Is a mini-split worth it?

The mini-split question usually arrives framed as an efficiency argument, and the efficiency argument is usually made with the wrong pair of numbers.

You will see SEER2 33.5 held up against a window unit’s CEER 17.6 and concluded that a mini-split is nearly twice as efficient. It is not. Those two ratings are not measured the same way, and comparing them is the single most common error in this category.

SEER2 is a seasonal figure. It averages performance across a modeled cooling season including long stretches of part-load running, where a variable-speed compressor is at its most efficient.

EER2 is the full-load figure at a single rating condition. CEER, the room air conditioner rating, is much closer to that in method — it is a combined figure at a rating point, with standby included.

Read the third row carefully, because it is the one nobody prints. The median certified mini-split is EER2 12.50. The median certified room air conditioner is CEER 15.0. On the comparable basis, a typical mini-split is not more efficient at cooling than a typical window unit — the best ones are, by single digits.

None of that makes SEER2 a lie. Over a real season with long part-load hours, a mini-split does better than its EER2 suggests. But if someone is selling you a mini-split on a cooling-efficiency argument, ask for the EER2.

This is where the category earns its price, and it is barely about summer at all.

• Median COP of 2.00 at 5°F across 16,667 certified records that publish one. Two units of heat per unit of electricity, in genuinely cold weather. The best in the field is 3.28.
• Median 85% of rated capacity retained at 5°F against the 47°F rating. The old objection that heat pumps collapse in the cold is measurably out of date on certified equipment.
• 14,229 of 16,792 US records are cold-climate certified — about 85% of the field, not a niche.

Electric resistance heat is fixed at a COP of 1 and cannot be improved by any manufacturer. A COP of 2.0 halves the electricity for the same heat. If your room is currently heated by baseboards or a fan heater, that is the number that justifies an installer — how heat pumps and air conditioners actually differ.

The dataset holds 204 outdoor-unit brand names across the US market, of which 64 carry 100 or more certified models. Across all 64, the median EER2 runs from 11.85 at the bottom to 13.00 at the top.

That is a spread of under 10% between the least and most efficient brand, while the spread within a single brand — from its entry model to its flagship — is several times larger. Choosing a brand is not how you choose efficiency in this category. Choosing a model is.

The same machine also appears repeatedly under different labels. One 18,000 BTU unit on this page is certified under at least five separate badge records. Contract manufacturing is the norm here, not the exception.

Plainly: we sell equipment, not installations, and we are not going to quote you an installed price we cannot stand behind. A mini-split is an installed system involving refrigerant lines, a wall penetration, a condensate route, an outdoor pad or bracket and usually a dedicated circuit. The equipment is a minority of the total on most jobs.

What we can tell you is the shape of the decision:

1. Cooling only, one room, rented or temporary. A window or portable unit wins on nearly every axis, and the efficiency gap on a fair comparison is single digits.
2. Cooling only, one room, owned and permanent. Closer. A mini-split is quieter, leaves the window free, and does not come out every autumn. That is a comfort argument, not an energy one.
3. Cooling and heating, replacing resistance heat. This is where it makes sense. You are halving the electricity for winter heat, every winter, which is a far bigger annual number than the cooling difference.
4. Multiple rooms. Multi-zone changes the arithmetic again, in the mini-split’s favor, because the install cost is shared — cooling a house against cooling a room.