Compare yearly heating cost on your own gas and electricity rates. The rates decide it, not the slogan.
$758 vs $1,185furnace vs heat pump, per year on the defaults60 million Btu of heat. Furnace: load ÷ AFUE ÷ 100,000 Btu per therm × gas price. Heat pump: load ÷ HSPF2 ÷ 1,000 × kWh price.
The short answer
It depends on your two rates. On the defaults (95% AFUE furnace at $1.20 a therm, HSPF2 8.1 heat pump at 16¢/kWh, 60 million Btu a year), the furnace is cheaper by $427 a year.
At average US prices, a heat pump needs about HSPF2 11.2 to beat a condensing gas furnace. Against oil, propane or electric resistance, the heat pump usually wins.
Compare running costs
Your yearly heating load, two ratings and two rates. Nothing is emailed.
Seasonal running cost, not cold-day capacity. Dual fuel is not modeled. No federal credit: §25C ended for property placed in service after December 31, 2025. HyreHVAC does not sell equipment.
What HSPF2 does a heat pump need to beat a gas furnace?
The HSPF2 a heat pump needs to beat a 95% AFUE gas furnace on running cost, 2024. Darker is harder.HyreHVAC calculation from EIA-861 electricity prices and EIA Natural Gas Navigator series N3010. Read the electric vs gas price study.
About HSPF2 11.2 at US average prices. Every heat pump proposal shows an HSPF2 number, but almost nobody says what it needs to be. The calculator answers that for your two rates; this section answers it for your state, from EIA’s 2024 residential prices.
HSPF2 is Btu of heat per watt-hour, and a watt-hour is 3.412 Btu, so seasonal COP = HSPF2 ÷ 3.412. A heat pump beats a gas furnace when its COP beats the furnace efficiency times the ratio of the two fuel prices. Multiply that back by 3.412 for the HSPF2 you need.
The median cold-climate model in the ENERGY STAR database is HSPF2 9.8, and the 90th percentile 11.1. At average US prices, the heat pump you need is above the 90th percentile of the market.
HyreHVAC analysis: that does not make heat pumps a bad buy. Running cost is one factor, and the picture changes completely against oil, propane or resistance heat. It means a running-cost claim needs checking against your state.
Which three numbers decide it?
11.2HSPF2 needed at US average pricesAgainst a 95% AFUE furnace, 2024
9.8median HSPF2 of certified cold-climate models9,271 distinct outdoor units
5 of 51jurisdictions where no model on the market clears the barThe best certified HSPF2 is 16.2
In which states does a heat pump beat gas?
Easy: 20 jurisdictions need HSPF2 9.8 or less, which the median cold-climate model already delivers. They include WV, KY, OR, VA, WA, UT, HI, OK: mostly southern states, where gas is pricey because few homes use it.
Out of reach on gas: 5 jurisdictions need more than HSPF2 16.2, the highest rating in the ENERGY STAR database. In AK, MI, CT, WI, CA, no heat pump made today beats a condensing gas furnace on running cost at 2024 average prices.
That is about the price of gas, not cold-weather capacity. Those heat pumps still heat the house well.
The big caveat: all of this assumes the house burns natural gas now. Against oil, propane or electric resistance, the heat pump wins nearly everywhere, and against resistance always. About half of US homes do not heat with gas.
What HSPF2 does your state require?
State
Electricity ¢/kWh
Gas $/Mcf
Breakeven COP
HSPF2 needed
Available today?
