with bond amount and shipment value.

Stack the fixed components on a formal entry and you get, very roughly, $234–$684 per shipment before a cent of duty. Against goods value:

Note that informal entries carry a different and lower MPF treatment than the formal-entry floor described above. The exact informal figure varies by filing method, and it is worth asking your broker for their specific number rather than relying on any published summary — including this one.

The decision framework

Once you see the fixed-cost stack, the question stops being "how do I reduce duty" and becomes "how many entries am I paying for."

Get a continuous bond if you import three or more times a year. A continuous customs bond covers unlimited shipments at every US port and renews annually. Published pricing for the minimum $50,000 bond runs roughly $400–$600 per year, with wider ranges depending on credit and import history. The break-even against single-entry bonds sits at about three shipments in a twelve-month period. If you clear that bar, this is straightforward money left on the table — and it requires no agent, no consultant, and no change to your supplier relationships. Do this first.

Stop splitting orders across shipments to smooth cash flow. This is the most common self-inflicted cost. Two $2,500 shipments carry two full fixed stacks; one $5,000 shipment carries one. The saving is the entire duplicate stack.

Reconsider the small pilot order. Testing a new product with a $500 shipment now carries fixed costs that can exceed the value of the goods themselves. That does not make pilots wrong — validating a product before committing to volume is still usually cheaper than being wrong at scale. But the true cost of a pilot is now the goods plus a few hundred dollars of fixed overhead, and it should be budgeted that way rather than discovered at clearance.

Combine orders from multiple factories into one shipment where you can. This is where the fixed-cost logic points hardest, and it is the one move that requires coordination you may not have in place. Each separate factory shipping separately is a separate entry with its own full stack.

Where this framework does not apply

Worth stating plainly, because a framework that claims to cover everything is not useful:

• This is US-specific. The EU, UK, and Australia each have their own low-value import regimes with different thresholds and mechanics. If you import into those markets, none of the numbers above transfer.
• Quota, antidumping and countervailing duty goods force formal entry at any value. If your product sits in one of those categories, the $2,500 line is irrelevant to you.
• Duty itself is not addressed here. This article is about the cost of entry — the fixed administrative stack. Tariff rates are a separate and much larger question depending on your HTS classification and origin.
• The figures have a shelf life. CBP adjusts user fees annually, and FY2027 rates take effect 1 October 2026. The mechanism — a floor that makes MPF regressive on small orders — will outlast the specific numbers, but re-check the numbers.
• This is not customs or legal advice. Classification and entry decisions should be confirmed with a licensed customs broker.

Where a sourcing partner changes the maths

Most of what is above, you can act on yourself. The continuous bond in particular is a pure in-house win, and you should not pay anyone a percentage to tell you to get one.

The part that is genuinely hard to do alone is the last item in the framework: consolidating orders from multiple factories into a single shipment and therefore a single entry. If you are buying drinkware from one factory, packaging from a second, and tote bags from a third, three separate factory shipments mean three entries and three full fixed stacks. Combining them requires someone in China who can receive, hold, check, and consolidate across suppliers before anything sails — because the factories will not coordinate that with each other on your behalf.

That is the structural reason a China-side partner can reduce your landed cost on multi-supplier orders: not by negotiating a better factory price, but by collapsing several entries into one. The saving is the duplicate fixed stacks you no longer pay.

This is what we do at Source with Tophney. We receive goods from each factory in China, check them, hold them, and consolidate them into a single outbound shipment — which clears as one entry instead of several. Our 1Stop Depot sits on top of that, giving you access to a consolidated supply chain across six core product categories ready to customize with your brand colors and logo, so multi-category ordering through one channel is practical in the first place.

Put numbers on it. Using the indicative fixed-cost stack from earlier — roughly $234–$684 per formal entry — a four-supplier order clearing as four separate entries carries about $936–$2,736 in fixed costs. The same order consolidated into one entry carries $234–$684. The difference, roughly $700–$2,050, is money that buys you nothing: it is the same paperwork filed four times.

Consolidation is not a line item we bill for. It is part of the service fee, which is published in full on our service page — a refundable sample fee of $100–300 per shipment, and a service fee of 5–10% on the bulk order, scaled to how custom the product is. There is no separate warehousing or consolidation charge on top, which is why the entry saving above is a genuine saving rather than a cost moved from one line to another.

Whether the maths lands in your favour still depends on the shape of your order, and it genuinely does not always:

• Single factory, single shipment. There is nothing to consolidate. The entry saving is zero, and you are paying a service fee for sourcing and quality control, not for landed-cost reduction. Judge it on that basis.
• Multiple suppliers, small total value. A four-supplier $6,000 order saves perhaps $700–$2,050 in duplicate entries, against a service fee of $300–$600. The consolidation alone can cover the fee.
• Multiple suppliers, high customization. At the 10% tier the fee scales with order value while the entry saving does not — it is a flat function of how many entries you collapse. On a large fully-customized order the fee is buying product development, not entry savings, and should be evaluated as such.

The honest version: consolidation is a real, computable saving on multi-supplier orders and no saving at all on single-supplier ones. If you buy from one factory, get a continuous bond, file your own entries, and skip the agent.