Web Analytics

WE ARE A PACKAGING COMPANY AND ALL OUR PACKAGES ARE SOLD EMPTY

What Counts as a "Container" Under the 0.4 mg THC Cap?

by Roshan Timsina

The FDA has not said. Here is how the packaging formats break down, what it costs to repackage each one, and how many reorder cycles are left before 11 December 2026.

By Roshan Timsina, Head of Marketing, Black Unicorn Hub · Published 24 September 2026 · Update log

Two deadlines, not one

Most operators we speak to believe there is one federal deadline. There are two, and which one applies to a given product depends on how its cannabinoid was made.

Date What loses hemp status
12 November 2026 Cannabinoids that cannot be naturally produced by a Cannabis sativa L. plant. This date was not extended
11 December 2026 Everything else caught by the new definition, including anything over 0.4 mg total THC per container

The second date moved when H.R. 6500, the Continuing Appropriations and Extensions Act 2027, passed the House 370 to 48 on 1 September and was signed the following day. The first did not move at all. Cannabinoids synthesised from hemp-derived CBD stay on the November date.

So a brand with both a delta-8 line and a naturally derived line is working to two separate deadlines, 29 days apart.

The cap that matters for packaging is 0.4 mg of total THC per container. Not per serving. A 5 mg gummy carries 12.5 times the limit. A 10 mg beverage carries 25 times it. The arithmetic is done on the whole package, which is why this is a packaging problem and not only a formulation one.

And the FDA has not defined "container."

That single undefined word decides whether a ten-pack of individually wrapped gummies is one container or ten, and therefore whether a large part of the market is compliant or not. The agency has also not published the cannabinoid lists the law requires. Four bills that would delay, repeal or rewrite the framework sit in committee with no scheduled votes, and the White House has said the 30-day extension was the last one it will support.

Who we are and why we are writing this. Black Unicorn Hub sells packaging, and we are launching custom production. That is a commercial interest and you should read this with it in mind. It is also why we can tell you what a printing cylinder costs and how long a proof cycle takes, which is the part of this problem the legal analyses stop short of. Every figure below is sourced, and where we have estimated rather than sourced, the text says so.

What a container probably means

Here is our read. We may be wrong, and if the FDA publishes something that contradicts it we will say so on this page rather than quietly editing it.

A container is the sealed unit a consumer opens to reach the product.

That gives one practical test, and it is the only one we have found that sorts the hard cases consistently:

Is the inner unit independently sealed?

If yes, the inner unit is the container and the outer is packaging.
If no, the outer is the container and everything inside it counts together.

Under that test, a carton holding ten individually sealed 0.4 mg sachets is ten containers. A resealable pouch holding ten loose 0.4 mg gummies is one container carrying 4 mg, which is ten times the cap.

The physical difference between those two products is small. The regulatory difference, if our read holds, is the difference between a compliant SKU and an unlawful one.

Why we think this reading holds

The cap exists to stop a consumer accessing an intoxicating dose from one purchase unit. A seal is what makes a unit separate in any practical sense: it is what a consumer has to breach, what a child would have to breach, and what child-resistance standards are already written around. A reading that ignored seals would make an outer shipping carton of 500 sachets a single container, which nobody thinks is intended.

Why it might not

The strongest argument against us is that the statute says "container" and a carton is, in ordinary English, a container. A regulator worried about operators gaming the rule by subdividing could read it that way deliberately.

That worry is not hypothetical. CannabisRegulations.ai notes that operators are already reacting by splitting a 20 mg product into 50 single-dose containers, and warns the approach carries real legal uncertainty because the result is "distinct from but practically very similar to" the original product.

So the honest position is this: the sealed-unit reading is the most defensible one available today, it is not settled, and anyone betting a full production run on it should know they are taking a position rather than following a rule.

Every format, and where it stands

Applying the sealed-unit test to the structures actually in use.

