Bitcoin’s Hashrate Growth Is Back, and Price Didn’t Lead It

For most of 2026, the story of Bitcoin’s mining network was a story about a plateau. The hashrate had slipped off its 2025 record and stayed there, difficulty fell for the first time in years, and the headlines were about miners walking away from the blockchain to rent their power to artificial-intelligence tenants. Then, in a single month, the network changed its mind.

Between the middle of October and the end of the month, Bitcoin’s mining difficulty rose twice, by 5.92 percent and then 5.54 percent, lifting it roughly 11.8 percent in four weeks to 148.36 trillion, according to a November mining update from MillionMiner. That is the highest reading since 27 November 2025 and only 4.9 percent below the all-time record of 155.97 trillion set on 29 October 2025. The seven-day average hashrate, the number those adjustments are chasing, climbed back above 1,000 exahashes per second (one zettahash) on 11 October for the first time since June and peaked at 1,088 EH/s on 26 October, its highest since December 2025. By early November it sat near 1,069 EH/s, within about 8 percent of the all-time high.

The strange part is what the price was doing while all of this happened. Bitcoin did not rally into the recovery. It fell. BTC closed October at $77,437, down 7.3 percent on the month, per MillionMiner, and traded around $79,641 in early November, roughly 37 percent below its October 2025 peak of $126,080, per CoinGecko. Hashprice, the revenue a miner earns per unit of computing power, fell 17.5 percent over the same month. Hashrate growth, in other words, outran the price that is supposed to drive it. This piece explains how that happened, who brought the machines back, and whether a new all-time high is the next thing on the board.

What hashrate growth actually measures

Hashrate is the total number of guesses per second that the entire Bitcoin network makes while searching for a valid block. Every mining machine repeatedly runs the SHA-256 hashing function on slightly different versions of a candidate block, hunting for an output below a target set by the protocol. More hashes per second means more attempts, which means more of two scarce things committed to the chain: specialized silicon and electricity. That is why hashrate growth is read as a proxy for conviction. When it rises, capital and power are flowing in, and whoever is spending that money is betting it will pay off.

One thing to keep in mind before reading any hashrate chart: nobody measures the number directly. There is no meter on the network. Hashrate is estimated from two things we can observe, the difficulty and how fast blocks are actually arriving, using the relationship that hashrate is roughly equal to difficulty times 2 to the 32nd power, divided by the target block time of 600 seconds. Because blocks are found at random intervals, short windows are noisy, and different trackers publish different figures on the same day. That is why analysts lean on seven-day and thirty-day averages, and why this article quotes smoothed figures rather than a single spot reading.

The 2026 round trip: record, collapse, and back

To understand why the October recovery matters, you have to see the arc it reversed. Bitcoin’s hashrate climbed almost without interruption for a decade. It crossed 100 EH/s in January 2020, roughly eleven years after the genesis block. It first held one zettahash on a seven-day average in early September 2025, a milestone CoinDesk flagged as the first time the level stuck rather than briefly spiking. The peak came weeks later: about 1,156 EH/s on 18 October 2025, with difficulty topping out at 155.97 trillion on 29 October.

Then, for the first time in the network’s history, the climb reversed without a government forcing it. The hashrate fell back below one zettahash in mid-January 2026 and kept sliding, bottoming near 850 EH/s in early February, a peak-to-trough decline of about 27 percent on CoinShares’ numbers. CoinDesk noted the first-quarter drop was the first in six years. This was not the 2021 China ban or the 2018 and 2022 capitulations, when miners were forced off or priced out. Many simply chose to redirect power to AI and high-performance computing, which paid better. The number fell by decision, not by decree.

That is what makes the autumn rebound notable. The same network that voluntarily shrank has voluntarily grown back, clawing from 850 EH/s to nearly 1,090 EH/s and from a June difficulty low back to within touching distance of a record, in a matter of months. A contraction that took most of a year to bottom was substantially reversed in a single quarter.

