Is Bitcoin Mining Still Profitable in 2026

Yes, but only if you run the numbers honestly. The days of plugging in an ASIC and printing money are long gone. Post-halving economics, rising network difficulty, and energy costs have squeezed margins hard. That doesn’t mean mining is dead. It means lazy mining is dead.

What the 2024 halving actually did to your revenue

*Photo by RDNE Stock project on Pexels*

The April 2024 halving cut block rewards from 6.25 BTC to 3.125 BTC. That’s a 50% reduction in the primary revenue stream for every miner on the planet. If your operation was barely profitable before the halving, it’s underwater now.

Bitcoin’s price has climbed since then, which partially offset the reward cut. But price appreciation doesn’t scale linearly with difficulty. As of mid-2026, the network hashrate sits above 750 EH/s, according to Blockchain.com data. That’s roughly double what it was two years ago. More hashrate means more competition for the same block rewards. Your share of the pie keeps shrinking unless you keep adding machines.

The math is straightforward. Your daily BTC revenue equals your hashrate divided by the network hashrate, multiplied by the daily block rewards plus transaction fees. Run those numbers with today’s figures and you’ll see why efficiency matters more than ever.

The real cost breakdown

*Photo by Monstera Production on Pexels*

Electricity is still the single biggest variable in your profitability equation. It’s not even close.

A modern ASIC like the Antminer S21 XP pulls about 3,600W and produces roughly 270 TH/s. At $0.07 per kWh, that machine costs about $6.05 per day to run. At $0.12 per kWh, it jumps to $10.37. At $0.04 per kWh (think hydro in the Pacific Northwest or certain Middle Eastern operations), you’re looking at $3.46.

Here’s a rough daily profitability snapshot at current conditions:

Electricity cost	Daily revenue (est.)	Daily electricity	Daily profit
$0.04/kWh	~$14	$3.46	~$10.54
$0.07/kWh	~$14	$6.05	~$7.95
$0.10/kWh	~$14	$8.64	~$5.36
$0.12/kWh	~$14	$10.37	~$3.63

These numbers shift daily with Bitcoin price and difficulty adjustments. But the pattern holds: electricity cost is the difference between a healthy operation and a money pit.

Hardware ROI in 2026

*Photo by Towfiqu barbhuiya on Pexels*

An Antminer S21 XP runs about $5,000 to $6,000 at current market prices. At $0.07/kWh electricity and current difficulty, you’re looking at roughly 750 to 900 days to break even on hardware alone. That’s over two years, and it assumes difficulty doesn’t spike further.

Compare that to 2021, when an S19 Pro paid for itself in under 300 days at similar electricity rates. The ROI window has stretched dramatically. You need to factor in:

• Hardware depreciation: ASICs lose value fast. An S19 that cost $10,000 in 2021 is worth under $500 now.
• Hosting fees: If you’re not running your own facility, hosting costs add $0.01 to $0.03 per kWh on top of your electricity rate.
• Maintenance: Fans fail. PSUs die. Hashboards need replacement. Budget 2% to 5% of hardware cost annually.
• Difficulty growth: The network has been growing roughly 30% to 50% year-over-year. Your revenue per TH/s drops accordingly.

If you can’t get electricity below $0.08/kWh, you need to seriously question whether buying new hardware makes sense. At higher rates, you’re essentially subsidizing the network with your capital.

Where the profit actually is

*Photo by George Pagan III on Unsplash*

The miners making money in 2026 share three traits: cheap power, efficient machines, and operational discipline.

Cheap power means sub-$0.05/kWh. That’s achievable through direct power purchase agreements with renewable energy providers, stranded gas operations, or government-subsidized industrial rates in certain countries. If you’re paying retail residential rates, stop mining. Seriously.

Efficient machines means the latest generation only. Running an S19 at 29.5 J/TH when the S21 XP does 15.0 J/TH is burning money. Older machines should be sold or retired unless your electricity is essentially free.

Operational discipline means tracking every expense, monitoring uptime, and making hard calls. If a machine isn’t covering its electricity cost after three consecutive difficulty adjustments, unplug it. Sentimentality has no place in a mining operation.

