How long does coffee stay hot in a vacuum mug?
Six hours, twelve hours, twenty-four: the numbers on the box are real and they are also nearly useless without the conditions behind them. What the physics actually does, what the figures assume, and how to get a truthful answer for your own mug in an afternoon.
By Marta Iken · published 24 August 2026 · updated 24 August 2026
Every insulated vessel on sale carries a pair of numbers, and every one of them is technically true and practically slippery. Six hours hot. Twenty-four hours cold. Nobody prints the room temperature they used, the fill level, whether the lid was on from the first second, or, the one that changes everything, what temperature counts as still hot.
This is not an accusation. There is no agreed test in this category the way there is for, say, a fridge. What follows is what the numbers depend on, what Ideus states for its own range, and how to measure your own in an afternoon with a kitchen thermometer, which is the only figure that will ever describe your actual morning.
Vacuum does not keep anything hot
The first thing worth unlearning: a vacuum wall does not add heat. It slows the escape of the heat already in the drink. That sounds pedantic until you notice what it implies, everything about how long a drink stays hot is decided by how much heat went in and how fast it can get out, and the flask only controls the second half.
Heat leaves a vessel three ways. Conduction through the wall, which is what the vacuum nearly eliminates: with no air between the two skins of steel there is almost nothing to conduct through, and the only bridge left is the rim where the two skins meet. Convection, which is what the lid controls: an open drinking hole is a chimney, and warm air leaving the top pulls colder air in. Radiation, which is what the mirrored inner surface reduces.
The rim and the lid are where the real differences between vessels live. Two flasks with identical vacuum walls can differ by hours if one has a wide mouth and a loose lid and the other has a narrow mouth and a gasket.
Mass is the quiet variable
A 40 oz bottle beats a 16 oz mug by hours, and not because the steel is better. It has more than twice the drink in it, while the surface it loses heat through has not grown at anything like the same rate. Double the volume and the surface area rises by roughly a factor of one and a half. More heat, proportionally less exit.
The same maths runs in reverse on the same vessel: a half-full flask behaves like a smaller one, and it behaves worse than that because the air gap above the drink is now big enough to convect. If you want the number on the box, fill the thing.
- Full beats half by a wide margin, on the same vessel.
- Big beats small at the same fill level.
- A narrow mouth beats a wide mouth, and a stein, all mouth, states the shortest hours in the whole range.
- Preheating with boiling water for two minutes buys back a real fraction of the first hour, because the steel itself starts warm instead of stealing heat from the coffee.
What Ideus states, and what it does not
The figures below are the maker's own, read from its listings. They are not measurements of ours and this site does not present them as such.
| Piece | Hot | Cold |
|---|---|---|
| 50 oz thermal carafe | up to 12 h | up to 24 h |
| 68 oz thermal carafe | up to 12 h | up to 24 h |
| 68 oz thermal carafe with a pour-over dripper | up to 12 h | up to 24 h |
| 68 oz thermal carafe with a 74 oz airpot dispenser | up to 12 h | up to 24 h |
| 68 oz thermal carafe with a 20 oz soup container | up to 12 h | up to 24 h |
| 24/34 oz insulated beer stein | up to 1 h | up to 5 h |
| 14 oz insulated bottle | up to 12 h | not stated |
| 24 oz insulated bottle | not stated | up to 24 h |
| 32 oz insulated bottle with a straw and two lids | up to 12 h | up to 24 h |
| 32 oz insulated bottle and two lids | up to 12 h | up to 24 h |
| 40 oz insulated bottle and two lids | up to 12 h | up to 24 h |
| 14 oz tumbler with a straw | up to 6 h | up to 12 h |
Two things stand out when you lay them side by side. The first is that the range is internally consistent, the biggest vessels carry the biggest numbers, exactly as the physics predicts, which is a small sign that nobody is inventing figures. The second is the stein: an hour warm, five hours cold. That is a modest number and the maker printed it anyway, which is worth more than a bold one.
What no listing here states is the condition behind the figure. So read them as a comparison inside one range, which is the job they can honestly do, and not as a promise about your Tuesday.
Measure your own in one afternoon
This takes a kitchen thermometer, a kettle and about six hours of leaving something alone. It is the only way to get a number that describes your mug, your room and your lid habits.
- Preheat: fill the vessel with boiling water, put the lid on, wait two minutes, tip it out. Skipping this step is the biggest single source of disagreement between two people testing the same mug.
- Fill to the level you actually use, most people fill to about an inch below the rim, not to the brim.
- Record the starting temperature with the thermometer in the middle of the liquid, not against the wall.
- Close the lid the way you close it in real life. If you leave the sliding hatch open at your desk, leave it open for the test. The result will be worse and it will be your result.