United States
16.48
14.50
3.28
11.2
Only at the top of the market
AK
24.82
11.73
6.11
20.8
No model on the market clears it
MI
19.30
10.76
5.18
17.7
No model on the market clears it
CT
28.75
16.83
4.93
16.8
No model on the market clears it
WI
17.18
10.14
4.89
16.7
No model on the market clears it
CA
31.97
19.14
4.82
16.4
No model on the market clears it
NM
14.20
9.05
4.53
15.5
Only at the very top of the market
IN
14.77
9.50
4.49
15.3
Only at the very top of the market
MT
12.66
8.22
4.45
15.2
Only at the very top of the market
NY
24.43
16.58
4.25
14.5
Only at the very top of the market
NJ
19.34
13.21
4.23
14.4
Only at the very top of the market
MN
15.45
10.65
4.19
14.3
Only at the very top of the market
IL
15.87
11.11
4.12
14.1
Only at the very top of the market
CO
14.92
10.58
4.07
13.9
Only at the very top of the market
SD
12.86
9.32
3.98
13.6
Only at the very top of the market
ID
11.52
8.51
3.91
13.3
Only at the very top of the market
MA
29.35
21.90
3.87
13.2
Only at the very top of the market
RI
28.65
21.66
3.82
13.0
Only at the very top of the market
PA
17.77
13.62
3.77
12.9
Only at the very top of the market
IA
13.40
10.33
3.75
12.8
Only at the very top of the market
ME
24.29
18.95
3.70
12.6
Only at the very top of the market
ND
11.51
8.99
3.70
12.6
Only at the very top of the market
VT
21.90
17.34
3.65
12.5
Only at the very top of the market
NH
23.40
18.74
3.61
12.3
Only at the very top of the market
OH
15.99
13.77
3.35
11.4
Only at the very top of the market
TN
12.42
10.75
3.34
11.4
Only at the very top of the market
DE
16.57
14.78
3.24
11.1
Yes: needs a good model
MD
17.86
16.14
3.19
10.9
Yes: needs a good model
NE
11.53
10.63
3.13
10.7
Yes: needs a good model
KS
14.15
13.14
3.11
10.6
Yes: needs a good model
DC
17.71
16.49
3.10
10.6
Yes: needs a good model
WY
12.47
12.39
2.91
9.9
Yes: needs a good model
WV
15.07
15.26
2.85
9.7
Yes: a typical model clears it
KY
12.79
13.61
2.71
9.2
Yes: a typical model clears it
OR
14.70
15.71
2.70
9.2
Yes: a typical model clears it
VA
14.41
15.40
2.70
9.2
Yes: a typical model clears it
WA
11.90
12.72
2.70
9.2
Yes: a typical model clears it
UT
12.22
13.15
2.68
9.1
Yes: a typical model clears it
HI
42.86
48.88
2.53
8.6
Yes: a typical model clears it
OK
12.24
13.99
2.53
8.6
Yes: a typical model clears it
MS
13.39
15.42
2.51
8.6
Yes: a typical model clears it
NC
14.13
16.42
2.49
8.5
Yes: a typical model clears it
SC
14.23
16.91
2.43
8.3
Yes: a typical model clears it
TX
14.94
17.98
2.40
8.2
Yes: a typical model clears it
AL
15.18
18.63
2.35
8.0
Yes: a typical model clears it
NV
15.00
18.69
2.32
7.9
Yes: a typical model clears it
GA
14.08
18.19
2.24
7.6
Yes: a typical model clears it
AR
12.32
16.09
2.21
7.5
Yes: a typical model clears it
MO
12.91
16.86
2.21
7.5
Yes: a typical model clears it
AZ
14.91
20.47
2.10
7.2
Yes: a typical model clears it
LA
11.73
16.38
2.07
7.1
Yes: a typical model clears it
FL
14.14
25.37
1.61
5.5
Yes: a typical model clears it
2024 residential fuel prices and the HSPF2 they require, from most to least favorable for the heat pump. Assumes a 95% AFUE condensing furnace.
For an 80% AFUE furnace, multiply the HSPF2 column by 0.842. Against oil, propane or resistance heat the bar is far lower and not shown. See the full price-ratio study.
What doesn’t this calculator model?
Capacity on the coldest day
Running cost and holding the house at −5°F are different questions. Cold-climate models keep a median 83.3% of their 47°F capacity at 5°F, so capacity is largely solved. Efficiency is not: the median COP at 5°F is 1.97.
Backup heat
When a heat pump cannot keep up, electric resistance strips take over at a COP of 1. Those hours cost full electricity price, and a seasonal average hides them.
Dual fuel
A dual-fuel system runs the heat pump above a set temperature and the furnace below it, using the cheaper fuel at every temperature. That beats both columns here and is not modeled.
Your actual rate versus the average
The state table uses average revenue per unit, fixed charges included. If your utility has a heat pump or time-of-use rate, enter that instead.
The AC you may be replacing anyway
A heat pump is an AC that runs in reverse. If the AC is due anyway, compare the extra cost of a heat pump over a new AC and furnace, not the full price.
Federal tax credit
§25C is $0 for property placed in service after December 31, 2025. Do not subtract $2,000 from the heat pump column. Utility rebates are separate.
