Storage containers compared.

The short version: for food you seal once and leave alone, the published barrier numbers point at foil-laminate mylar inside a rigid outer container: a 4.3 mil foil laminate is specified at 0.001 cc/m²/24 hrs oxygen transmission [1], while a 4.0 mil nylon/PE vacuum-sealer film from the same manufacturer is specified at 4.25 cc/100 in²/24 hrs [3]. For food you open every week, a jar earns its place on reseal, not on barrier. And no manufacturer in this set publishes a light-transmission figure for any finished food bag, so light is the one criterion here that cannot be ranked at all.

The five containers, compared

Every cell below is what a manufacturer, extension service or regulator publishes, reproduced in the units they published it in. Blank cells are absences, not omissions. The searches behind them are listed further down.

The unit trap, before you compare any two rows. Three oxygen-transmission unit systems appear across these datasheets: cc/m²/24 hrs [1][2], cc/100 in²/24 hrs [3], and cc per 100 sq in/24 hrs at 100% oxygen [4]. One square metre is roughly 1,550 square inches. A reader who sets "0.001 cc/m²/24 hrs" [1] beside "4.25 cc/100 in²/24 hrs" [3] without converting will understate the gap by about three orders of magnitude. Test conditions make it worse: only the 7.5 mil sheet states its temperature and humidity (77°F at 90% RH) [4]; the 4.3 mil and vacuum-film sheets give the ASTM method and no conditions at all [1][3].

Thickness is measured two ways by two sellers. One manufacturer publishes "7.5 mil per side (15 mil total)" [10]; the datasheet grade of the same nominal thickness states 7.5 mil as the film gauge, i.e. per wall [4]. This matters at the seal more than at the barrier: a bag manufacturer points out that a 7-mil bag presents "7-mil on each side = 14 mil total sealing thickness" [20], and one 12" tabletop sealer rated to 8 mil for ordinary bags caps "Metalized Food Bags" at 2.5 mil on the same machine [21].

For food kept sealed

This is the case the foil laminate is built for, and the case where the container is the smaller half of the decision. The published mylar specification for freeze-dried food is narrow: "5–7 mil thickness, no clear plastic window," plus an oxygen absorber [17]. The window is disqualifying, which is exactly why the kraft-window pouches in the table above sit in a different section.

Two preconditions travel with every shelf-life figure attached to a sealed container. The food has to be dry ("Foods should be below 10% moisture content for storing with absorbers" [18]), and a low-oxygen package does not make a wet food safe. NC State Extension states the consequence plainly: "This low-oxygen environment is just right for the development of several pathogens, but one to be specifically concerned about is botulism," and adds that "these disease-causing bacteria may not change the color or look of the food" [22]. Vacuum sealing does not substitute for either condition [11][22]. Extension measurement of small-scale freeze-dryer output found "the water activity did not ever rise above 0.6, indicating that no food safety or spoilage risk occurred" [26], and USU puts finished freeze-dried product at "about 0.3 or below, with many products falling in the range of 0.08–0.33" [17].

Absorber sizing is where sources stop agreeing. For a one-gallon container, the published figures in this set run 300 cc from extension [18] and 700 cc from the freeze-dryer manufacturer, carrying only the condition "For use in gallon-size Mylar bags" [27]; it is the highest figure found and the thinnest condition attached to one. Extension's 2011 numbers are quart 50 cc, gallon or #10 can 300 cc, and three 500 cc packets for a 5-gallon container [18]. Sizing is by residual air volume, not food weight, and the reference volumes are 18,942 cc for a 5-gallon pail and 3,980 cc for a #10 can [19].

Expect no visual confirmation. Absorbers take the oxygen fraction only, so "1/5 of the air volume will be gone but 4/5 of the air will always remain" [15], and a vacuum-packed look is not a test: "It's a misconception that a 'vacuum-packed' look confirms the oxygen has been removed" [14]. The published checks are mechanical: once cool, "try to pull the mouth of the bag open. If it doesn't come open…" the seal is holding, and folding the bag into a pail until trapped air pillows up and watching whether it deflates [13]. Foil micro-cracks at the folds are normal ("Tiny breaks in the foil layer…are not holes… their barrier properties are not significantly affected" [12]), and a bag that passes light through a flashlight test is not thereby defective, since "inside a metalized bag in a cool, dark storage place, there is virtually no effect from light" [12]. Absorbers themselves have a short open window: use within 15 minutes, and a spent one is "brown and clumped together compared to a fresh pink or gray powder" [18].

Seal temperature, where anyone publishes it: 400°F / 40 PSI / 1 sec for the 4.3 mil foil laminate [1], 375–425°F for a flat iron on 7-mil bags [20], and 250–300°F published by one seller for the bare PET layer only, not the laminate [8]. A home user who takes the 250–300°F figure to a foil bag will under-seal it. And a vacuum sealer is not disqualified as a sealer: one manufacturer endorses using a FoodSaver with the pump off, "just use the sealing feature" [15], while a bag maker says the machine cannot manage 14 mil of stack [20]. Both positions are published; neither is tested.