- Read at one, two, four and six hours, opening the lid for as little as possible each time. Note the room temperature too, a 19 °C kitchen and a 4 °C car are different experiments.
The number people care about is when the drink drops below the temperature they will still drink it at. For most coffee drinkers that is somewhere near 55 °C, which is a long way above cold and usually arrives long before the vessel has stopped being technically warm. That personal threshold is what makes one person's four-hour mug another person's two-hour mug.
What the curve actually looks like
Heat loss is not a straight line, and knowing its shape changes what you expect. A hot drink in a vacuum flask falls fastest in the first thirty minutes and then settles into a long, shallow decline. That first drop is not the vacuum failing. It is the drink giving heat to the steel, the lid and the trapped air, all of which started at room temperature.
This is why preheating matters more than any other single habit. Two minutes of boiling water tipped out before you fill costs nothing and removes most of that first drop, because the steel is no longer a cold body waiting to take heat from your coffee. In a small mug the difference is the difference between drinkable at eleven and disappointing at ten.
It also explains a complaint that looks like a fault. People fill a 40 oz bottle a third full with ice water, find it lukewarm by mid-afternoon and conclude the insulation is poor. A third full means two thirds air, and air convects: the drink is cooling into a warm chimney above it. Same bottle, same vacuum, entirely different result.
Hot and cold are not symmetrical
Every listing in this category quotes a bigger number for cold than for hot, and it is not marketing. Room temperature is around 20 °C. Coffee starts near 90 °C, so it is fighting a 70-degree gap. Iced water starts near 2 °C and fights an 18-degree gap. The cold drink has a quarter of the pressure on it, and it also has an ally: melting ice absorbs a great deal of heat without changing temperature at all.
That last point is worth using. A bottle filled with ice and water stays at drinking-cold until the last ice melts, then warms quickly. Add more ice than looks sensible and the number on the box becomes conservative rather than optimistic. It is the one case where a buyer can beat the maker's figure without buying a better flask.
In the other direction there is no trick. Nothing you can put in a mug holds heat the way ice holds cold, and once a drink has fallen below drinking temperature, no vessel brings it back.
The lid decides more than the wall
Two vessels with the same vacuum wall can differ by hours, and the difference is almost always the opening. Convection through an open drinking hole moves heat far faster than conduction through a rim, and that is why a mug drunk from with the hatch open all morning behaves like an uninsulated cup with a good handle.
Mouth width matters for the same reason even when the lid is closed: a wide mouth means a wide column of trapped air, and trapped air moves. This is the physical reason the stein in this range states the shortest hours of anything on the site. It is not a lesser product. It is a shape with a large opening and a job that lasts an hour, not a day.
| Habit | What it costs you | Why |
|---|---|---|
| Not preheating | Most of the first hour | The steel takes its share of heat before the drink keeps any |
| Filling half way | An hour or more | Trapped air above the drink convects |
| Leaving the hatch open at a desk | Hours | An open hole is a chimney |
| Opening to check every twenty minutes | Half an hour, cumulatively | Each opening swaps warm air for cold |
| Leaving it in a cold car | An hour or more | The gap the vacuum is fighting has doubled |
Two figures that get confused
There is a difference between still hot and still safe, and listings blur it. Hot is a taste question and lands somewhere around 55 °C for most people. Food safety is a different threshold entirely, and it matters for anything with milk or a broth in it: a vacuum flask that keeps a drink pleasantly warm for six hours has also kept it in a temperature range where bacteria are comfortable.
For coffee, tea and water this is academic. For soup, milky drinks or anything you would refrigerate at home, the honest advice is the boring one: fill it very hot, drink it within a few hours, and do not treat a flask as a lunchbox that can sit out until evening.
This site does not publish food-safety guidance beyond that, because we have no expertise to add and the topic deserves a source better than a drinkware catalogue.
Comparing brands honestly
If you are cross-shopping, the only fair comparison is between figures measured the same way, and outside a laboratory nobody can promise you that. What you can do is compare shapes and sizes, which the physics makes predictable: bigger holds longer, narrower holds longer, sealed holds longer, and a preheated vessel beats a cold one of the same design.
Where two products differ only in the number printed on the box, treat the difference as marketing rather than engineering. Where they differ in mouth width, wall thickness or lid type, believe the difference.
What to do with all this when buying
If your question is whether the coffee will survive a meeting, almost anything with a vacuum wall and a closed lid will do it, and the choice comes down to shape, handle and colour. If your question is whether water will still be cold at four in the afternoon, buy volume, a 32 or 40 oz bottle, filled, with ice, will do things a mug simply cannot.
And if you drink slowly with the hatch open all morning, no flask on the market gives you the number on the box, including this one. That is not a fault in the product. It is a chimney you opened yourself, and the fix costs nothing.