Two retargets, nearly 12 percent: reading the October surge

Difficulty is Bitcoin’s thermostat. Every 2,016 blocks, roughly every two weeks, the protocol adjusts how hard it is to find a block so that blocks keep arriving about every ten minutes regardless of how much hashrate is online. When machines pour in, blocks come too fast, and the next adjustment raises difficulty to slow them back down. October produced two of the sharpest upward adjustments of the year, back to back.

The first, on 16 October, lifted difficulty 5.92 percent to 140.58 trillion, the third-largest increase of 2026 and the highest reading since March, news.bitcoin.com reported. The second, on 29 October, added another 5.54 percent to reach 148.36 trillion. Taken together they lifted difficulty almost 12 percent in a single month and pushed it to its highest level since late November 2025. For a network whose difficulty had fallen for most of the year, two back-to-back increases of that size are the clearest signal available that hashpower is returning in bulk.

The surge is already cooling at the edges. With only a fraction of the current epoch mined, early estimates for the next adjustment, due around 13 November, point to a small decline of roughly 2 to 3 percent, according to Pickaxe. That would not undo the recovery; it would simply mean the thermostat slightly overshot the hashpower that actually stuck around. Over the full year the network has logged 21 adjustments, 10 up and 11 down, a genuinely volatile stretch rather than a smooth climb.

Why the machines came back, and it was not only price

The cleanest account of the recovery comes from Luxor’s Hashrate Index, which framed the earlier decline as the product of three pressures that have since reversed. The first was price: Bitcoin recovered off a cycle low near $59,000 on 30 June, lifting mining revenue off the floor. The second was weather: a late-January Arctic blast across the United States, where most reported hashrate sits, had knocked machines offline, and conditions normalized through the year. The third was grid economics in Texas, where miners curtail during the ERCOT 4CP peak-demand windows; that curtailment season ended after a final peak on 12 September, freeing capacity to run flat out again. As Hashrate Index laid it out, when all three eased at once, the hashrate had room to climb back above a zettahash.

Revenue is the thread tying it together. Miners collectively earned about $1.215 billion in October, up 8.5 percent from September and the strongest month since November 2025, MillionMiner reported. News.bitcoin.com put the causal chain plainly: better mining revenue brought more computing power back into the competition, pushing the hashrate above one zettahash per second. When the math works, idle rigs get switched back on and new ones get plugged in. That is the ordinary physics of this market, and in October it finally ran forward.

The speed and breadth of the rebound are the tell. This was not one flagship operator energizing a single site; it was a broad re-entry across the fleet, which is why difficulty needed two outsized jumps to catch up. The marginal machine in Bitcoin mining is remarkably responsive: it can be powered down in an afternoon and powered back up just as fast, so the network expands and contracts with the margin rather than on multi-year capital cycles. That responsiveness cuts both ways, and it is exactly why the recovery arrived so abruptly after months of stagnation.

Hashrate rose while price and hashprice fell

Here is the part that breaks the simple mental model. The usual rule of thumb is that price leads and hashrate lags: coins get more valuable, miners earn more, more machines come online, and the hashrate follows weeks or months later. In October that lag became visible in an awkward way. Price was falling even as the hashrate climbed. BTC dropped 7.3 percent on the month, and hashprice fell from $40.24 on 1 October to $33.18 at the close, a 17.5 percent decline, MillionMiner data shows, yet difficulty jumped almost 12 percent.

Pickaxe named the dynamic directly: hashrate growth has outpaced price movements in recent periods, contributing to the compression in hashprice. Think of it as a delayed reaction. The machines arriving in October were ordered and financed during the late-summer rally, when BTC briefly touched the high $80,000s; they switched on regardless of where the price drifted afterward. More hashrate competing for the same fixed 450 BTC per day of block subsidy, against a lower coin price, mechanically squeezes what each unit of computing power earns. Growth and profitability pulled in opposite directions, which is exactly why a rising hashrate is a poor day-to-day price signal: it tells you about capital that was committed in the past, not about the next candle. For the business behind it, that squeeze is the whole story, as our look at mining margins in 2026 traces in detail.

