2025 Heat Pump Model Showdown
2025 equipment models come with increases in efficiency and new refrigerants.
January 2026 update: after a complete installation season, here are our thoughts on each one of these brands from an installation standpoint, accounting for their quirks, and issues. You can read about that below.
As you probably know, new heat pump models with R-410A refrigerant are phasing out of the market, with newer R-32 and R-454B models taking their place. We’ve been keeping an eye on them at Laminar Collective.
But today, it’s time. Let’s compare every single cold-climate model we can find!
Summary
The good news? A lot of these models are just better. Every single model that we review is a cold climate model that retains the vast majority of its heating output at 5°F, which is important b/c we are a Cambridge/Boston-based research team.
Performance: Most brands have notably better specs showcasing higher capacity ranges and efficiencies both for cooling and heating
Cooling: SEER2 ratings look great compared to to R-410A equivalents
Winners: For ductless & mixed, Mitsubishi on top, but LG is surprise #2. Everyone else is pretty good. No clear leader for unitary ducted.
Caveat: We do not factor in cost or supply house until 2nd half of article.
Refrigerant: R-454B has supply constraints, R-32 more available
Want to see the spreadsheet yourself and get access to all the product links? Go here.
We use SEER2 & HSPF2 efficiency ratings to rank the models, and you can read more about that in our footnote here1. Again, we factor in cost & supply chain in Pt. 2.
Ductless: Major Improvements over R-410A
Mitsubishi, LG, & Midea top the rankings.
Update: LG is actually quite annoying to install, so that’s one significant downside. You can read about it here at our January 2026 update.
Commonly used for triple deckers, 1st floor of 2 family buildings, and cape-styled houses, a ductless/mini-split only setup is the most efficient system you can get. The air feels cooler due to lower discharge temperature, which is particularly useful for heat waves. The downside, of course, is that you need wall space.
But importantly, every single model out right now blows their prior model equivalent out of the water in SEER2 & HSPF2 performance:
With Mitsubishi’s minimum heating BTU finally coming down to 12,600 in the R-454B model, everyone’s minimum heating BTU is now at a level that should help a lot to prevent inefficient short-cycling & oversizing.
Mixed: Mostly the Same as Ductless
Midea top the rankings based on SEER2, but we think heating is slightly more important in MA so Mitsubishi pulls ahead with a 12.1 HSPF2. LG is okay.
This setup is pretty common in colonial-styled single family homes with open attics/ 1st floors. While discharge temperature on the 2nd floor isn’t as cool/hot as with mini-splits, it does free up wall space. Be warned, though: this requires an empty attic, and becomes somewhat more difficult to install if you want the attic finished.
I think for us, because LG is pulling ahead on HSPF2 for ductless installations, it’s just about even between LG & Midea right now.
Ducted: Fujitsu Pulls Ahead
Note: Midea ducted is hard to find, since we only know of 1 supply house that carries it directly. Otherwise, I haven’t seen any white-labeled version (e.g. Carrier) cold-climate heat pumps in the field, so I can’t attest to their cold-climate performance. You can read about it here at our January 2026 update.
Fujitsu’s standard model looks good (e.g. non-unitary ducted) model looks good. By unitary, we mean condensers that are made specifically to be compatible with an indoor air handler unit, and usually nothing else.
It’s important to note that every model is better than a 2-ton SVZ model from Mitsubishi’s prior generation, and roughly on par with Mitsubishi’s R-410A PUZ model.
We typically do not recommend fully ducted systems unless you already have ductwork in place for a house. It is less efficient due to the heat loss in the ductwork, which is why all these SEER2 & HSPF2 ratings are lower. If you already have ductwork & it’s good condition, however, this is a viable option.
Shockingly, Mitsubishi’s vaunted light commercial PUZ/PVA setup doesn’t top the chart at the 2-ton level. And LG clocking in with a 17.5 SEER2 isn’t great.
Mitsubishi’s residential SUZ/SVZ setup isn’t that good, either, though substantially better at a SEER2 of 18.6 vs. the 16.0 of the prior generation:
Comparatively, Daikin looks real good with a 19.2 SEER2 rating, and Bosch, Midea + LG offer an HSPF2 of 10.