Transaction fees: the wild card

*Photo by Ivan S on Pexels*

Block rewards will keep halving every four years. Transaction fees are supposed to eventually replace them as the primary miner revenue. So far, that transition hasn’t materialized in a meaningful way.

Fees spiked during the Ordinals and BRC-20 craze in 2023 and 2024, briefly making up 30% to 40% of block rewards. They’ve since normalized to around 5% to 10% most days. Layer 2 solutions like the Lightning Network reduce on-chain fee pressure, which is good for Bitcoin adoption but bad for miner revenue.

Don’t build your profitability model on fee spikes. They’re unpredictable and unsustainable. Base your projections on block rewards plus a conservative fee estimate.

Should you mine or just buy Bitcoin?

*Photo by Kanchanara on Unsplash*

This is the question every rational miner has to ask themselves. If you took the $6,000 you’d spend on an ASIC and just bought Bitcoin instead, you’d have roughly 0.06 BTC at current prices. If your mining operation produces less than 0.06 BTC over the hardware’s profitable lifespan (after electricity costs), you would have been better off buying and holding.

For operators with electricity above $0.10/kWh, buying Bitcoin is almost certainly the better play. For operators with $0.03 to $0.05/kWh power and access to bulk hardware pricing, mining still wins. The middle ground ($0.06 to $0.09/kWh) depends on your specific setup, tax situation, and conviction on future Bitcoin price.

The honest answer

*Photo by Magda Ehlers on Pexels*

Bitcoin mining in 2026 is profitable for well-capitalized operations with cheap power and modern hardware. It’s break-even or worse for everyone else. The network doesn’t care about your feelings or your sunk costs. It rewards efficiency and punishes everyone else equally.

Run your own numbers. Use a mining calculator like WhatToMine with your actual electricity rate and hardware. Factor in difficulty growth. If the math doesn’t work, don’t force it. There’s no shame in buying Bitcoin directly instead of mining it at a loss.

You’re about to drop $15,000 on a batch of Antminer S21 Pros from a supplier in Shenzhen, and your bank just flagged the transaction. Sound familiar? If you run a mining operation of any size, you’ve hit this wall before. International hardware purchases are the backbone of crypto mining, and traditional payment rails make them unnecessarily painful.

Why mining hardware procurement is a payment headache

Mining hardware isn’t cheap, and it doesn’t come from around the corner. The best ASICs ship from manufacturers in China, Southeast Asia, and increasingly from secondary markets in Eastern Europe. A single unit can run $3,000 to $10,000 depending on the model and market conditions. When you’re ordering in bulk, you’re looking at five- or six-figure invoices that need to clear fast.

Traditional bank wires work, but they’re slow. Three to five business days is standard, and that’s if your bank doesn’t freeze the transfer for “review.” Wire fees eat into margins too. A $45 fee on a $50,000 order isn’t catastrophic, but it adds up when you’re placing orders monthly. Credit cards would be ideal for the speed and buyer protection, but most corporate cards carry foreign transaction fees of 2.5% to 3%, and many suppliers don’t accept them for large orders anyway.

Then there’s the currency conversion problem. Your supplier quotes in USD or CNY. Your bank converts at a rate that benefits the bank, not you. You lose another 1% to 2% on the spread before the payment even arrives.

How stablecoin corporate cards change the equation

A stablecoin corporate card solves most of these problems in one move. You fund the card with USDC or USDT, and the card processes payments in whatever fiat currency the supplier accepts. No bank freezing your transaction because it looks “unusual.” No waiting for wire confirmations. No losing money on bad exchange rates.

Here’s how it works in practice. You load your corporate card with stablecoins from your treasury or exchange account. When it’s time to buy hardware, you pay the supplier just like any card transaction. The settlement happens in fiat on their end. They get paid in their local currency. You paid in crypto. Everyone’s happy.

The speed difference alone makes this worth considering. A wire transfer takes days. A card payment authorizes in seconds and settles within one to three business days. When a new batch of miners drops and inventory moves fast, that speed is the difference between securing hardware and watching it sell out.