The thermostat, running in reverse

JPMorgan’s Nikolaos Panigirtzoglou spent much of 2026 explaining why the hashrate fell, and his explanation doubles as a map of why it is rising now. Earlier in the year he told TFTC: “When bitcoin trades below its production cost, higher-cost miners power down, the hashrate declines, and difficulty adjusts lower.” His team estimated a difficulty-to-price sensitivity, a beta, of about 0.62, meaning difficulty tends to move a little more than half as much as price does over time.

Run that mechanism backward and you get October. Revenue climbed back above the cost line for a large share of the fleet, so the rigs that had powered down came back, and difficulty rose to meet them. The self-correcting loop that protected miners on the way down, by shedding the least efficient machines and lowering the bar, is the same loop that lets the network re-accelerate when conditions improve. It is worth stressing how robust this makes Bitcoin’s security model: difficulty does not need a committee or a vote to rebalance the supply and demand for block space. It simply responds, every 2,016 blocks, to whatever hashpower actually showed up.

Efficiency is the quiet engine of the comeback

If revenue is why miners wanted to grow, efficiency is how they could. The machines returning to the network in 2026 are not the machines that left in 2021. A decade ago, the Antminer S9 did about 98 joules per terahash. In 2026 the newest immersion-cooled units crossed below 10 joules per terahash for the first time in mass production: Bitmain’s Antminer S23 Hydro at roughly 9.5 J/TH and Bitdeer’s SealMiner A4 Ultra at about 9.45 J/TH. That is close to a tenfold efficiency gain in ten years.

Efficiency changes what growth even means. A fleet operator can retire old rigs and install new ones that produce far more hashes for the same power draw, so the hashrate can rise while the power bill stays flat. Increasingly the newest machines replace old fleets rather than expand the network outright, which is why the hashrate can approach records without the energy footprint doing the same; we unpack that split in our piece on Bitcoin’s energy and water footprint. The practical upshot for the recovery is that much of the hashpower that came back in October was not new power at all. It was better silicon running inside the same walls.

The AI pivot did not stop the climb

The dominant mining narrative of 2026 was that listed miners were abandoning Bitcoin for artificial intelligence, and the economics behind it are real. CoinShares estimates that AI and high-performance-computing hosting can generate roughly $1.5 million in annualized profit per megawatt, against about $0.5 million for mining, and the firm expects listed miners could draw up to 70 percent of their revenue from AI by the end of 2026, up from roughly 30 percent today. Riot Platforms signed a 20-year, 191-megawatt lease with Anthropic worth about $9.1 billion, CoinDesk reported, and TeraWulf, IREN and Core Scientific struck similar deals. Demand for that compute is not slowing, as our coverage of the race to build AI demand describes.

And yet the hashrate still recovered to within 8 percent of its record. The reason is that the AI diversion and hashrate growth are not a simple tug of war. Listed miners are only part of the network, the efficiency gains above let the remaining operators do far more with the same power, and the private, offshore and sovereign miners who do not issue press releases kept building. James Butterfill, CoinShares’ head of research, has called this one of the most challenging periods the mining industry has faced, and his firm still projects the network could reach 1.8 zettahashes by the end of 2026 if the price cooperates. The AI pivot bent the growth curve; it did not break it.

Some in the industry see the two businesses converging rather than competing. Fred Thiel, chief executive of MARA, has argued that power, not hashing, is the real prize: “By 2028, you’ll either be a power generator, be owned by one, or be partnered with one,” he told CoinGeek. In that view the megawatts a miner controls are the durable asset, and whether they point at Bitcoin or at an AI cluster in any given month is a routing decision that flips with the margins. October was a month when the routing favored Bitcoin.