Packaging structure Our read Exposure What we would do now
Single sealed pouch, one dose One container, the pouch Low if dosed at or under 0.4 mg Confirm the dose is per container, not per serving
Resealable pouch, loose units inside One container, the pouch. Contents count together High Reformat to sealed inner units, or cut the count to fit the cap
Carton, individually sealed inner units Ten containers, not one. The open question Medium Defensible, but do not bet a full run on it. Keep proof of the inner seal
Rigid jar, loose units One container, the jar High Same as the resealable pouch. No amount of labelling fixes it
Rigid jar, sealed sachets inside Sachets are the containers Medium Same open question as the carton
Beverage, single can or bottle One container High at any normal dose. 10 mg is 25 times the cap Reformulate or exit the format
Beverage, multipack in a carrier Each can is a container. Carrier is not sealed Low to medium The strongest multipack position, since the seal argument is unarguable
Bulk or wholesale pack One container Highest No packaging fix exists. This is a business model question

The three rows that matter

The resealable pouch with loose units is the biggest exposure in the market and the most common format in hemp edibles. There is no clever packaging answer. Either the total across the pouch comes under 0.4 mg, or the units get individually sealed, or the SKU goes.

The carton with sealed inner units is where the money is riding. If our read holds, this is the escape route for most multipack SKUs and it requires new inner packaging but keeps the product. If it does not hold, everyone who took it repackaged twice.

The beverage multipack is the safest position of any multipack, because a cardboard carrier is self-evidently not a sealed container. If you are arguing the multipack case to a regulator, that is the example to start from.

One thing that does not help

Labelling. Printing "contains 10 separate containers" on an outer pouch does not make it ten containers if the units inside are loose. The seal does the work, not the words. We mention it because we have already been asked twice whether a label change is enough, and it is not.

What it costs to repackage one SKU

Nobody has published this, so here it is with the working shown.

These figures are derived from pricing published by five packaging suppliers, aggregated into ranges rather than attributed, because individual quotes are commercially sensitive. They are a starting point, not a final answer. Where a number is our own estimate rather than a published figure, it says so on the line.

Rotogravure, the standard path for an established SKU

Line Basis Cost
Engraved cylinders One per colour at $120 to $200 published. Worked here on a six colour design $720 to $1,200
Minimum run 10,000 units, the published floor for rotogravure pouches
Pouch cost 10,000 at $0.25 to $0.40. Our estimate, narrowed from published ranges $2,500 to $4,000
Artwork file setup Published rate, one additional SKU $125
Total per SKU $3,345 to $5,325

That is a wide range, deliberately. The cylinder charge and the minimum run are published figures. The per-unit pouch cost is our own estimate, narrowed from published ranges spanning $0.05 to $3.50, so it is the number most likely to move.

Colour count is the lever most people miss. Every colour is another cylinder, so simplifying artwork from eight colours to four takes several hundred dollars off the setup before anything is printed. Six is the worked example here, not a claim about what is typical.

Cylinders are a one-off. Reorder the same design and they cost nothing, which is why the second run is dramatically cheaper than the first and why changing artwork twice is the expensive mistake.

Digital, the path for a smaller or uncertain SKU

Line Basis Cost
Plate or setup charges None on digital, published $0
Minimum run 6,000 units, published
Pouch cost 6,000 at $0.35 to $0.55. Our estimate. No supplier we found publishes digital per-unit pricing $2,100 to $3,300
Artwork file setup $125
Total per SKU $2,225 to $3,425

Digital costs more per unit and less in total. The crossover sits somewhere around 12,000 to 15,000 units, so below that digital usually wins even before you count the risk of committing cylinders to a design that may change when the FDA finally publishes.

That last point is the one we would press. Anyone cutting rotogravure cylinders before the container definition is settled is spending $720 to $1,200 on a design that a single FDA notice could invalidate. Digital costs more per unit and buys you the option to change your mind. Until the definition lands, that option is worth more than the unit-cost saving.