Real numbers: what you actually save

Let’s run the math on a typical order. Say you’re buying 10 Antminer S21 XP units at $4,800 each. That’s $48,000 total.

Wire transfer route:

• Wire fee: $45
• Exchange rate loss (1.5%): $720
• Time cost: 3 to 5 days (opportunity cost if price moves)
• Total friction: ~$765

Stablecoin corporate card route:

• Card processing fee: often 0% on the buyer side
• No foreign transaction fees with the right card
• Exchange rate: near mid-market
• Settlement: 1 to 3 days
• Total friction: minimal

You’re saving $700+ per order. Over a year of monthly purchases, that’s $8,000 to $10,000 in fees you’re not paying. That’s another ASIC or two, just from switching payment methods.

Managing mining spend with corporate controls

The other advantage nobody talks about is spend management. Mining operations often have multiple team members placing orders, paying for hosting, or covering electricity deposits. A corporate card program lets you issue multiple cards with individual spending limits.

You can set a $10,000 monthly limit on the card your warehouse manager uses for parts. You can issue a separate card for hosting facility deposits with a $25,000 cap. Everything feeds into one dashboard. No more chasing receipts or reconciling a dozen wire transfers at the end of the month.

This matters more than most operators realize. When you’re scaling from 20 machines to 200, the administrative overhead of payments becomes a real time sink. Corporate card programs with stablecoin funding let you scale your payment infrastructure alongside your hash rate.

What to watch out for

Not every supplier accepts card payments for large orders. Some still prefer wire transfers or crypto direct. Ask before you commit to a card-based workflow. Also, verify the card’s daily and per-transaction limits. A $50,000 hardware order might exceed your default limit, and you don’t want to find that out at checkout.

Tax treatment varies by jurisdiction. In the US, spending stablecoins is generally treated as a taxable event. Consult your accountant. In other jurisdictions, the rules are different or still evolving. Keep clean records regardless.

Finally, watch your stablecoin exposure. If you’re holding your entire treasury in USDT and the peg wobbles, that’s a risk. Diversify your payment rails and don’t park more on a card than you need for near-term purchases.

The bottom line

Mining is a margin business. Every dollar you save on procurement, electricity, and operations goes straight to your bottom line. A stablecoin corporate card won’t make or break your operation, but it removes friction from one of the most common tasks you face: buying hardware internationally, fast, without getting gouged on fees.

If you’re still wiring money the old way, you’re leaving money on the table. Set up a corporate card funded with stablecoins, test it on your next order, and see the difference in your settlement times and fee statements.

Every rig has a second life as parts. Cards get retired, risers pile up, a whole frame gets torn down when you upgrade, and there is a real market for all of it. Buyers who mine already hold crypto, so a lot of them would rather pay you in it than dig out a card and a bank transfer. That works in your favor if you set it up cleanly.

Price the parts honestly. A used mining card is worth less than a boxed one, and pretending otherwise just kills the sale. State the hours it ran, whether it was undervolted, whether you repadded the memory, and post a photo of the actual card with a temperature reading. Buyers who know hardware will pay a fair price for a card that was run cool and documented, and they will walk from a vague listing. Honesty is the cheapest way to close.

When you take payment in crypto, confirm the transaction has settled before the parts leave your hands. Wait for the network confirmations to land. A screenshot of a pending send proves nothing. A pending transaction can be dropped or replaced, and once the box is with a courier you have no way to claw it back. Confirmed funds first, then ship.

Getting paid across borders is where crypto earns its place. A buyer in another country paying by bank wire means fees on both ends, a currency conversion spread, and a wait of several business days while the money crawls through the system. A crypto payment settles in minutes for a flat network fee, regardless of which country the buyer sits in, and you are not handing a chunk to an intermediary for the privilege. For a $600 used card going from one continent to another, that is the difference between clean money in your wallet by lunch and a week of watching a pending transfer.

