The best window air conditioner brands, tested against the certified data

“Which brand should I buy?” is one of the most common questions in this category and one of the least useful, because the certified data says the badge is frequently the least informative thing on the box.

We have not tested any of these units and we do not hold reliability data. What we have is the ENERGY STAR Certified Room Air Conditioners dataset, which records every certified model with figures the manufacturer cannot write itself. We grouped all 485 records by the brand_name field and ranked on four things:

1. How many certified models the brand actually files. A brand with two models and one good one is not a track record.
2. Median CEER across its range. The typical unit you get if you buy the badge blind, rather than its best.
3. Best CEER. What the brand is capable of when it tries.
4. Share of its range on the Most Efficient list. The strictest tier in the program — only 171 of 485 room units reach it.

We deliberately did not rank on price, warranty length or reputation. Price we cannot verify on the day, and reputation is what we are testing rather than assuming.

Before the rankings, the thing that changes how you should read them.

We grouped every record by an exact match on capacity, CEER, annual kWh, all three dimensions and weight, then kept only the groups spanning more than one brand name. There are 40 such groups, and they contain 282 of the 485 certified units — 58.1% of the field. 38 of the 47 brand names appear in at least one.

The largest single group is worth seeing in full:

Nine brand names, one set of numbers. These are not similar units; every published figure matches to the decimal, including the weight.

That does not make any of them a bad buy — a CEER 15 unit at 400 kWh a year is a perfectly reasonable machine. It does mean that comparing those nine on specification is comparing a machine to itself, and that whichever of them is cheapest on the day is the correct one to buy.

Three things stand out. The highest Most Efficient rates belong to the smallest lines — Kuhl, Brothers and Seasons file eight to ten models each and certify almost all of them, because a narrow premium line is easier to keep at the top. The volume brands are mixed: Midea and TCL both have a median of CEER 15 and reach the top of the field only with specific models. And eleven brands with six or more models have none at all.

Note also how flat the median column is. Sixteen of these twenty-one brands sit at exactly CEER 15, which is what you would expect if a large part of the market is buying the same platforms.

This is the single most useful thing on the page for anyone shopping by badge, because all three names appear on shelves and they perform nothing alike.

• GE — 3 models, median CEER 16, 3 of 3 Most Efficient.
• GE Profile — 9 models, median CEER 15, best 15.7, 0 of 9 Most Efficient.
• GE Appliances — 2 models, CEER 13.2, 0 of 2.

“Profile” is the premium-sounding badge, and in the certified data it is the weaker set — nine models, none of which clear the Most Efficient bar, against three plain-GE models that all do.

Sub-brands are marketing tiers, not efficiency tiers. The lesson generalizes: whenever a manufacturer runs several badges, check which one the specific model is filed under rather than assuming the fancier name is better.

Friedrich’s Kuhl range gives an unusually clean natural experiment. Nine model families appear twice in the dataset, with model numbers differing only in a trailing A or B — a revision, not a redesign.

The B revision is more efficient in all nine of nine families, and in seven of the nine it is the difference between clearing the Most Efficient bar and not clearing it. On KCVS12B30 that is CEER 15.0 against 16.8, and 68 kWh a year — from one letter.

Both revisions are sold. Both look like “a Friedrich Kuhl 12,700 BTU” in a listing. The efficiency difference is invisible unless you read the full model number, which is the strongest practical argument on this page for shopping by model rather than by name.

Having spent the page arguing the badge is weak, it is worth being fair about where it is not.

• Which platform you are getting. Midea holds 184 certified records and 100 of the Most Efficient units. If you buy almost anything in this market you are quite likely buying its engineering, whatever the sticker says.
• Whether a premium line exists. Kuhl at 8 of 8 and plain Friedrich at 6 of 28 is a real distinction, even though both are the same company.
• Distribution and availability. The nine brands in that top identical-spec group exist largely because retailers want their own label. That is genuinely useful to you — it means the same machine is available at several prices.
• Warranty and parts. Not in the dataset, and not uniform. Read the terms before you decide the cheapest identical unit is the right one.
• What it does not tell you: efficiency. That is on the model number, and the specification sheet reads the same regardless of who printed it.

1. Size the room first. Capacity dominates running cost far more than brand does — work out what you need before you look at a single badge.
2. Write down the full model number, including any trailing letter. That letter is worth more than the brand.
3. Check it against the certified list. CEER and annual kWh are published per model. If your candidate is not on it, that is itself information.
4. If two units have identical specs, buy the cheaper. There are 282 units in the field where this applies, so it comes up often.
5. Then weigh the things we cannot measure — warranty, parts, service — and let those break the tie.
6. Check the install before you commit. Weight, bracket requirement and whether it straddles the sill decide more about your weekend than the badge does.