Find your smart speaker power consumption by entering your wattage, the hours a day your smart speaker runs and your electricity rate. Click Calculate and you get your cost per day, month and year along with the kWh it uses. Those same results also cover smart speaker electricity consumption.
Smart speaker electricity consumption is tiny compared with almost everything else plugged in around your home, but the exact figure depends on how you use the device and what it is connected to. This guide shows how much energy these always connected devices draw, what that adds to your yearly bill, and the one pairing with a television that can quietly change the picture.
Retailers call the product a voice assistant, a virtual home assistant or simply a home assistant, and a smart speaker by any of those names never fully switches off. A microphone and a small processor stay awake so the device can hear its wake word, and a wireless chip holds a connection to the internet at all times. That is why the unit draws a few watts around the clock instead of dropping to zero when nobody is talking to it. Next, look at electric car power consumption.
Independent testing, including a well-known study by the Natural Resources Defense Council, found that popular models from Amazon, Apple and Google each cost only a few dollars of electricity per year. A published report from that work also measured the Harman Kardon Invoke, a speaker built around Microsoft's voice assistant, and it used more than the others in standby, which shows that two products in the same category can still differ by a factor of two.
Manufacturers rarely publish a clean figure, so researchers measure two numbers with a plug-in meter. Standby power is what the speaker draws while it sits idle waiting to be addressed. On-mode power is what it draws while playing music or speaking an answer, and it rises with volume because the amplifier works harder. Reading both numbers is the only reliable way to compare the electricity a smart speaker draws from one model to the next.
Because the speaker keeps listening, it spends most of its life sitting idle in the lowest state it can reach, not in the loudest. Over a full year that quiet state accounts for the largest share of consumption, even though each hour of it costs almost nothing. That idle state is what drives a smart speaker's yearly energy use, and it adds up across twelve months despite looking harmless on any single day.
You can work out the yearly figure for any device with the same two steps. First convert watts into kilowatt-hours by multiplying the power by the hours of use and dividing by 1,000. Then multiply the result by your unit rate.
$$\text{Annual cost} = \frac{\text{Watts} \times \text{Hours per day} \times 365}{1000} \times \text{Rate per kWh}$$
The power rating printed on the power supply is a maximum, not what the unit draws day to day, so use a measured value when you have one. A plug-in meter costs very little and reads in watts directly. Your electricity bill states the average unit rates you pay, and a smart meter display can confirm the result by showing live demand with the speaker unplugged and plugged back in.
Take a speaker that spends 20 hours a day idle at 2.1 watts, 3 hours a day playing music at 6.8 watts, and 1 hour a day answering questions at 4.3 watts. At a rate of $0.17 per kWh, the table below shows the cost per hour, the cost per week and the cost per year for each mode.
| Usage mode | Power (watts) | Hours per day | Cost per hour | Cost per week | Cost per year |
|---|---|---|---|---|---|
| Idle, waiting for wake word | 2.1 | 20 | $0.00036 | $0.050 | $2.61 |
| Streaming music | 6.8 | 3 | $0.00116 | $0.024 | $1.27 |
| Answering requests | 4.3 | 1 | $0.00073 | $0.005 | $0.27 |
| Total | - | 24 | - | $0.079 | $4.14 |
That adds up to 24.35 kWh over the year, and the total is $4.14 when calculated from unrounded values. Notice that idle time contributes more than half the bill even though the speaker is doing nothing useful, which is the pattern researchers keep finding.
Context matters more than the raw number. Consumer guides that rank appliances by running cost put heaters, tumble dryers, electric showers and kettles at the top, each costing tens or hundreds of dollars or pounds a year, while a smart speaker sits near the bottom alongside lamps, chargers and a record player. Quoted prices in the UK are based on the energy price cap and your tariff, so the same speaker costs a little more or less depending on where you live.
Alongside other small appliances and kitchen appliances, a typical home might own one speaker in the kitchen, one in the living room and a small one in each bedroom. Five units like the example above would add roughly $20.69 to a year of electricity bills, which is less than most people spend on a single month of streaming subscriptions.
The exception to the low numbers comes from linking a speaker with a smart TV. When you set up voice control so the speaker can switch the television on, the TV must stay partly awake to hear that signal. Testing on newer sets found that some drew close to 20 watts continuously in that state instead of the usual fraction of a watt. Devices that behave this way are often called energy vampires because they feed on power even when they appear to be off. Compare with side-by-side refrigerator electricity consumption.
The ability to wake and control a set with your voice commands is the feature that raises standby mode demand, whereas using the physical remote control leaves it low. Here is what the difference looks like over a year for a television left in standby, using the same $0.17 rate:
For one home that is noticeable but not dramatic, and it shows how a speaker's setup can raise a home's electricity use even though the speaker itself draws little. Spread across millions of households, the same behavior equals the output of several large power plant units, with matching pollution. Researchers who raised the issue with TV manufacturers heard that design improvements and software updates were already lowering the draw to a watt or two, so a recent model may not show the problem at all.
For video streaming, streaming devices such as sticks and small boxes sit in a similar low range as speakers. They draw a few watts in use and often less than that in sleep. Older set-top boxes and DVRs supplied by pay TV companies draw much more. If your provider offers the same channels through an app on a streaming stick, returning the old box is one of the easier ways to cut a bill.
Next to a smart speaker, a streaming stick draws a similar few watts. Counted at national scale, the energy consumption of both is measured in gigawatt hours a year, which comes from the sheer number of units, not from any one of them.
Because the speaker is so cheap to run, the biggest wins come from the equipment around it. These tips help you keep your savings without giving up the convenience:
Lowering smart speaker use this way trims your bill and a small share of generation-related emissions, in line with net zero goals and a smaller carbon footprint. Every one of these energy-saving steps helps the consumer twice, since the watts that cost money also produce carbon.
For the price of roughly one coffee a year, you get a music player, a timer, a home controller and an answering service. Judged on cost per useful hour, few devices in the house are better value, and the efficiency of modern designs keeps improving. The product is described as energy efficient in nearly every independent review for exactly that reason.
Names you will see in test results include the Amazon Echo, the Google Home and its smaller Google Home Mini sibling, the Apple HomePod and the Harman Kardon Invoke. Larger speakers with better drivers tend to draw more in both modes, and the leader in standby efficiency is usually a compact unit. A digital assistant embedded in a phone or watch uses the device's own battery rather than the wall socket, so it does not appear in this comparison at all.
The right way to compare is always annual energy use in kWh, because it combines both power levels and your habits into one number. Convert to annual electricity costs by multiplying with your unit rate, and use the result to judge whether a difference between two models is worth any attention.