A few practical habits keep it smooth. Use a stablecoin if you want to avoid price swings between the moment you agree a price and the moment it settles. Keep a simple record of what sold, for how much, and the wallet or transaction it came through, because that trail matters when you square things up later. And give the buyer a clean payment destination, whether that is a wallet address or a checkout link, so there is no fumbling with copied strings and no room for a costly typo.

Sell honest, confirm before you ship, and let the payment cross borders without a bank taking a cut. The hardware market rewards operators who make the transaction easy.

Undervolting is the closest thing to a free lunch in this hobby. You feed the card less voltage, it makes less heat, it pulls fewer watts, and on a lot of memory-bound mining workloads your output barely drops or even climbs because the card stops throttling. Lower bill, cooler room, longer-lived hardware. The catch is that it takes an afternoon of patient testing, and most people skip that afternoon.

The idea is simple. Cards ship with more voltage than they need, because the factory has to guarantee stability across millions of chips, including the weak ones. Your specific card is almost certainly better than worst-case, so there is headroom to pull voltage down. Less voltage means less power, and power is what becomes heat.

The stock-clock method

On Nvidia cards the cleanest approach uses MSI Afterburner’s voltage-frequency curve editor. Open the curve with Ctrl+F. It looks like a scatter plot of clock speed against voltage. What you want is to lock the core to a modest clock at a low voltage, so the card stops chasing high boost clocks it does not need.

Pick a point on the curve, say 800mV, and drag it up to a core clock around 1,300MHz. Then flatten everything to the right of that point so the card never runs past that voltage. Apply it. The core now runs at roughly 1,300MHz while sipping 800mV instead of the 1,050mV it wanted at stock.

Then push the memory. Mining is memory-bound on most algorithms, so memory clock is where your output comes from. Raise the memory clock offset in steps of 50MHz, testing stability at each step, until you find the wall, then back off 100MHz for a safety margin. Watch memory junction temperature the whole time, because faster memory runs hotter even as the core cools.

Real numbers

Take an RTX 3070. At stock it might pull about 220W at the wall for the card and run the core hot. Undervolted with a locked core around 1,100 to 1,300MHz and a memory offset near +1,100MHz, that same card commonly lands near 130W while holding full mining output, sometimes a touch higher because it stopped throttling. That is roughly 90W saved per card.

On a six-card rig, 90W each is 540W off your wall draw. Run that 24/7 for a month and you have cut around 388 kWh. At $0.20/kWh that is about $78 a month, every month, for one afternoon of setup. The cards also run 10 to 20 degrees cooler, which means the fans spin slower, wear less, and the whole rig gets quieter.

Testing and stability

Change one thing at a time. Drop the core, confirm output holds and the card is stable for an hour. Then raise memory in small steps. If the driver crashes, the screen artifacts, or your miner throws invalid or rejected shares, you have gone too far on memory. Back off 50 to 100MHz.

Invalid shares are the tell that matters. A card can look stable and still be quietly producing garbage that gets rejected, which is worse than a clean crash because you keep paying for power while producing nothing usable. Let a new setting run for a few hours and watch the rejected-share count before you call it done.

Save your profile once it is dialed in, and set it to load on startup, because a driver update or a reboot will wipe it and the card will quietly go back to pulling full power. Check your temps and wattage again after any driver update. Ten minutes of verification protects the afternoon you spent tuning.

Power is the bill that shows up whether your rig makes anything or not. The heat, the wear, the noise, all of it is downstream of watts pulled from the wall. Before you plug in a single card, you should be able to do this math in your head, because the wall meter does not care about your optimism.

Start at the wall, not at the spec sheet. A card rated for 320W does not pull 320W total for the system. The rest of the rig pulls power too: the motherboard, the CPU, the RAM, the fans, and the power supply itself wastes a slice as heat. A rig of six 320W cards is not 1,920W. Measure it with a plug-in power meter and you will usually see 2,100 to 2,300W at the wall once you count everything and the PSU’s efficiency loss. That gap is real money, and it only shows up if you measure.

Now turn watts into a bill. The unit you pay for is the kilowatt-hour, which is 1,000 watts running for 1 hour. A rig pulling 2,200W at the wall uses 2.2 kWh every hour. Run it 24 hours and that is 52.8 kWh a day. Over a 30 day month that is 1,584 kWh.

