What the certified numbers actually show
Manufacturers publish AHRI-certified heating capacity at three outdoor temperatures - 47F, 17F and 5F. The 47F figure is the one that becomes the marketing number. Here is how the other two compare across our catalog:
| 5F output as a share of the 47F rating | Systems |
|---|---|
| Under 50% | 22 |
| Under 70% | 177 |
| Median | 77% |
| 100% or more | 212 |
So the typical system's certified 5F output is about three-quarters of its 47F figure. But the distribution is what matters, because roughly one system in five is certified higher at 5F than at 47F, and 22 come in under half.
Read that ratio as a rough sorting tool, not as a physical fade curve - the next section explains why the two ends of the range mean different things.
The units that gain capacity in the cold
This surprises people, and it is not an error - we audited it against the saved AHRI records specifically because it looked wrong.
The strongest examples are all nominally 6,000 BTU systems rated at 9,000 BTU of heating at 47F and 13,500 BTU at 5F - 150% - including the Midea Elevate Max, the Carrier Infinity 35 and the Blueridge XS6 Max Efficiency. GREE's Sapphire 9,000 BTU reaches 144%, and Fujitsu's AIRSTAGE Orion XLTH+ 9,000 BTU hits 142%.
The reason is that these are variable-capacity inverter systems, and the three certified points are separate rated conditions rather than three readings from one continuous curve. An inverter has no single fixed output - it modulates across a wide range, and the low-temperature ratings are not all taken at the same compressor speed as the 47F rating. So a system can legitimately publish a larger number at 5F than at 47F without anything being wrong.
You can see it clearly in Fujitsu's AIRSTAGE Orion XLTH+ at 9,000 BTU: 12,000 BTU at 47F, dipping to 7,400 at 17F, then back up to 17,000 at 5F. That is not a machine getting stronger as it gets colder. It is three certified measurements taken under three different sets of conditions.
The practical reading: treat each figure as "what this system is certified to deliver at that temperature," not as a percentage of the headline number. The cooling BTU on the box is a poor guide to heating output, and a poorer one still to output in the cold.
The practical consequence is a happy one: a small, well-chosen cold-climate unit can carry far more winter load than its nameplate implies. It is also why "buy bigger for winter" is often the wrong instinct - see how to size a mini split.
The units that fall off a cliff
The other tail is the one to watch for. Della's 18,000 BTU Motto and Umbra are rated 18,700 BTU at 47F and 3,600 BTU at 5F - 19%. The 12,000 BTU versions run 12,300 down to 2,400, about 20%.
Those are real published figures, not transcription errors; they are in the certified records. A unit like that is a perfectly reasonable cooling system with heating as a shoulder-season bonus. It is not a winter heating system, and sizing one to a January design load using the 47F number would leave a room cold by a factor of five.
This is the single most expensive mistake in cold-climate mini split sizing, and it is invisible if you only read the headline capacity.
How to use this when you buy
- Find your design temperature. That is the outdoor temperature your area actually reaches on a cold winter night, not the record low. Our state cost pages list climate context.
- Read the capacity at the temperature closest to it, not at 47F. If your design temperature is near 5F, the 5F number is your real capacity.
- Compare that against your heat loss, not against the room's cooling load. They are different numbers, and in a cold climate the heating load usually governs.
- Check the low-ambient rating too. Capacity tells you how much heat you get; the low-ambient heating rating tells you how cold it can get before the system stops being rated at all. A unit rated to -22F with modest capacity may serve you better than a stronger one rated only to +5F.
- Be sceptical of a heating number quoted without a temperature. It is almost always the 47F figure.
Why this is not on most spec sheets
Retail listings quote one BTU number because one number is easier to shop. The three-point certified data exists for every system in this comparison, but it usually sits in an AHRI record or a submittal rather than on the product page.
That gap is precisely why two systems that look identical on a shelf can differ by a factor of eight in the only conditions that matter for heating. Every model page in our catalog publishes the certified figures at 47F, 17F and 5F where the manufacturer has them, so you can check yours directly.
Look up your system in the product finder, or start with the systems built for this in best cold-climate mini splits. If you are weighing a mini split against another heat source entirely, cold weather performance covers the wider question.