What “high efficiency” is measured against
What counts as a high-efficiency heat pump?
It depends on the yardstick: the legal minimum, the ENERGY STAR label, or the market. They are far enough apart that one machine can honestly be called well above standard and below the middle of the market.
The HSPF2 ladder: the first five bars are thresholds, a legal minimum and four label criteria. The last four are the certified market, above all of them.Thresholds from 10 CFR 430.32(c)(5) and (c)(6) and the ENERGY STAR key product criteria. Market figures are a HyreHVAC calculation from ENERGY STAR certified product data, September 4, 2026. Read the cold-climate heat pump study.
The legal floor: under 10 CFR 430.32(c)(5) and (c)(6), a split heat pump made on or after January 1, 2023 must reach 14.3 SEER2 and 7.5 HSPF2. Heat pumps have no regional standard; the same minimum applies everywhere.
The label: ENERGY STAR asks for 7.8 HSPF2, 15.2 SEER2 and 11 EER2 on a split system, and 7.2 HSPF2, 15.2 SEER2 and 10 EER2 on single-package equipment.
Cold Climate:8.1 HSPF2 ducted or 8.5 non-ducted, plus a COP at 5°F of at least 1.75 and at least 70% of the 47°F heating capacity at 5°F.
HyreHVAC calculation: across the certified catalog, the median model is HSPF2 9.5, the 90th percentile 10.9, and the highest 16.2. The ENERGY STAR bar of 7.8 sits below even the 10th percentile (8.0).
HyreHVAC analysis: an ENERGY STAR label is a floor, not a ranking. Judge a quoted HSPF2 against the market, and then against the number your fuel prices demand, which the calculator above gives you.
A useful check: the lowest SEER2 in our certified data is 15.2, exactly the ENERGY STAR bar, and the lowest HSPF2 is 7.2, exactly the single-package bar. The catalog stops at the criteria, as it should.
What does each heat pump rating claim actually prove?
Three of these show up in marketing. One shows up only if you ask. The last column says what a missing claim tells you.
Criterion
What it requires
What it really tests
If it is missing
Federal minimum
14.3 SEER2 and 7.5 HSPF2 for a split heat pump made on or after January 1, 2023
Whether the unit may legally be made for the US market
Nothing is claimed. Every new split heat pump meets it
ENERGY STAR
7.8 HSPF2, 15.2 SEER2, 11 EER2, split system
A bar the whole certified catalog clears; our lowest SEER2 is 15.2, the bar exactly
The model is outside the certified catalog, below everything in the figures on this page
Cold Climate designation
8.1 HSPF2 ducted or 8.5 non-ducted, COP 1.75+ at 5°F, and 70% of the 47°F heating capacity at 5°F
Performance at a temperature you care about in January
Nobody tested it at 5°F to a published bar. In a cold climate that matters, and HSPF2 cannot fill the gap
Whether the quoted ratings belong to the system you are buying
An outdoor unit on its own does not carry a certified SEER2. A quote that names only a condenser model number has not yet told you what the system is rated at. Two quotes with the same outdoor unit are not always the same system
Capacity at your design temperature
A BTU-per-hour figure from the maker’s performance tables at your area’s design temperature
Whether it can hold your house on the coldest days without backup heat
The unit is named but not shown to meet the load. This decides how many hours you pay backup-heat rates
Five claims a heat pump proposal may make, and what each one proves.
HyreHVAC analysis: only the bottom two rows are about your house. The top three are about the product line, and those are the ones in the brochure.
Where the old objection is wrong, and where it is not
Do heat pumps work in cold weather?
Yes, for heating capacity: certified cold-climate models keep most of it. But they are not cheap to run in the deep cold. Those are two separate questions.
Capacity kept against the 47°F rating, 9,271 certified cold-climate outdoor units. The 10th-percentile and median models at two low temperatures. Lab ratings, not field measurements.HyreHVAC calculation from ENERGY STAR certified product data, September 4, 2026. Read the cold-climate heat pump study.
HyreHVAC calculation: across 9,271 certified cold-climate outdoor units, the median keeps 83% of its 47°F capacity at 5°F, and even the 10th percentile keeps 70%. At 17°F the median keeps 80%. Cold capacity is largely solved.