What this excludes

Design work. The $125 above is file setup, not design. Actual artwork revision to add or change compliance text is a separate cost and we could not source a reliable figure for it, so it is excluded rather than guessed. If you have a real number from a studio, send it and we will add it with attribution.

Regulatory label review. A lawyer reading the label against the rule. Also excluded for the same reason.

Write-off of existing stock. Any run printed before September with the old declaration is scrap on 11 December. At the figures above that is $2,500 to $4,000 a SKU gone before the replacement is ordered. For a brand with twenty SKUs, the write-off alone runs into six figures.

So treat the totals above as the floor, not the estimate.

Ordering 10,000 does not get you 10,000

Custom pouch production carries a manufacturing tolerance. One supplier publishes theirs at plus or minus 20 per cent, and we could not find a second who states it openly, which is itself worth knowing. Order 10,000 on those terms and the contract allows delivery of anywhere between 8,000 and 12,000. Both are a complete order and both are billed.

If your supplier has not told you their tolerance, ask. It is in the terms whether or not it is on the website.

In normal trading this is a rounding annoyance. Against a fixed legal deadline it is a real cost, because a short delivery cannot be topped up inside the remaining calendar.

To be confident of 10,000 compliant units you have to order about 12,500:

Order 10,000 Order 12,500
Worst-case delivery 8,000 10,000
Pouch cost at $0.32 $3,200 $4,000
Cylinders $960 $960
Total $4,160 $4,960
Cost per guaranteed unit $0.52 $0.40

The headline per-unit price is not the price. Guaranteeing supply costs roughly 19 per cent more than the quoted total, and a brand that budgets off the quote and receives the low end of the tolerance is short with no time to reorder.

What we would do: order for the worst case, not the midpoint. The overage is cheap compared with being 2,000 units short on 5 December with every converter in the country running at capacity.

One more cash detail that catches people out. A 50 per cent deposit up front is standard, with the balance due before shipment. On a 12,500-unit run that is roughly $2,500 leaving the business weeks before the stock arrives and months before it sells through. For brands already carrying dead inventory they are about to write off, that timing is the harder problem, not the total.

The calendar is tighter than the countdown suggests

78 calendar days remain until 11 December. That is about 55 working days once weekends and Thanksgiving come out.

A production run is not 55 days of work, but it is not the 15 days a supplier quotes either. The quoted time starts at proof approval and ends at despatch. The real cycle is longer at both ends.

Stage Duration Basis
Artwork revision and internal sign-off 3 to 7 days Our estimate
Supplier quote and order placement 2 to 5 days Our estimate
Proof, revision, approval 3 to 10 days Our estimate
Production 15 to 18 business days Published
Freight and inbound 5 to 15 days Our estimate
Full cycle 4 to 6 weeks

Only the production line is a published figure. The rest is our estimate of how these jobs actually run, and we would welcome correction from anyone who does this every day.

Even read optimistically, at six weeks a cycle, 78 days is two full attempts. At four weeks it is three, and only if nothing goes wrong.

The date to put in your calendar

30 October is the last date to start a run with room for a failed proof. Start then, allow six weeks, and you land on 11 December with the second attempt used up.

After roughly 13 November you have one attempt and no margin. A rejected proof, a colour that comes back wrong, a freight delay, and the stock arrives unlawful.

This is also why the 30-day extension helped less than it looked. It moved the deadline by 29 days, which is most of one production cycle, in a market where the binding constraint was never the deadline but the number of cycles before it.

Operators are already acting on that arithmetic rather than waiting. MJBizDaily reported on 15 September that hemp businesses are unwinding despite the reprieve, quoting Upstate Elevator chief executive Dylan Raap on raw materials needing months of lead time, co-packing needing 90 days, and packaging taking months to produce. Torch Drinks chief marketing officer Ryan Bouton told the same reporter his company plans in 30-day increments and cannot stock inventory that far ahead.

