Bitcoin's Difficulty Drops 5% in the Month's Second Big Cut
Bitcoin's mining difficulty fell 5.00% at block 957,600 on July 11, dropping from 133.87 trillion to 127.17 trillion and taking roughly 6.7 trillion off the target that miners have to beat. It was the network's fourteenth difficulty adjustment of 2026, and the eighth of those to move down; six have moved up. The cut followed a fast decline in computing power over the first ten days of the month, with the seven-day average hashrate sliding from 986 exahashes per second on July 1 to about 908 by the day of the retarget, a fall of 7.9%. It is the second large downward move in under a month, after a 9.91% cut on June 13, and it leaves difficulty 14.22% below where it started the year on January 8.
The Mechanism: A Self-Correcting Target Reset Every Two Weeks
Mining difficulty is the number that sets how hard it is to find a valid block. The protocol recalculates it every 2,016 blocks — about two weeks at the intended ten-minute block interval — by comparing how long those blocks actually took against how long they should have taken. If machines have left the network, blocks arrive slowly and the next adjustment lowers the target; if machines have joined, blocks arrive fast and difficulty rises. Nothing about the schedule is discretionary, which is why the number is read as a hardware census taken every fortnight. The July 11 reset was the arithmetic consequence of the hashrate that had already gone: about 78 exahashes per second of capacity stopped hashing between July 1 and the retarget.
The Pattern: Two Big Cuts in a Month, With a Rise in Between
The two downward moves are not consecutive retargets. June's 9.91% cut, one of the largest on record, took difficulty to 125.19 trillion, and enough capacity returned over the following weeks to carry it back above 133 trillion before the July reset pulled it down again. That round trip is the more informative part. Difficulty rising and falling by several percent inside a month says the marginal machine is being switched on and off according to what it earns that week, rather than running continuously the way an owned, fully paid fleet would. Eight negative adjustments against six positive over fourteen resets describes the same thing across the year: a network whose installed base has been shedding its least efficient hardware, with intermittent capacity filling the gap when conditions allow.
The Economics: What the Machines Were Earning When They Left
Hashrate follows revenue with a lag, and the revenue is a straightforward calculation: a fixed block subsidy plus fees, divided among everyone hashing, measured against power and hosting costs. Coverage of June's cut named two drivers: unprofitable rigs powering down, and the pivot by mining companies into hosting AI and high-performance computing workloads, which is a decision to sell the same power and data-centre space to a different customer. The June retarget landed with bitcoin near $64,007 and the Fear and Greed index at 13, deep in fear; that was the price miners were selling their production into. One report circulating around the June cut put the hashrate ahead of it at 918 exahashes per second, below the roughly 967 cited elsewhere, so the size of the exit is not agreed on.
What It Means for the Miners Who Stayed: Hashprice at $31.1
Difficulty falling is a transfer to the operators still running. Hashprice — the daily revenue a miner earns per petahash per second of computing power — rose 12.5% to $31.1 on the easing, because the same block rewards were being split among less competing hardware. That is a real margin improvement for anyone whose costs did not change, and it is also how the network stabilises: cheaper mining draws capacity back, which raises difficulty again and closes the gap. The limit on the relief is visible in the level itself. At $31.1, hashprice remains far below the $49.4 peak of October 2025, so a 5% cut restores a fraction of what the past nine months removed, and the machines that moved to AI and high-performance computing hosting are unlikely to be the ones that come back at the next retarget.