Then multiply by your actual rate. Look at your power bill and find the number in dollars per kWh, and use the real one, including delivery charges and taxes, not the headline supply rate. Rates swing hard by region. Some places sit near $0.10 per kWh. Plenty of places are $0.30 or higher, and parts of Europe have seen $0.40 and up.

Run the same rig at three rates and the picture changes completely:

• At $0.10/kWh: 1,584 kWh times $0.10 is about $158 a month.
• At $0.20/kWh: about $317 a month.
• At $0.35/kWh: about $554 a month.

Same hardware, same heat, same wear. The only thing that changed is the rate, and it nearly quadrupled the cost. This is why two people running identical rigs can have wildly different outcomes, and it has nothing to do with the hardware.

What “pays for itself” means in power terms

Set coin prices aside. In pure energy terms, a rig pays off when what it produces is worth more than the power it burns to produce it. If your rig eats $317 of electricity a month, everything it earns below $317 is you paying to heat a room. The rig is running at a loss the moment the value of its output drops under its power cost, and it will happily keep running at that loss all night because it has no idea.

This is where undervolting earns its keep. Trimming each card from 320W to 230W on a six-card rig cuts wall draw by roughly 540W, which is about 388 kWh a month. At $0.20/kWh that is $78 a month back in your pocket, for output that barely moves. Power is the lever you actually control.

Two costs people forget

First, time-of-use pricing. Many utilities charge more during peak afternoon and evening hours and less overnight. If your rate doubles from 2pm to 8pm, a rig that runs flat out around the clock is paying the premium for 6 hours a day. On a time-of-use plan you may come out ahead running hard overnight and backing off during peak. Read your rate plan and find out if you have peak hours at all.

Second, the cooling tax. Every watt your rig burns becomes heat in the room, and in summer you may be paying a second time to air-condition that heat back out. A rig throwing off 2,200W of heat into a small room in July can add meaningfully to your cooling bill, so the true cost of running it is the power plus the cooling to survive it. In winter that heat offsets your heating, which is a genuine rebate for a few months a year.

One more thing worth checking before you commit: the circuit itself. A standard 15-amp household circuit at 120V tops out near 1,800W of continuous safe load, and code says stay under 80 percent of the breaker rating on a continuous draw. A 2,200W rig will trip that breaker or, worse, cook the wiring in the wall if the breaker is tired. Spread big rigs across separate circuits, or run a dedicated 20-amp line. The power math protects your wallet and your wiring both, and an overloaded circuit is how a mining setup becomes a house fire.

Do the arithmetic before the rig is on the floor. Wall watts times 24, times 30, times your real rate. If that number scares you, it is better to be scared now than on the first bill.

A mining rig is a space heater that happens to do math. Run it hard enough for long enough and it will try to destroy itself with its own waste heat. Your job is to move that heat out of the case, out of the room, and out of the building faster than the cards can make it. Do that well and a GPU lasts years. Do it badly and you are shopping for replacement fans in 3 months and a replacement card in 8.

Start with the number that actually matters: memory junction temperature. Core temp is the friendly number the manufacturer likes to show you, and it lies by omission. On cards with GDDR6X, like the RTX 3080 and 3090, the memory sits under the same cooler as the core but runs much hotter, and the memory is what you are hammering when you mine. A card can report a core at 60C while the memory junction is sitting at 104C. Above 110C that memory throttles hard, and living in the high 90s and low 100s for months cooks the thermal pads until they turn to chalk. If your monitoring software shows a memory junction reading, watch that one. Keep it under 100C. Under 90C is where you want to be.

Ambient is the ceiling you can’t beat

Every temperature inside the case is stacked on top of the temperature of the room. If the air going into the rig is 30C, no fan on earth gets your card down to 25C. You are always fighting from ambient upward. This is why the same rig that runs fine in February starts throttling in July. A garage that sits at 22C in winter can hit 38C in a summer afternoon, and that 16 degree swing lands directly on your silicon.