For scale on what is riding on the calendar: the hemp sector is estimated at $38.7 billion in annual sales supporting roughly 225,000 US jobs, according to Whitney Economics figures cited in that reporting.

What we would do

Three recommendations. These are opinions, held with the commercial interest declared at the top, and you should weigh them accordingly.

  1. Sort your SKUs by container structure before you do anything else. Not by product line, not by revenue. Run every SKU through the one test above: is the inner unit independently sealed? That sort takes an afternoon and it tells you which SKUs are fine, which need reformatting, and which have no packaging answer at all. Most operators we have spoken to are still thinking about this product by product, which is how the third group stays invisible until November.

  2. Print digital for anything you might have to change. Rotogravure is cheaper per unit and the saving is real at volume. It is also $720 to $1,200 of cylinders committed to a design that the FDA has not yet made final. Until the container definition is published, the ability to change a design cheaply is worth more than the unit-cost saving on most SKUs. Revisit that once there is guidance.

  3. Order for the bottom of the tolerance, and order by 30 October. Both of those cost money you would rather not spend. Both are cheaper than discovering on 5 December that you are 2,000 units short with every converter in the country booked solid.

If you are relying on the carton reading

The sealed-inner-unit position is the one keeping most multipack SKUs alive, and we think it is right. If you are betting on it, two things are worth doing now.

Keep documentary proof that the inner units are genuinely sealed: specifications, photographs, supplier confirmation. If the reading is ever challenged, the seal is your argument and you will want evidence it was there from the start, not reconstructed afterwards.

And split the risk. Run the first order at the minimum rather than the full quantity, so that if guidance lands against the reading you have written off one minimum run instead of a year of stock. The per-unit cost is worse. The exposure is a fraction.

What we do not know

Listing these because a guide that pretends to certainty on an unsettled rule is worse than useless.

  • Whether our reading of "container" is right. The FDA has not defined the term. We have given our reasoning and the counter-argument. Nobody has authority on this yet.
  • What the FDA will publish, and when. The law requires cannabinoid lists that have not appeared. There is no published timeline.
  • Whether the deadline moves again. Four bills sit in committee covering delay to 2028, full repeal, a federal framework and a beverage carve-out. None has a scheduled vote and the White House has said the last extension is done. We would not plan around another one.
  • How enforcement will work. Whether it targets manufacturers, distributors or retailers first, and whether early action is warning letters or seizures, is unknown.
  • How states will respond. Federal compliance does not guarantee state compliance and several states are moving separately.
  • The real cost of artwork and regulatory review. Excluded from our model because we could not source it reliably. If you have a figure from a studio or a firm, send it and we will publish it with attribution.

If you have better information on any of these, particularly the first, we would rather be corrected in public than leave a wrong answer up. The contact is at the bottom and corrections go in the update log with credit.

This is not legal advice. We sell packaging. Decisions about whether a specific product complies belong with a lawyer who has seen it.

Method and sources

How the cost figures were built. Per-unit and setup costs are aggregated from pricing published by five packaging suppliers, presented as ranges rather than attributed to any one of them, because individual quotes are commercially sensitive. Cylinder costs, minimum order quantities, tolerances, deposit terms and lead times are taken from suppliers' own published pages, linked below. Where we have estimated rather than sourced a figure, the text says so at the point it appears.

This is version one. The cost figures rest on published pricing. We are gathering live quotes and first-party purchase data, and will revise these numbers and record the change in the log below. If you work in this trade and our figures are wrong, tell us and we will correct them with credit.

Legal and legislative facts come from law firm analyses and congressional reporting, linked below. We have not interpreted the statute ourselves beyond the container reading, which is flagged throughout as our opinion.

Sources

Update log

Date Change
24 Sep 2026 First published

Corrections and media

We answer media enquiries within 24 hours. If a figure here is wrong, tell us and it gets corrected with credit in the log above.

Roshan Timsina
Head of Marketing, Black Unicorn Hub
info.roshantimsina@gmail.com

Back to News