Is Bitcoin mining profitable in 2026? What actually decides it

Bitcoin mining can still turn a profit in 2026, but the conditions are tight. A miner needs cheap power, efficient machines, and a Bitcoin price high enough to cover both. Miss one and the math goes red. Even the largest listed miners lost money this year as the price slipped and the network got harder to mine.

Profit here is simple to state and hard to reach. You earn newly issued bitcoin plus transaction fees for finding blocks. You pay for electricity, machines, cooling, rent and pool fees. When the first number beats the second, you make money. When it does not, you burn cash.

What counts as profit in bitcoin mining

Revenue comes from two places. The block reward, which has been 3.125 BTC since the April 2024 halving, and the fees users attach to their transactions. A single miner almost never wins a block alone, so most join a pool and share the rewards in proportion to the work they contribute.

Costs are where miners live or die. Power is the biggest one. Then come the machines, which age fast, plus hosting, maintenance and a pool fee of about 2 percent. Depreciation on the hardware is a real cost too, even though no cash leaves the account each day.

Fees matter more than they once did. When blocks fill with transactions, the fees attached can add a real top-up to the fixed subsidy, and in quiet periods they fade to almost nothing. A pool smooths this out, paying members a steady share of what the group finds rather than the all-or-nothing swing of mining alone. Most pools take a small cut and pay out in bitcoin.

The three things that decide the outcome

Three inputs move the result more than anything else: your electricity price, how efficient your machines are, and the price of Bitcoin. Network difficulty sits behind all three, because it sets how much hashing the whole network must do to find each block.

Electricity is quoted per kilowatt-hour. Industrial miners who pay $0.03 to $0.05 per kWh can run almost any modern rig at a profit. At a home rate of $0.10 to $0.15, only the newest and most efficient machines stand a chance, and often not even those.

Hardware efficiency is measured in joules per terahash, or J/TH. Lower is better. Bitmain's Antminer S21 Pro runs at 15 J/TH and the base S21 at 17.5, according to mining data firm Hashrate Index. An older S19 XP sits at 21.5 and the S19j Pro at 29.5. On the same power bill, the old machine earns far less bitcoin.

Price is the lever nobody controls. Bitcoin traded near $85,500 at around 00:30 UTC on October 7, 2026, according to CoinGecko, below where it sat through much of 2025. The network also kept getting harder. Difficulty held at a record 132.76 trillion and total hashrate hovered around 969 EH/s on October 2, 2026, according to mining tracker Pickaxe. More machines chasing the same reward means a smaller slice for each.

Difficulty adjusts itself about every two weeks, or every 2,016 blocks, to keep new blocks landing roughly 10 minutes apart. When more machines switch on, difficulty rises at the next retarget and each miner's share of the reward shrinks. When machines switch off, it eases back. That feedback loop is why a miner who stands still slowly falls behind the pack.

Why the block reward keeps shrinking

The reward is not fixed. About every four years, the protocol cuts it in half. It dropped to 3.125 BTC in April 2024 and is set to fall to 1.5625 BTC around 2028. Our guide to how the Bitcoin halving works walks through the full schedule and why supply stops at 21 million coins.

Each halving doubles the pressure overnight. The same electricity bill now buys half the new bitcoin it did the week before. Fees are meant to fill the gap over the long run, though on most days they still make up a small part of what a miner earns.

What it costs to run a single machine

Take one S21 Pro rated at 234 TH/s and 3,510 watts. Run it all day and it draws about 84 kilowatt-hours, more power than many homes use in two days. At $0.06 per kWh, that is roughly $5.05 a day in electricity for one box. At a home rate of $0.12, it is $10.11. Multiply by thousands of machines and the bill gets serious.

To judge revenue, miners watch hashprice, the daily dollar revenue per unit of hashing power. In illustrative 2026 scenarios published by pool operator ViaBTC, hashprice ranged from about $35 to $65 per petahash per day, depending on the coin price and fees. When hashprice falls below a rig's running cost, that machine loses money every hour it stays on.

Put the two sides together for that one machine. At 234 TH/s it holds 0.234 of a petahash, so at a $50 hashprice it would bring in about $11.70 a day before fees. Take off the $5.05 in power and the gross margin is near $6.65, which still has to cover the machine itself, cooling and the pool cut. Halve the hashprice and that margin is gone. This is why miners obsess over power contracts and fleet efficiency, not slogans.

No single number captures the true cost of one coin. Each operation has its own power deal, machines and uptime, so the all-in figures miners publish vary widely and are not directly comparable.

What the big miners' results show

The listed miners give the clearest read, because they report hard figures. MARA Holdings mined 2,422 BTC in the second quarter of 2026 at an average price near $71,325, ran 70.3 EH/s of capacity, and cut its cost per petahash per day by 4 percent to $27.7, according to The Block. Revenue still fell 27 percent to $174.9 million, and the company posted a net loss of $611.3 million.

CleanSpark told a similar story, with a net loss of $239.8 million in its fiscal third quarter. Both firms are now building high-performance computing and AI capacity alongside mining, chasing steadier income than block rewards alone provide. For the basics of the work itself, see our plain guide to what Bitcoin miners actually do.

That pivot says plenty. When the most efficient, best-funded miners on the planet lose money and start looking for other work, the idea that mining is easy money does not survive contact with a spreadsheet.

Competition for power is the bigger threat now. AI data centers want the same cheap electricity and ready grid connections that miners spent years locking up, and they can often pay more per megawatt. Some operators have started leasing part of their sites to AI tenants, turning a mining campus into a hybrid. For a pure miner, that rivalry is one more reason the power price, not the coin price, often settles who survives.

Can a home miner still make money?

For most people at home, the honest answer is no, or not by much. Residential power usually costs too much, one or two machines throw off real noise and heat, and solo mining a block is a long-odds bet against industrial farms. Some hobbyists run a miner anyway, for the hardware, the winter heat, or a lottery-style shot at a whole block.

Hosting is the middle path. You buy the machine and a data center runs it at a cheaper power rate for a fee. It lowers the hassle, not the underlying risk. If the Bitcoin price drops or difficulty climbs, a hosted miner can still lose money.

What to watch

Four forces decide where mining economics go next. The Bitcoin price, which sets revenue. Network difficulty, which keeps grinding higher as capacity returns. The 2028 halving, which will again cut the reward in half. And the race for cheap power, now shared with AI data centers that can outbid miners for both electricity and sites.

None of this is a nudge to mine or not to mine. It is the arithmetic every operator runs before switching on a machine. Get the power price, the hardware and the coin price right, and mining can pay. Get one wrong, and the rig quietly drains the wallet instead.