As of mid-2026, Bitcoin uses roughly 140 to 200 terawatt-hours (TWh) of electricity per year -- about as much as a mid-size country such as Poland or Thailand, and around half a percent of total global electricity consumption. The exact figure depends on the model, but every credible estimate puts annualized Bitcoin electricity use in the mid-hundreds of TWh.

The range is not sloppiness -- it reflects two respected sources that model the network differently. The Cambridge Bitcoin Electricity Consumption Index (CBECI) sits at the lower end (~138 TWh in its most recent study), while Digiconomist sits near the top (~197 TWh). This page explains how that number is measured, why the two estimates diverge, how it has changed over time, what fuels it, and why the popular "per-transaction" figure is misleading.

You cannot meter Bitcoin directly, so researchers estimate it. The Cambridge Centre for Alternative Finance builds the CBECI by taking the network's total hashrate (the collective computing power securing the chain), mapping it to the mix of mining hardware likely in use, and multiplying by each machine's power draw -- then annualizing as if that power level held constant for a full year.

Because the hardware mix is uncertain, Cambridge reports a range rather than a single number:

Cambridge's 2025 sustainability study, based on a survey representing about 48% of global mining activity, estimated Bitcoin's annual electricity consumption at roughly 138 TWh -- about 0.5% of global consumption. Digiconomist puts the figure closer to 197 TWh.

Here is the useful part most summaries skip: the two estimates diverge because they model the network differently. Cambridge works bottom-up from observed hashrate and a probable mix of mining hardware, then takes a mid-point. Digiconomist works from mining economics, assuming miners spend a roughly fixed share of their revenue on electricity -- so its estimate climbs faster when Bitcoin's price is high. Neither is "wrong"; they answer slightly different questions. That is exactly why quoting a sourced range (roughly 140-200 TWh as of mid-2026) is more honest than a single decimal, and why anyone citing "Bitcoin uses X TWh" should say which model they mean.

Abstract TWh figures are hard to picture, so here is the country comparison:

• Bitcoin's ~140-200 TWh/year is comparable to the total annual electricity consumption of a country like Poland (lower estimate) to Thailand (higher estimate) -- Digiconomist explicitly benchmarks its ~197 TWh figure against Thailand.
• It is roughly twice Finland's national electricity demand (~85-95 TWh).
• It is a small slice -- about half a percent -- of the world's total electricity use.

So Bitcoin uses a meaningful amount of power in absolute terms, but a small share globally.

Bitcoin's energy footprint has grown with its price and hashrate, but not linearly -- it dips during bear markets when unprofitable miners switch off, and climbs during bull runs. The long-run direction is up, roughly tracking adoption: as more people own Bitcoin and its price rises, more computing power competes to secure the network each cycle. However, efficiency gains matter: newer ASIC miners do far more hashing per watt than older models, which softens the growth in energy use relative to the growth in hashrate.

Bitcoin mining's fuel mix has shifted meaningfully toward lower-carbon sources. The Cambridge 2025 study found that sustainable energy sources now account for 52.4% of Bitcoin mining -- 42.6% renewables plus 9.8% nuclear -- up from about 37.6% in 2022.

The full breakdown from that study:

Notably, natural gas has overtaken coal as the single largest source (coal fell from 36.6% in 2022 to 8.9%), and miners increasingly use otherwise-flared or stranded gas that would be wasted anyway.

You will often see headlines like "one Bitcoin transaction uses as much energy as a household for X weeks." That framing is misleading. As Bitcoin Magazine and the International Energy Agency both explain, Bitcoin's energy use is driven by mining, not by transaction throughput.

The network consumes the same power whether it processes one transaction or one thousand in a block. Mining energy scales with Bitcoin's price and the resulting competition among miners, not with how many payments settle. Dividing total energy by transaction count produces a number that swings wildly and says nothing about the marginal cost of a payment -- especially since Layer-2 solutions batch huge numbers of transfers off-chain.

Comparing Bitcoin to a whole country is popular but not very useful. A more apples-to-apples comparison is to the industries Bitcoin partly competes with. One widely cited analysis estimated:

By this estimate, Bitcoin uses less than half the energy of either the banking system or gold mining. The caveat cuts both ways: Bitcoin's energy use is transparent and easy to measure, while banking and gold footprints are far harder to pin down -- which makes Bitcoin an easy target for scrutiny that comparable industries rarely face.

Roughly 140-200 TWh annually as of mid-2026, depending on the model. The Cambridge CBECI puts it near 138 TWh, while Digiconomist estimates about 197 TWh. That is around half a percent of global electricity consumption.

It depends on the fuel mix. The Cambridge 2025 study found 52.4% of Bitcoin mining now runs on sustainable energy (renewables plus nuclear), with natural gas replacing coal as the top fossil source. Bitcoin also uses less than half the energy of the banking or gold sectors by some estimates.

Because Bitcoin's energy use comes from mining, not transaction volume. As Bitcoin Magazine and the IEA note, the network draws the same power regardless of how many payments it settles, so dividing total energy by transactions produces a misleading figure.

No. Spending or sending Bitcoin adds a transaction to a block that would be mined anyway, so the marginal energy cost of a payment is essentially zero. On efficient chains like Solana, the per-transaction energy footprint is negligible by design.

Bitcoin's ~140-200 TWh/year is comparable to a mid-size country like Poland or Thailand, and about twice Finland's national demand -- but still only about half a percent of global electricity.

If your goal is simply to use crypto for everyday purchases, the energy question is largely beside the point -- a single payment adds no meaningful load to the network, and stablecoins on efficient chains settle for a fraction of a cent. A crypto debit card lets you pay with crypto anywhere Visa and Mastercard are accepted, so you can spend Bitcoin like cash without ever touching a mining rig. SolCard settles top-ups over Solana in seconds, making the per-payment energy footprint effectively nil.

• Cambridge Bitcoin Electricity Consumption Index (CBECI)
• Cambridge CBECI Methodology
• Cambridge Judge Business School -- Sustainable energy rising in Bitcoin mining (2025)
• Digiconomist -- Bitcoin Energy Consumption Index
• IEA -- Bitcoin energy use, mined the gap
• Bitcoin Magazine -- Energy Per Transaction Is a Misleading Metric