Who actually brought the hashrate back

It is tempting to credit the big public miners, but they are a shrinking minority of the network. By CoinShares’ second-quarter 2026 accounting, the realized hashrate of listed miners was around 319 EH/s, roughly a third of the total, and their share has been falling as they tilt toward AI. The rest, the majority, is the part you cannot see on an earnings call: private farms, offshore operators, state-linked sites, and the idle rigs that re-energized the moment the margin math turned positive. Much of October’s rebound came from that invisible majority switching back on, not from a single company flipping a switch.

Mining pools are where that scattered hashpower gets coordinated, and the pecking order keeps shifting. In early November, AntPool led with about 23.6 percent of network hashrate, followed by Foundry USA at 21.5 percent and F2Pool at 18.75 percent, per Pickaxe. A piece of plumbing is changing too: Binance Pool is scheduled to shut down on 10 November, and its hashrate will redistribute to rivals. Pools coordinate work and smooth payouts; they do not own the machines, and the miners pointing hashpower at them can repoint it elsewhere at will. That detail is what keeps pool concentration from being the same thing as network control, and it is why the top-three share tells you less about who runs Bitcoin than it first appears.

What a near-record hashrate buys: security

The reason anyone outside the mining business should care about hashrate growth is security. A higher hashrate means anyone who wanted to rewrite recent transaction history by out-mining the honest network would have to marshal more machines and more power to do it. At current levels, the network’s annual security budget, the sum of block subsidy and fees paid to miners, runs on the order of $13 billion to $14 billion a year (450 BTC per day at roughly $79,600, plus transaction fees).

Campbell Harvey, the Duke finance professor, has put a sharper number on the cost of a 51 percent attack. He estimates it at about $8 billion, or “about 50 basis points of the value of bitcoin,” and warns that “the difference today is the derivatives markets,” because an attacker could short Bitcoin derivatives and profit from the crash the attack would cause rather than needing to recoup the cost through mining. Not everyone is convinced the threat is practical. Matt Prusak, president of American Bitcoin, told crypto.news that “economic feasibility kills the 51% thesis,” arguing that quietly acquiring that much hardware and power is implausible in the real world. Either way, the arithmetic moves in defenders’ favor as the hashrate grows: a recovering network is a more expensive network to attack.

It is also worth remembering what the security budget has to survive. Bitcoin’s block subsidy halves roughly every four years, with the next cut due in 2028, so over time a larger and larger share of the miner revenue that funds this security has to come from transaction fees rather than new issuance. A near-record hashrate funded mostly by the subsidy is comforting now; the open question for the next decade is whether fee revenue grows fast enough to keep the attack cost high as the subsidy shrinks. Bitcoin’s main chain has never been successfully 51 percent attacked, unlike smaller proof-of-work coins such as Ethereum Classic and Bitcoin Gold, and a rising hashrate is the most direct reason that record holds.

Growth has a cost: the hashprice squeeze

For the people running the machines, a near-record hashrate is a mixed blessing. The flip side of more competition is thinner pay. Hashprice sat around $34 per petahash per day after the 29 October adjustment, against about $46 at the previous hashrate all-time high, Hashrate Index noted. Same near-record hashrate, far less revenue per unit, because the coin price is lower and the difficulty is higher. It is still about 20 percent above June’s record monthly low, so the picture is one of recovery rather than crisis, but the margin of safety is slim and the least efficient rigs are running close to break-even.

That is the quiet tension inside every hashrate recovery. Growth is good for security and good for the network’s credibility, but it is paid for by the operators whose margins it compresses. When the hashrate races ahead of price, as it did in October, the oldest machines are the first to feel it, and a stretch of sub-break-even hashprice is what eventually forces some of them back off, which is how the difficulty thermostat finds its level again. The cycle that produced the recovery contains the seed of the next pullback. The same push and pull drives the longer arc we explore in Bitcoin’s halving-cycle math.

Is a new all-time high next?

