Skip to content
PlainTotal
Energy Costs

Air Conditioner Running Cost Calculator

What air conditioning costs depends on three things: how much cooling your climate demands, how efficient your equipment is, and what you pay per kilowatt-hour.

Last updated Free, no sign-upHow it works

Your inputs
Your air conditioning

For a window or portable unit, the room. For central air, the whole conditioned area.

SEER2 for central systems and mini-splits, CEER for window and portable units. Check the yellow EnergyGuide label. Current federal minimum for central air is about 13.4 to 14.3 SEER2.

Location

Optional. Picks the published price for your state; otherwise the US average is used.

Why we ask

A ZIP code is matched to its state with published postal prefix ranges and used only to look up that state’s row in a federal price table. We never ask for a street address, nothing is stored, and your inputs appear only in this page’s URL.

Results update as you type. Nothing you enter is sent anywhere or stored.

Cost to run your air conditioning: $700, ranging from $570 to $880.

Cost to run your air conditioning

low confidence
$700Per year, at your rate and climate
$570 lowhigh $880
  • United States average
  • Assumptions 2026.07.1

Why low: no state was resolved, so national averages were used; the modeled range spans about 44% of the typical value; building envelope and occupancy are not modeled and can move cooling energy by a third in either direction.

Key figures

Per month during cooling season
$176($143 – $220)Spread over about 4 months of active cooling in your climate.
Electricity used
3,836 kWh(3,116 kWh – 4,794 kWh)
Cooling capacity
43,500 BTU/hrSized from 2,000 sq ft at 550 sq ft per ton.
Equivalent full-load hours
942 hrDerived from 1,177 cooling degree days. This is how many hours the system would run if it ran at full capacity the whole time.
Cooling degree days
1,177Unweighted mean of state degree days (2021–2025 mean); no state resolved.
At 20 SEER2 instead
$492($400 – $616)What the same cooling would cost with more efficient equipment.
Electricity rate used
$0.18/kWhUnited States residential average of $0.1834/kWh (EIA, June 2026).

Cost breakdown

Estimated cost of each component, with low, typical, and high values.
ComponentLowTypicalHigh
Cooling electricity — central air conditionerEnergy3,836 kWh per cooling season at $0.1834/kWh.$570$700$880
Cost to run your air conditioning$570$700$880

The low and high columns above will not add up to the headline range. Each line has its own uncertainty, and every line landing at its worst case at the same time is unlikely, so the total combines them as partially independent rather than summing the extremes.

What moved your estimate

  • Upgrading to 20 SEER2 — decreases the estimate

    Would cut running cost by about $211 a year at your rate and climate. Weigh that against the installed cost before deciding.

How this number was produced
  1. Equivalent full-load hours1,177 CDD × 0.8942 hr
  2. Load adjustment1.14 setpoint × 1.15 distribution1.31
  3. Annual energy43,500 BTU/hr × hours × load ÷ (14 × 1,000)3,836 kWh
  4. Annual costkWh × $0.1834$703

Adjustments applied

  • Local cooling load×0.98

    Unweighted mean of state degree days (2021–2025 mean); no state resolved.

  • Thermostat setpoint 74°F×1.14

    Relative to a 76°F baseline at 7% per degree.

  • Distribution losses×1.15

    Ducted systems lose conditioned air through leakage and to unconditioned attic and crawlspace runs.

  • Electricity rate×1.00

    United States residential average of $0.1834/kWh (EIA, June 2026).

Notes and sources3 notes, 2 sources

Worth knowing

  • SEER2 is a seasonal average across a range of outdoor temperatures, so it already accounts for part-load operation. Using it with equivalent full-load hours is the standard way to estimate seasonal energy, but it is an estimate and a real house varies with insulation, shading, window area, and how many people are in it.
  • This covers cooling only. If you have a heat pump, winter heating is a separate and usually larger number.
  • Degree days are a five-year average. An unusually hot summer will push your actual bill above this.