For food opened often

Here the ranking inverts. A mylar bag is a single-seal package that gives up height every time you cut it [14][15], and an oxygen absorber is single-use: once its capacity is spent it does not recharge, which is why re-adding one is part of every reseal. A jar closes and reopens indefinitely, and extension puts glass with an absorber or a vacuum seal at "many years" while zip-top bags get "a few months" and vacuum-in-plastic 2–3 years [16].

The sorbent also changes. For repeatedly opened containers the documented failure mode is moisture, not oxidation: a packaging manufacturer names freeze-dried goods in re-opened containers, along with salt and sugar, as the desiccant cases, and gives the physical separation rule if you run both: "Oxygen absorbers have built in moisture to make them work; it is imperative to keep the desiccants separated from the oxygen absorbers" [28]. Published desiccant sizing is "about 1 gram of silica gel for roughly 120 cubic inches of space," which works out at 1–2 grams for a standard mason jar and 5–10 grams for a quart or gallon mylar bag [29]. Whether desiccants and absorbers belong in the same container at all is disputed between manufacturers, and none of them publishes a test.

One practical consequence of the jar: it is clear. Extension recommends darkened jars for light-sensitive items such as dried milk [11], and the freeze-drying guidance rules out clear windows in bags for the same reason [17], while a manufacturer argues light is irrelevant in a dark cupboard [12]. Nobody publishes a number that would settle it, which is covered below.

For carrying or selling

A pouch that travels is judged on presentation and closure as much as on barrier, and this is where the two kraft products separate. The foil-lined kraft laminate is a real barrier material with a published OTR of 0.01 cc/m²/24 hrs [2]; the window pouches that look like it are not, and their manufacturers say so: "This product does not have foil lining" [4], with structures that contain no metal layer [4][6]. Neither window pouch publishes an oxygen or moisture transmission figure [4][6]. If a customer stores your product for a year, that difference is the whole product.

The rules on selling are state law, and they are not uniform. Some examples of the shapes they take, as published:

• Ohio restricts the packaging method directly: "Cottage food products may not be packed using reduced oxygen packaging" [23]. That reaches an absorber-in-mylar pack.
• Illinois permits dehydrated foods and states a target rather than a filing: "Are dehydrated foods allowed? Yes, as long as no prohibited ingredients (such as meat, dairy, etc.) are included. For safety, aim for a water activity of 0.85 or less" [24]. Its pH language, by contrast, requires a lab report [24].
• New York lists "Freeze-dried foods" among products prohibited under its home processing exemption, and prohibits home drying of fruits and herbs while permitting repackaging of commercially dried ones [25].

Labeling elements recur across states: producer name and address, product name, ingredients by weight, allergens, net weight. But the required disclosure sentence is written per state and does not port across a state line [23][24][25]. Sublime Pantry describes what these documents say; it does not tell you what is permitted where you live. The operative step is to ask your own state agency in writing and keep the reply.

What we could not source

• A light-transmission figure for any finished food bag. Every source treats light qualitatively. The only numeric optical property found in the whole search is 3.0% haze for a bare PET layer [8]. No container in the table above can be ranked on light, and no extension source publishes a lux figure, an exposure duration, or a vitamin-loss percentage to rank them against.
• Oxygen or moisture transmission for glass jars and HDPE buckets. Neither appears in any retrieved datasheet. The bucket case rests on a general statement about plastics at 1–2 years [11] and on the instruction to line them [18].
• Barrier numbers from the largest consumer mylar seller. Site-restricted searches for an OTR or WVTR value returned product and FAQ pages carrying marketing language only: no numeric barrier figure, no layer identities beyond "three layers," no seal temperature. The specification data in this article comes from a converter that publishes datasheets [1][2][3][4].
• Retail per-bag pricing for the case-only items. The kraft and vacuum SKUs above are sold by the case, so there is no comparable small-quantity price [3][5][6].
• Whether "7.5 mil" means one wall or two. One seller publishes per side and total [10]; the datasheet grade publishes film gauge [4]. Unreconciled, and it decides whether a given sealer is in or out of spec [20][21].
• A freeze-dried class on any absorber sizing chart. The charts stop at "pasta and thicker, less dense foods." The one freeze-dryer manufacturer figure found, 700 cc per gallon bag [27], carries no food class and no fill level.
• Sublime Pantry has bench-tested none of this. Every figure above is somebody else's published claim, with its units, conditions and publisher attached.

Next

• Kraft versus mylar — the bag decision, in detail.
• Six beliefs the manufacturers contradict — the folklore around sealing and absorbers.
• When stored food fails — what each container's failure mode looks like.

Sold by Sublime Pantry: Snack Bags 6×6 — 50 Pack — the mylar half of this comparison, with the absorber already matched to the bag.