Bitcoin mining difficulty fell 10.09% on June 13, 2026 — the 11th-largest downward adjustment in network history and the second-largest of 2026. The decline, triggered at block height 953,568, dropped difficulty from 138.96 trillion to 124.93 trillion, its lowest level since July 2025. Network ha...
"The miners winning in 2026 are the ones that closed hyperscaler deals first. Power capacity alone isn't enough — the market is pricing contracted backlog, delivery timelines, and the quality of counterparties." — James Butterfill, Head of Research, CoinShares
Bitcoin mining difficulty fell 10.09% on June 13, 2026 — the 11th-largest downward adjustment in network history and the second-largest of 2026. The decline, triggered at block height 953,568, dropped difficulty from 138.96 trillion to 124.93 trillion, its lowest level since July 2025. Network hashrate slid from above 1,000 EH/s to approximately 893–918 EH/s as a 15% BTC price decline in June squeezed miner margins below breakeven.
The adjustment lands in a sector already undergoing structural transformation. Public mining companies have collectively announced over $70 billion in AI and high-performance computing (HPC) contracts. Core Scientific, TeraWulf, and IREN now derive between 60–70% of revenue from non-mining operations. Mining stocks have outperformed BTC by roughly 70% year-to-date, with markets pricing these firms as power-infrastructure plays rather than pure bitcoin producers. The separation between miners who secured hyperscaler contracts and those who did not is now the defining fault line of the industry.
At block 953,568, Bitcoin's difficulty algorithm executed a -10.09% adjustment, resetting the mining difficulty from 138.96 trillion to 124.93 trillion. The epoch preceding the adjustment took 15.6 days to complete — well beyond the standard 14-day target interval — indicating that miners were leaving the network faster than new capacity was connecting.
The proximate cause: Bitcoin's spot price fell approximately 15% during June, from above $76,000 at the start of the month to approximately $63,500–$66,500 in mid-June. On June 15, BTC traded at $65,695, rebounding slightly on news of a U.S.-Iran ceasefire agreement. The price decline compressed miner revenue per hash unit below the operating cost threshold for a significant portion of the fleet.
The adjustment immediately improved economics for remaining miners. Active equipment now generates approximately 9–11% more bitcoin per unit of hashrate. Hashprice — the estimated daily revenue per petahash per second — climbed 13% from a trough in the high $20s to approximately $32.31/PH/s/day, according to the Hashrate Index.
The CoinShares Q1 2026 Bitcoin Mining Report placed the weighted average cash cost to produce one bitcoin among publicly listed miners at approximately $79,995–$90,000. With BTC trading around $65,700, the average listed miner is losing roughly $14,000–$24,000 per coin produced on an all-in economic basis.
The breakeven line is shaped by three variables:
These economics explain the hashrate exodus. When revenue per hash falls below the marginal cost of electricity, rational operators shut down machines. The 10% difficulty decline is the network's self-correcting mechanism working as designed.
The mining sector's response to margin compression has not been primarily operational — it has been strategic. Over $70 billion in cumulative AI and HPC contracts have been announced across the public mining sector since 2025.
Three companies illustrate the scale of reallocation:
Core Scientific (CORZ) signed high-density colocation contracts with CoreWeave representing over $10 billion in potential revenue across a twelve-year term, covering approximately 590 MW under contract. In Q1 2026, colocation revenue reached $77.5 million out of $115.2 million in total revenue — approximately two-thirds of income now comes from AI-related hosting rather than mining. Core Scientific sold $175 million of its bitcoin holdings to accelerate the transition, reducing its treasury from 2,537 BTC to roughly 630 BTC.
TeraWulf (WULF) has locked in $12.8 billion in contracted AI revenue. In Q1 2026, it generated $21 million in HPC revenue out of $34 million total. Management has stated its intention to exit bitcoin mining operations entirely by year-end. Approximately 27% of revenue comes from AI currently, projected to reach 70% by December.
IREN holds a Microsoft AI cloud partnership and a 4.5 GW power pipeline. The company targets more than $3.4 billion in annualized AI cloud revenue by the end of 2026, supported by an expansion to 140,000 NVIDIA GPUs across its development pipeline. HPC revenue is projected to reach 71% of total by year-end.
The economic logic is straightforward. These companies sit on assets that hyperscalers need: permitted sites with grid connections, low-cost power purchase agreements, cooling infrastructure, and power-engineering expertise. An AI GPU rack generates materially higher revenue per megawatt than a bitcoin ASIC rack at current hashprices. The pivot does not require abandoning existing infrastructure — it requires repurposing it.
The market has priced the bifurcation clearly. Mining stocks have outperformed BTC by roughly 70% year-to-date. The sector split into two distinct cohorts:
Infrastructure-pivoted miners (WULF, CORZ, HUT) posted YTD gains of 40–73%:
Pure-play miners (MARA, CLSK, RIOT) have tracked closer to BTC's price, which sits approximately 12% lower since January 1.