Efficiency is another story. The median COP at 5°F is 1.97, with the 10th percentile at 1.80. A COP near 2 is still twice the heat you pay for, good against resistance heat, but far below the seasonal figure this comparison uses.
HyreHVAC analysis: that is why the yearly number can be right and the January bill still a surprise. HSPF2 averages a season of mostly mild weather. It suits a yearly comparison, not the coldest two weeks, which is when people form their opinion.
What to ask: the heating capacity and COP at your design temperature, from the maker’s performance tables. Compare the capacity to your Manual J heating load. Any shortfall is covered by resistance backup at a COP of 1, the most expensive heat there is in a cheap-gas state.
These are certification ratings under test conditions. Real installs add defrost cycles, duct losses and control quirks, and each pushes the number down.
What do heat pump and furnace terms mean?
Most confusion in a heat pump conversation is about words, not facts. These eight carry the argument.
COP (coefficient of performance)
Heat delivered divided by energy used, in the same units. COP 3 means three units of heat per unit of electricity. Resistance heat is exactly 1, and no burner exceeds 1, which is why a heat pump can beat a 95% furnace on energy and still lose on money.
HSPF2
Heating Seasonal Performance Factor under the current test: Btu of heat per watt-hour across a modeled season. Divide by 3.412 for seasonal COP. It is an average, and averages hide the coldest weeks.
AFUE
Annual Fuel Utilization Efficiency: the share of fuel energy that becomes heat over a season. A 95% furnace uses about a fifth less fuel than an 80% one for the same heat.
The outdoor temperature where the heat pump’s capacity equals the house’s heat loss. Below it, something else must help. It depends on this machine in this house.
Switchover temperature
On a dual-fuel system, the outdoor temperature where the control stops the heat pump and starts the furnace. It is a setting that can be changed, and it decides which fuel you buy most of the winter.
Auxiliary or backup heat
Usually electric resistance strips at a COP of 1, covering the gap below the balance point. Those hours cost full electricity price, and a seasonal average will not show how many there were.
Emergency heat
A thermostat mode that skips the heat pump and runs backup only. It is for a failed compressor, not a cold snap. Left on through a cold week, it produces a bill people remember.
Defrost cycle
The outdoor coil ices up, and the unit briefly reverses to clear it, often firing backup heat. It is normal, counted in the rating, and why a heat pump sometimes blows cool air in winter.
Is a dual-fuel heat pump cheaper to run?
Usually, yes. The calculator compares two columns. A dual-fuel system uses whichever is cheaper at each outdoor temperature, which beats either one.
How it works: a heat pump and a gas furnace share the ducts. Above a set switchover temperature the heat pump heats; below it, the furnace does. Both are normal units with a control that knows about both.
Why it saves: a heat pump’s efficiency drops as it gets colder; a furnace’s does not. So there is a crossing temperature, set by your two rates, above which electricity is cheaper and below which gas is. Dual fuel stays on the cheap side all winter.
HyreHVAC analysis: so the state table, showing places where no heat pump beats a condensing furnace on yearly cost, is not the case against heat pumps it seems. It compares two single-fuel options. Dual fuel is neither.
The catch: you buy and maintain two heat sources. The saving must cover that over the equipment’s life. If your furnace is recent and sound, you only add the heat pump, and if the AC is due anyway, the extra cost is smaller still.
A free check: if you already have dual fuel, ask what the switchover temperature is set to. It is often left at the installer’s default, and it decides which fuel you buy most of the winter.
We do not calculate a switchover temperature for you. That needs your home’s heat-loss curve, the unit’s capacity and COP at each temperature, and your marginal rates. We have none of those.
What should a heat pump proposal include?
Each item is a number or a document, not a promise, because promises cannot be compared between quotes.
The AHRI reference number for the proposed pairing
A certified rating belongs to a specific combination of outdoor unit, indoor coil and, where present, air handler or furnace. Each certified combination carries its own AHRI Reference Number. Without it, the quoted ratings belong to some pairing, not necessarily yours.
Heating capacity at your design temperature, in Btu per hour
Not the nominal tonnage, which is a cooling rating in mild weather. This figure decides whether backup heat runs ten hours a winter or two hundred.
The calculated balance point
The outdoor temperature where capacity meets your heat loss. It needs a load calculation, so asking also tests whether one was done.
The backup heat size, in kilowatts
Backup runs at a COP of 1. Multiply the kilowatts by your electricity rate to see what an hour of it costs.