The honest answer is that the network is close enough that the question is reasonable for the first time in a year. Difficulty is 4.9 percent below its record and the hashrate is roughly 8 percent below its peak. Two or three more adjustments the size of October’s would do it. But the early estimate for the 13 November retarget points to a small decline, a reminder that the surge has already slightly outrun the hashpower that stayed online. Records are not set by a single hot month; they are set when the margin math holds long enough for the hashpower to stick.

What would push the network to a fresh record is the same thing that started the recovery: durable revenue. CoinShares’ 1.8-zettahash target for the end of 2026 was always contingent on Bitcoin recovering toward six figures, and with BTC near $79,600 that condition is not met. A sustained move back above the mid-$80,000s, helped along by the kind of spot-ETF demand we track in our coverage of the crypto ETF wave, would pull the remaining efficient rigs online and likely set a new high. A breakdown in price, or another leg of the AI diversion, would stall it again. For once, the hashrate is not waiting on a distant forecast; it is one or two good months of price away from a record.

What the recovery signals, and how to read the number

For US readers, the regulatory backdrop here is unusually settled for crypto. In March 2025 the SEC’s Division of Corporation Finance stated that proof-of-work mining does not involve the offer or sale of securities, because miners earn rewards through their own efforts rather than the efforts of others. Mining itself is not a licensed securities activity in the United States; the SEC’s attention remains on staking and on the companies that package mining into investment products, not on the act of hashing. That clarity is part of why the US still hosts the largest reported share of the network, and why the recovery was centered there.

Two takeaways for anyone watching the number climb. First, read it as a statement about committed capital and conviction, not as a price signal; October proved the two can move in opposite directions for weeks at a time. Second, treat every record claim as an estimate, because it is: the hashrate is inferred, not measured, so prefer seven-day and thirty-day averages to any single spot reading, and be skeptical of trackers that flash an implausible all-time high on a noisy day. What is not in doubt is the direction. After the first voluntary contraction in Bitcoin’s history, the machines came back on their own, and the network is once again knocking on the door of a record it set a year ago.

Frequently Asked Questions

What is Bitcoin’s hashrate right now in November 2026?

Bitcoin’s seven-day average hashrate was around 1,069 EH/s in early November 2026, having peaked near 1,088 EH/s on 26 October. That is roughly 8 percent below the all-time high of about 1,156 EH/s set on 18 October 2025. Mining difficulty stood at 148.36 trillion, about 4.9 percent below its record.

Why did Bitcoin’s hashrate rise while the price fell?

Hashrate lags price. The efficient machines that came online in October were ordered during the late-summer rally, so they switched on even after BTC slipped. More hashpower chasing the same fixed daily block reward, against a lower coin price, compressed hashprice. Growth reflects capital committed earlier, not where the price is today.

What drove the 2026 hashrate recovery?

Luxor’s Hashrate Index points to three pressures reversing at once: Bitcoin’s price recovered off its June low, the January weather disruptions normalized, and the ERCOT 4CP curtailment season in Texas ended after mid-September. Stronger mining revenue, the best month since November 2025, pulled idle and new rigs back online.

Will Bitcoin’s hashrate set a new all-time high in 2026?

It is within reach, roughly 8 percent for hashrate and 4.9 percent for difficulty, but not guaranteed. CoinShares’ 1.8-zettahash year-end target assumed BTC recovering toward $100,000, which has not happened. A durable price move back above the mid-$80,000s would likely bring enough efficient hashpower online to set a record; a price breakdown would stall it.

Does a higher hashrate make Bitcoin more secure?

Yes. A higher hashrate raises the cost of a 51 percent attack, since an attacker must out-compute the honest network. Duke’s Campbell Harvey estimates such an attack would cost about $8 billion today. Bitcoin’s annual security budget, subsidy plus fees, currently runs around $13 billion to $14 billion.

By Marcus Okafor, senior mining and markets correspondent at HOGE Wire.