Bitcoin difficulty drops 10.09%, its 11th-largest decline ever

June 14, 2026 — Bitcoin's mining difficulty fell 10.09% at block 953,568 on June 14, dropping from 138.96 trillion to 124.93 trillion, the 11th-largest downward adjustment in the network's history and the second-biggest negative retarget recorded in 2026.

The prior epoch ran 15.6 days against the 14-day target. Each 2,016-block cycle ends with the network measuring how far block production deviated from the ten-minute design interval, then resetting the required work for the next epoch proportionally. The May-to-June epoch arrived 1.6 days late because hashrate was leaving. Bitcoin's network hashrate, which topped 1,000 exahashes per second in late May, fell to approximately 893 EH/s ahead of the adjustment. Blocks that should average ten minutes were arriving at more than eleven.

This is where the adjustment moves from routine to significant. Miners were running well below profitability. Checkonchain's production-cost model estimated average mining cost at approximately $84,300 per BTC as of mid-June, against a spot price around $63,000. A roughly $21,300 per-coin loss forced operators to a choice: curtail operations, shut off inefficient machines, or redeploy capacity. Many did all three.

The displacement to other uses is what distinguishes this adjustment from a straightforward cyclical squeeze. As CoinDesk reported in March, publicly listed mining firms including Core Scientific, TeraWulf, and Hut 8 have signed multi-year, multi-billion-dollar agreements to convert power capacity and computing infrastructure to AI data centre hosting. Core Scientific locked in a 12-year, $10.2 billion contract with CoreWeave. TeraWulf has $12.8 billion in contracted HPC revenue. Hut 8 signed a 15-year, $7 billion lease for AI infrastructure at its River Bend campus. Per CoinDesk, those contracts deliver multi-year revenue visibility at margins well above what Bitcoin mining has produced in 2026, and the income is independent of BTC price, difficulty, or block reward cycles. Hashrate moved to AI workloads does not return when the price recovers. The machines are committed under binding agreements. That makes the current drawdown partly structural, not wholly cyclical.

Bitcoin's price fell roughly 15% through June, amplifying the pressure. The combination of depressed prices and AI-driven exits pushed the epoch above 15 days, enough to produce a cut of this size. The February 7 adjustment, this year's prior record at 11.16%, came as the post-halving mining buildout collided with compressed margins. The June 14 cut arrived with additional weight: more capacity committed to AI, and a price still in the low-$60,000s rather than recovering.

The mechanism is working as designed. Every 2,016 blocks, a deterministic calculation resets the proof-of-work target. No committee votes. No operator intervenes. The epoch ran slow, the algorithm registered the deviation, and it cut 10.09%. A faster epoch next time, with more hashrate and blocks arriving under ten minutes, will push difficulty back up. The protocol is indifferent to who turned their machines off or why.

For miners still online, the adjustment is an immediate improvement. Hashprice (revenue per petahash per second per day) recovered to approximately $32.31 after the cut, up from below $28 in the days immediately before the retarget. Each remaining unit of hashrate is now producing roughly 11% more BTC at equivalent power draw. That narrows the gap between Checkonchain's $84,300 production-cost estimate and the $63,000 spot price, though it does not close it. Most active miners remain in loss territory on a per-coin basis. The adjustment reduces the depth of that loss, not the loss itself.

The historical context calibrates the scale. The June 14 difficulty of 124.93 trillion is the lowest point of 2026 and the weakest reading since July 12, 2025, an 11-month low. The year's peak of 146.47 trillion, set in January, is now more than 21 trillion above the current level. Across the 12 epochs completed since January, seven have been negative and five positive, a downward skew that reflects post-halving price compression and a structural shift in how mining firms allocate capacity.

The security implication is real but bounded. Bitcoin's hashrate at 893 EH/s remains historically elevated. The difficulty algorithm prevents chain stalls regardless of how many miners exit. What erodes in a sustained below-cost environment is not chain function but security margin: the cost an adversary would need to exceed to mount a majority attack. A network maintained by a smaller and more concentrated pool of operators running thin margins offers modestly less resistance than the same hashrate spread across a broader independent base. The question is how long the below-cost environment persists and how much additional capacity AI conversions remove before economics change.

The next retarget epoch is projected around June 28. If current hashrate holds near 893 EH/s and blocks average close to ten minutes, the following adjustment will be near neutral. If exits continue through further price pressure or additional AI contract conversions, blocks slow again and another negative retarget follows. No material price recovery is in evidence at this writing.

---

Verified claims