So manage the room before you fiddle with anything on the card. Pull cool air in from the lowest, coolest part of the space and dump hot air out somewhere it won’t recirculate. Hot air rises, so exhaust high and intake low. If you are venting into an attic or a closet, you are just preheating your own intake air, and the whole rig slowly bakes over the course of a day. A cheap thermometer at the intake tells you more than most people’s entire monitoring setup.

Airflow beats fan speed

More static pressure through the case matters more than screaming fans on the cards themselves. A rig in a tight enclosure with three cards jammed together will always run hotter than the same cards on an open frame in the same room, because the middle card is breathing the exhaust of its neighbors. Space the cards out. Open-air mining frames exist for exactly this reason. If you are stuck with an enclosed case, add a box fan or an inline duct fan moving real volume through it, not just the stock case fans.

Fan curves are the next lever. The default curve on most cards is tuned to stay quiet, which means it lets the card get hot before it spins up. For mining you want the opposite: aggressive early, so the card never gets a chance to heat-soak. Set a custom curve that has the fans at 60 to 70 percent by the time the card hits 55C, and near 100 percent by 70C. Yes it is louder. A rig is not a living-room PC. The tradeoff is that the fans and the card both last longer when the card never sits in the danger zone waiting for the fans to catch up.

One caution: running fans at 100 percent 24/7 wears the bearings out. Sleeve-bearing fans in particular start rattling and dying after a year of full-tilt duty. Double ball-bearing fans handle it better. If a fan starts making a grinding or clicking noise, replace it before it seizes, because a dead fan on a loaded card is how you get a thermal shutdown at 3am, or worse, no shutdown and a scorched card.

Dust is a slow strangle

Dust is the killer nobody schedules for. A GPU heatsink is a dense stack of thin fins, and it works by having air pass between them. Dust bridges the gap between fins, and within a couple of months a card in a dusty room loses a noticeable chunk of its cooling capacity. The card compensates by ramping fans and, when that runs out, by thermal throttling down its clocks, which quietly drops your output while you’re not looking.

Blow the rigs out with compressed air every 4 to 6 weeks, more often if you have pets or the rig lives in a garage. Hold the fans still with a finger or a zip tie while you blow, because spinning a fan backward with a jet of air can generate voltage and damage it. Do this outdoors or you are just relocating the dust. A furnace filter taped over the intake side of an open frame catches most of it and is far easier to swap than cleaning fins.

Thermal pads and paste

The pads and paste under the cooler are consumables on a mining card, even though nobody sells them that way. Factory thermal paste dries out under constant load, and the memory pads compress and lose contact. If a card that used to run at 90C memory junction is now sitting at 104C with clean fins and good airflow, the pads are done. Repadding the memory with 1.5mm or 2mm pads and repasting the core can drop junction temps 15 to 20 degrees on a tired card. It is fiddly, it voids the warranty, and it is often the difference between a card that lasts another 2 years and one that dies this summer.

Watch it before it bites you

Heat problems announce themselves early if you are looking. Log memory junction temp, core temp, fan speed, and power draw, and glance at them once a day. A card that used to sit at 88C and now idles the same graph at 96C is telling you something clogged, dried out, or loosened. Most miners run monitoring on the whole rig anyway, so add a temperature alert that pings you when any card crosses 100C junction. That single alert has saved more cards than any fancy cooler.

Set the thermal limits in software too, as a backstop. Cap the card’s power limit and set a temperature target so it throttles itself before it hits shutdown. Throttling costs you a little output for a few minutes. A card that runs full-tilt into a thermal wall with a dead fan costs you the card. Give yourself the soft landing.

Placement matters more than people expect. A rig shoved against a wall, under a desk, or in a cabinet is choking on its own exhaust no matter how good the fans are. Give it 30cm of clear space on the intake and exhaust sides, keep it off carpet that blocks the bottom and sheds fibers into the fans, and point the hot exhaust away from anything you care about. The best cooler in the world cannot fix a rig that is breathing its own hot air.

Keep the memory cool, keep the room cool, keep the fins clean, and stay ahead of the fans. Everything else is detail.