Sources

  • U.S. Energy Information Administration, Electric Power Monthly, Table 5.6.A — Average Price of Electricity to Ultimate Customers by End-Use Sector, by StateprimaryJune 2026retrieved 2026-09-11
  • NOAA Climate Prediction Center, population-weighted daily degree days (StatesCONUS), 5-year mean. Alaska and Hawaii from U.S. Energy Information Administration State Energy Data System indicators.mixed2021–2025 meanretrieved 2026-07-31

This calculator takes the first from five-year average cooling degree days for your state, the second from your equipment’s efficiency rating, and the third from published rates — then shows the working.

Methodology

How this calculator works

Turning climate into running hours

Cooling degree days measure how much, and for how long, outdoor temperature sits above a base of 65°F across a year. We use NOAA’s population-weighted figures for your state, averaged over five years so one hot summer does not distort the answer.

Degree days are converted into equivalent full-load hours — the number of hours the system would run if it ran flat out the whole time — by dividing by the difference between a design outdoor temperature and the balance point. That works out to roughly 0.8 hours per degree day, and we carry a band from 0.65 to 1.0 to reflect how much building envelopes differ.

Turning running hours into dollars

An efficiency rating is BTU of cooling delivered per watt-hour of electricity. Dividing capacity by the rating gives the average power draw, and multiplying by full-load hours gives kilowatt-hours.

Ducted systems get a 15% penalty for distribution losses, because ducts leak and often run through hot attics. Ductless systems do not.

Thermostat setpoint is applied at about 7% per degree below 76°F. This is a well-established rule of thumb rather than a measurement of your specific house.

What this does not model

Insulation levels, window area and orientation, shading, air leakage, ceiling height, and how many people and appliances are adding heat inside. Together these can move real cooling energy by a third in either direction, which is why the confidence on this calculator is deliberately not shown as high.

Time-of-use pricing is also not modeled, and it matters here more than for most appliances, because air conditioning peaks in exactly the late-afternoon hours when time-of-use rates are highest.

Location and data

Where the numbers come from

Location is optional, and we only ask for a ZIP code or a state — never a street address. A ZIP code is matched to its state using published postal prefix ranges, and that state is used to look up labor and price rows in the tables below. If you leave it blank, or your ZIP cannot be matched, the estimate falls back to a national average and says so on the result.

Nothing you enter is transmitted or stored. The calculation runs entirely in your browser, and the only place your inputs appear is in the page URL, so you can bookmark or share a result if you choose to.

  • primaryJune 2026v2026.06

    U.S. Energy Information Administration, Electric Power Monthly, Table 5.6.A — Average Price of Electricity to Ultimate Customers by End-Use Sector, by State

    Residential sector average retail price. State averages blend all utilities and rate classes in the state; an individual household bill can differ substantially, especially in deregulated markets and for customers on time-of-use tariffs.

  • mixed2021–2025 meanv2026.07

    NOAA Climate Prediction Center, population-weighted daily degree days (StatesCONUS), 5-year mean. Alaska and Hawaii from U.S. Energy Information Administration State Energy Data System indicators.

    Degree days are population-weighted state means over a 5-year window, which smooths single-year weather swings. IECC zone is DERIVED from HDD65 threshold bands and is the dominant zone for the state only; IECC assigns zones by county and large states span several. Alaska and Hawaii are not in the NOAA CONUS product and use EIA SEDS values. The District of Columbia inherits Maryland.

Full detail on every dataset is on the data sources page. Spotted something wrong? Report incorrect data.

Calibration

The judgements behind this estimate

These are the figures we chose rather than measured. For each one we record the range of values found across references, what we selected, and why — because an estimate you cannot interrogate is not much better than a guess. The 1 marked high risk are the ones most likely to be wrong and most consequential when they are.