The switchover setting, if dual fuel
A temperature, and the reason for it. “The default” means it was not set from your rates.
Whether your electrical service can carry the backup
Resistance strips are a big electrical load. That is for the installer and electrician to confirm, not something to investigate yourself. See electrical panels and HVAC.
Do not accept a SEER vs SEER2 comparison
Since January 1, 2023, equipment is rated under 10 CFR part 430, subpart B, appendix M1, and the two metrics do not convert. A saving across them comes from the test change.
Do not accept a federal tax credit in the price
§25C is $0 for property placed in service after December 31, 2025. Utility rebates are separate and still exist, but they should be named with the program, not assumed.
What can this heat pump vs furnace tool not do?
Limits beyond the caveats above. These are the ones that could change which column wins.
It cannot find your yearly heating load
You supply it. The load does not change which column wins, but it sets the size of the gap entirely.
It cannot count hours on backup heat
Those hours cost full electricity price, and a seasonal average hides them. Counting them needs a balance point and local temperature data.
It uses only the rate you type
Tiered and time-of-use rates mean extra winter use may cost more or less than your average. If your utility has a heat pump rate, enter that.
It cannot compare installed prices
HyreHVAC does not install, service or sell HVAC equipment, and holds no data on quotes, invoices or finished jobs. Nothing here is a price we observed. That is also why these tools can tell you to keep the system you have.
It cannot tell what your ducts will do
A heat pump moves more air at a lower supply temperature than a furnace. Ducts sized for a furnace may not suit it, and that looks like the heat pump underperforming.
It can tell you to keep the furnace
In several states, no heat pump made today beats a condensing gas furnace on running cost at recent prices. We publish that result.
What would make a heat pump the cheaper choice?
Both prices move, and not together. Knowing what would flip your answer lasts longer than the answer.
The price ratio, not either price. A heat pump wins when electricity is cheap relative to gas. Both can rise while the ratio improves, so watching your gas bill alone tells you little.
The fuel you use now. Everything here assumes natural gas. Against oil, propane or electric resistance, the heat pump wins almost everywhere, and against resistance always. About half of US homes do not heat with gas.
Whether the AC is due anyway. A heat pump is an AC that also runs in reverse. If you must replace the AC regardless, compare the extra cost of a heat pump over a new AC, not a heat pump against a furnace.
A special rate. Some utilities offer heat pump or all-electric rates. Where one exists, it can reverse the state figures, though regulators can change it later.
HyreHVAC analysis: the state table uses 2024 prices and will move. Treat it as a method: put your current rates into the calculator and the answer updates.
Questions this calculator answers
Is a heat pump cheaper to run than a furnace?
It depends on your gas price, electricity price, AFUE and HSPF2. On 60 million Btu, a 95% furnace at $1.20/therm beats an HSPF2 8.1 heat pump at 16¢/kWh. Change the rates and the answer can flip.
What is HSPF2?
Heating Seasonal Performance Factor under the 2023 test: Btu of heat per watt-hour of electricity across a heating season. Higher costs less to run. Divide by 3.412 for seasonal COP.
Do heat pumps work in the cold?
Yes. Certified cold-climate models keep a median 83% of their 47°F capacity at 5°F. Efficiency drops in deep cold, though. Ask for capacity at your design temperature on the proposal.
What HSPF2 does a heat pump need to beat gas?
About HSPF2 11.2 at US average 2024 prices against a 95% furnace. It varies widely by state; the table on this page lists all 51. For an 80% furnace, multiply by 0.842.
Is a dual-fuel system worth it?
Often. It runs the heat pump in mild weather and the furnace in deep cold, so you buy the cheaper fuel at every temperature. It costs more to install, since you keep two heat sources.
Is there still a federal heat pump tax credit?
No, not for property placed in service after December 31, 2025, when §25C ended under Public Law 119-21. Utility rebates are separate and still exist.
Air-source heat pumps key product criteria (ENERGY STAR, retrieved 2026-09-05. Standard and Cold Climate criteria, including COP and capacity at 5°F. No version or effective date was stated.)
How to use the AHRI directory (Air-Conditioning, Heating, and Refrigeration Institute, retrieved 2026-09-05. Returned 403 to automated retrieval. The matched-system rule is taken from 10 CFR 430.32(c)(6)(ii) and 10 CFR 429.16.)