The divergence reflects a repricing event. Markets no longer value these companies primarily on hashrate, bitcoin production, or BTC treasury holdings. The valuation driver is contracted AI revenue backlog, power-capacity pipeline, and counterparty quality. TeraWulf's market capitalization exceeds several miners with substantially higher hashrate — because the market is pricing it as a power-and-AI-buildout story.
CleanSpark represents an exception among pure miners. Its February 2026 operating update reported 50.0 EH/s operational hashrate with fleet efficiency of 16.07 J/TH — among the best in the public mining sector. The company has maintained mining-focused positioning while achieving operational metrics that keep it competitive at current hashprices.
The Hash Ribbon indicator — which tracks the crossover between the 30-day and 60-day moving averages of hashrate — has been signaling miner capitulation for approximately three months. This is one of the longest capitulation periods on record.
Capitulation occurs when mining revenue drops below operating costs, forcing less efficient operators to shut down machines and liquidate BTC reserves to cover electricity, debt service, and overhead. Bitcoin is now trading below its estimated average production cost, a condition that last occurred in November 2022 when BTC bottomed near $15,500.
Historically, the end of Hash Ribbon capitulation periods has correlated with local or major bitcoin price bottoms. Following the mid-2024 yen carry trade unwind, a Hash Ribbon capitulation signal preceded bitcoin's move from approximately $49,000 to $100,000 over the subsequent five months. The current signal is approaching a potential recovery crossover, though the timing remains uncertain.
The pattern is consistent with economic theory: once unprofitable hashrate exits and difficulty adjusts downward, remaining miners achieve improved unit economics, reducing forced selling pressure. The question is whether this cycle follows the historical pattern or whether the structural AI pivot removes enough hashrate permanently to alter the dynamic.
MARA Holdings — formerly Marathon Digital, previously the loudest advocate of a strict bitcoin accumulation strategy — sold 20,880 BTC for $1.5 billion in Q1 2026. The sales occurred in two phases:
Between March 4 and March 25, MARA sold 15,133 BTC for approximately $1.1 billion, directing $1 billion toward convertible note repurchases that reduced outstanding debt from $3.3 billion to $2.3 billion. The remaining $400 million in sales occurred during the balance of Q1.
The company reported a $1.26 billion net loss in Q1 2026, more than double its $533 million loss a year earlier. Revenue fell 18% year-over-year to $175 million. Despite the liquidation, MARA retains 35,303 BTC — the fourth-largest corporate bitcoin treasury — with approximately 28% of holdings loaned or pledged as collateral.
MARA's strategic shift is significant because it signals that the "mine-and-hold" model has reached its limits at current price levels. When the cost to produce bitcoin consistently exceeds the spot price, accumulation becomes leveraged long exposure funded by operating losses. MARA has also begun exploring AI infrastructure at its Long Ridge facility, though it remains further behind the pivot curve than Core Scientific or TeraWulf.
The simultaneous hashrate decline and AI pivot raise a question about Bitcoin's long-term security budget. If mining companies permanently reallocate power capacity from ASICs to GPUs, the network's security model depends on two factors: whether remaining miners can sustain sufficient hashrate, and whether transaction fees can supplement the declining block subsidy.
At 3.125 BTC per block and approximately 450 BTC mined daily, the current annual block subsidy at $65,700/BTC is roughly $10.8 billion. This figure halves again in April 2028. If a significant portion of the mining industry's power infrastructure migrates permanently to AI workloads, the network hashrate may stabilize at a lower level than previously projected.
The difficulty adjustment mechanism ensures that the network continues to function — blocks will be produced regardless of total hashrate. The economic question is whether the equilibrium hashrate provides sufficient security for a $1.3 trillion asset. This is not an immediate crisis, but the structural trajectory is worth tracking.
The June 2026 difficulty adjustment is not an anomaly. It is the quantitative expression of a structural shift in bitcoin mining economics. The April 2024 halving compressed revenue. The subsequent price decline from $109,000 to $65,700 compressed it further. And the emergence of AI infrastructure demand has given mining companies a higher-margin alternative for their core asset — cheap power at scale.
The industry is splitting into two businesses that happen to share physical infrastructure. One produces bitcoin. The other rents power and cooling to hyperscalers. The market has made its preference clear: infrastructure-pivoted miners are trading at 40–73% YTD gains while BTC itself is down 12%.
For Bitcoin's network, the difficulty adjustment mechanism ensures continued block production. The longer-term question — whether the security budget remains adequate as miners redirect capacity — will depend on BTC price recovery, fee market development, and whether the AI pivot proves to be a permanent reallocation or a cyclical hedge. The data so far suggests the former.