  • Equivalent full-load cooling hours per cooling degree dayhigh risklow confidence
    Observed range
    0.55 — 1.15 (typical 0.8)hours per CDD65
    Why this value
    Derived from CDD x 24 / (design outdoor temperature - balance point). With a 95F design day and a 65F balance point this is 0.8. A tight, well-shaded, well-insulated house behaves like 0.55; a leaky one with large west-facing glazing behaves like 1.15. We narrowed to 0.65-1.0 because the extremes describe envelopes far outside the median United States housing stock. Cross-checked against published equivalent full-load hour tables for representative cities: Florida lands near 2,500 hours and Minnesota near 400, both consistent with this factor.
    Applies to
    United States. The factor is climate-independent by construction; climate enters through the degree-day input.
    Known limits
    Building envelope is not modelled at all, and it can move real cooling energy by a third in either direction. This is why the air conditioner calculator caps its own confidence.
    Sources
    Recorded before the richer schema; see rationale for provenance
    Last verified
    2026-07-31 · reviewed biennial
  • Cooling load sensitivity to thermostat setpointmedium confidence
    Observed range
    0.03 — 0.1 (typical 0.07)share of load per degree F
    Why this value
    Energy-efficiency guidance consistently quotes 6-8% per degree; measured studies range wider, from 3% in mild climates to 10% in hot ones, because the effect depends on how far the setpoint sits from the outdoor mean. We ship the widely published 7% as a point value and state it as a rule of thumb rather than a measurement of the user's house.
    Applies to
    United States. Genuinely climate-dependent; not currently varied by climate.
    Known limits
    No additional limitation recorded.
    Sources
    Recorded before the richer schema; see rationale for provenance
    Last verified
    2026-07-31 · reviewed biennial
  • Ducted distribution lossmedium confidence
    Observed range
    1.05 — 1.35 (typical 1.15)multiple of delivered load
    Why this value
    Duct leakage and conduction losses in unconditioned attics are commonly cited at 15-30% of delivered capacity; well-sealed ducts inside conditioned space approach 5%. We ship 15% as a single value for ducted systems and 0% for ductless, and disclose that leaky attic ducts can be considerably worse.
    Applies to
    United States. Worse in hot climates where ducts run through vented attics.
    Known limits
    No additional limitation recorded.
    Sources
    Recorded before the richer schema; see rationale for provenance
    Last verified
    2026-07-31 · reviewed biennial

Cost factors

What changes the price most

Climate
The largest single factor. A Florida cooling season demands roughly seven times the runtime of a Washington one.
Efficiency rating
Running cost is inversely proportional to it. Going from 13 to 20 SEER2 cuts cooling electricity by about 35%.
Thermostat setpoint
About 7% per degree. Two degrees warmer is a noticeable saving and usually barely perceptible.
Electricity rate
Varies more than fourfold across states, and air conditioning runs precisely when peak rates apply.
Duct condition
Leaky ducts in an unconditioned attic can waste a fifth of the cooling you paid to produce.
Envelope
Insulation, shading, and air sealing determine the load itself. Not modeled here, and the reason two identical systems can have very different bills.

Questions

Frequently asked

How much does it cost to run central air per month?

In a moderate climate with mid-efficiency equipment, roughly $60 to $130 a month during the cooling season at the national average rate. In a hot climate with expensive electricity it can exceed $300. Enter your own details above — the spread between states is the whole point.

Is it cheaper to leave the AC on all day or turn it off?

For most homes it is cheaper to let the house warm up while nobody is there. A house loses heat to the outside in proportion to the temperature difference, so a warmer house leaks less. The old argument that recooling costs more than you saved does not hold for typical absences. A programmable or smart thermostat makes this automatic.

What temperature should I set my thermostat to save money?

Every degree warmer saves roughly 7% of cooling energy. 78°F while home and warmer while away is the common efficiency recommendation, though comfort and humidity matter too — in a humid climate a slightly lower setting with good dehumidification can feel better than a higher one without.

Are portable air conditioners expensive to run?

Yes, disproportionately. Single-hose portables blow conditioned indoor air out the window, which draws hot outside air in through every gap to replace it. Their real-world performance is meaningfully worse than the rating suggests, and a window unit of the same nominal capacity will usually cost less to run.

Will a new air conditioner pay for itself?

Compare the figure above against the same calculation at a higher efficiency rating using the "compare against" field. In a long cooling season with expensive electricity the saving is real and can be several hundred dollars a year. In a short season with cheap power the equipment usually wears out before it pays back. Replacement is more often justified by reliability than by energy alone.