Here’s a direct comparison of LG and Mitsubishi’s 3.5 ton units on AHRI:
This seems to validate some early trends we’ve been spotting in this space:
Distribution Networks & Refrigerant
So far, we have been ranking purely based on efficiency ratings. But there are two other important considerations as you make a decision about which brand to install.
Distribution Networks
Every brand typically has a local supply house or distributor. Either that, or a contractor is large enough to directly procure from the manufacturer. For example, acquired ABCO, and now has a network of New England distribution centers, including one very conveniently in Somerville:
Meanwhile, due to Mitsubishi, early market entry & market share, Johnstone Supply, Ferguson & SG Torrice all carry Mitsubishi equipment.
Carrier & Bryant’s local distributor is the extremely generically named Distributor Corporation Of New England. That is like, an ACME Corp. kind of name:
This is all very important because Midea, as far as we can tell, does not have a direct supply house in the region. Therefore, we don't know exactly how you can get spare parts on the fly for Midea installations. So this shifts the advantage back towards the other models that do have high efficiency scores that also have a robust distribution network.
Refrigerant Availability
There is currently significant supply constraints for R-454B, both for the refrigerant and the canisters. Anecdotally, we have talked to local supply houses in Somerville that mentioned that R-454B is twice as expensive as R-32 right now.
This primarily impacts larger homes that need extra refrigerant, but in general, this is a challenging summer to procure R-454B if there are leaks, etc.
Because of this, we at Laminar Collective are slightly more inclined to use a model with R-32. We’ll be publishing more extensive notes on this situation in a separate article.
The Cost Factor
Most models in this generation perform as well as Mitsubishi’s top of the line products from the previous generation.
This opens the possibility for substantial cost savings because a lot of non-Mitsubishi brands are priced more affordably.
Prices at supply houses are notoriously murky, but in general, Mitsubishi is the most expensive and everyone else follows. Given the fact that you have multiple strong alternatives, we’re anticipating a lot more competition in 2026 and 2027.
Where did we find this product information?
We got most of the product information through looking at product submittals and engineering spec sheets. These are easier to get for some brands more than others.
For example, here’s Mitsubishi’s redesigned submittal/engineering documentation database:
It’s like they actually want you to find information about their products. Wow!
Then, there’s the topic of how the submittals are actually formatted. Midea’s sheets, for example, are very clear and easy to understand:
Carrier’s submittals on the other hand were very difficult to find and we ended up having to get a couple of them emailed from a supply house after making several phone calls. As a matter of fact, they don't even have submittals. They only have gigantic engineering documents that we had to really look through to find performance stats.
Conclusion: The Market Gets More Competitive
Five years ago, Mitsubishi had the only good cold climate game in town, and thank goodness it's starting to change. The market activity and continued investment into cold climate models mean that multiple viable competitors are now competing in the New England market. This is good for consumers.
In a way, this is the optimal outcome of heat pump incentives. Even when they’re gone, consumers are better off. The summers are just going to get hotter.
There's going to be a continued business case for investing in these models, because people will want air-conditioning during the summer. And if you can get 2x the capabilities of an AC for only 25% more of the cost, why wouldn't you do it?
There are two key ratios we look at to judge product efficiency and performance. Every HVAC device is rated for both of these. For both higher numbers are better.
SEER2 - Seasonal Energy Efficiency Ratio 2. It’s how effectively a device cools during the summer. HSPF2 - Heating Seasonal Performance Factor 2. It’s how effective a device is at heating is during the cold months.
Another good source for heat pump specs is AHRI, an independent testing agency. There are occasional differences between manufacturer's specs and what AHRI has found with their testing.
Good review! Just a comment about this statement:
It is less efficient due to the heat loss in the ductwork,
Although possibly not common in the Boston area, homes built with all ductwork within the conditioned space, like a ranch with conditioned basement or a bi-level home, have zero energy loss through ductwork. We own such a bi-level home. Chase in the lower level/basement ceiling for main trunk and branch ducts between floor joists to registers in lower level ceiling or upper level floor.