Bitcoin miners are losing approximately $19,000 on every coin they produce. With production costs near $90,000 per BTC and spot prices hovering around $60,000, the mining industry has entered its deepest profitability crisis since the April 2024 halving. Hashprice — the standard measure of daily ...
"The ability for miners to convert to AI is one of the largest infrastructure shifts of this decade." — Adam Sullivan, CEO, Core Scientific
Bitcoin miners are losing approximately $19,000 on every coin they produce. With production costs near $90,000 per BTC and spot prices hovering around $60,000, the mining industry has entered its deepest profitability crisis since the April 2024 halving. Hashprice — the standard measure of daily mining revenue per petahash — fell below $30/PH/s/day in early 2026, a five-year low. According to CoinShares, 15% to 20% of legacy mining rigs are now operating at a cash loss.
The response has been structural, not cyclical. Publicly listed miners have collectively signed more than $70 billion in AI and high-performance computing (HPC) contracts, liquidated over 15,000 BTC from their treasuries, and redirected hundreds of megawatts of power capacity toward GPU clusters. Bitcoin's network hashrate posted its first quarterly decline in six years during Q1 2026. Mining difficulty dropped 10.09% on June 14 — the second-largest negative adjustment of the year. The industry that secures a $1.2 trillion network is becoming, by revenue composition, an AI infrastructure business.
The weighted-average cash cost to produce one bitcoin among publicly listed miners rose to approximately $79,995 in Q4 2025, according to CoinShares' Q1 2026 Mining Report. By early 2026, factoring in incremental electricity cost increases and network difficulty, that figure climbed toward $90,000. Bitcoin traded at approximately $60,100 on July 1, 2026 — roughly 33% below the breakeven threshold for the average public miner.
The margin compression traces directly to two forces: the April 2024 halving, which cut block rewards from 6.25 BTC to 3.125 BTC, and the sustained decline in bitcoin's price from its $126,000 all-time high in October 2025. The halving effectively doubled production costs overnight. Average production cost per bitcoin among public miners rose from approximately $16,800 pre-halving to $37,856 immediately after, according to JPMorgan research. Post-halving bull market conditions temporarily masked this structural shift. The subsequent 52% price decline from $126,000 to below $60,000 exposed it.
Hashprice — daily USD revenue earned per petahash of computing capacity — peaked near $60/PH/s/day during the October 2025 price surge before collapsing to approximately $28/PH/s/day by late February 2026. It has since stabilized in the $30–$35 range, still well below the level required for most operators to generate positive cash flow. Mining ROI timelines have extended to approximately 1,000 days for new hardware deployments, according to CCN analysis, compared with 400–500 days during the 2024–2025 bull run.
Electricity costs are the primary determinant of survival. Miners operating at $0.05/kWh or below remain viable. Those paying $0.08/kWh or above face shutdown economics. With electricity accounting for up to 90% of operating costs, the 2026 energy price environment — elevated by Middle East-driven oil price increases — has turned power efficiency into the single most important competitive variable.
Bitcoin's total network hashrate reached multiple all-time highs above 1 Zettahash per second (ZH/s) in January 2026, a milestone that represented years of uninterrupted growth in mining infrastructure deployment. That growth reversed in Q1 2026. According to CoinDesk, Bitcoin's hashrate posted its first quarterly decline since 2020, falling approximately 4% year-to-date to hover around 1 ZH/s.
The decline is not weather-related or temporary. It reflects a deliberate reallocation of capital and power capacity away from bitcoin mining and toward AI infrastructure, according to CoinShares. Industry hashrate growth forecasts have been revised downward. The network was previously projected to reach 2.5+ ZH/s by Q1 2027; that estimate has been reduced to approximately 1.8 ZH/s by end of 2026, according to BlockEden analysis.
The hashrate stagnation marks a departure from bitcoin mining's historical pattern of relentless capacity expansion through every price cycle. Even during the 2022 bear market, when bitcoin traded below $17,000, hashrate continued to rise as miners deployed pre-ordered hardware and locked in power contracts. The 2026 dynamic is different: miners are not just pausing deployment — they are actively converting existing capacity to non-mining uses.
The scale of the mining industry's pivot to AI infrastructure is measured in tens of billions. More than $70 billion in cumulative AI and HPC contracts have been announced across publicly listed mining companies, according to multiple industry reports. The economics explain why: AI training clusters are willing to pay 3–5x more per megawatt-hour than bitcoin mining operations generate, with multi-year contracted revenue that mining's variable block-reward model cannot match.
The largest individual contracts include:
More than 600 megawatts of long-term AI hosting deals were signed between late September 2025 and March 2026, involving hyperscalers including AWS, Google-backed Fluidstack, and Microsoft. According to Bloomberg reporting from April 2026, AI revenue is set to surpass bitcoin mining revenue for leading crypto mining companies, with some firms projected to derive up to 70% of total revenue from AI operations by end of 2026.
MARA Holdings CEO Fred Thiel framed the strategic rationale: "By 2028, you'll either be a power generator, be owned by one, or be partnered with one." Riot Platforms CEO Jason Les described Q1 2026 as an "inflection point," signaling its transition toward a revenue-generating data center operator rather than a pure mining company.
The firms furthest along in the transition — IREN, TeraWulf, Cipher, Core Scientific, Hut 8, and Applied Digital — together hold tens of billions in contracted AI revenue and have effectively become hybrid infrastructure operators. JPMorgan upgraded Cipher and CleanSpark while trimming MARA and Riot price targets, reflecting the market's differentiation between miners executing the AI pivot and those still primarily dependent on block rewards.
To fund the capital-intensive pivot to AI, miners have been systematically liquidating their bitcoin reserves — a reversal of the "HODL" treasury strategy that characterized the 2023–2025 bull market.
Documented treasury sales include:
Publicly listed miners have collectively reduced their bitcoin holdings by over 15,000 BTC from peak levels, according to CoinDesk reporting. The proceeds are being directed primarily toward GPU procurement, data center construction, and power infrastructure upgrades required for AI workloads.
Bitcoin's difficulty adjustment algorithm — which recalibrates every 2,016 blocks (approximately two weeks) — has produced significant negative adjustments in 2026, reflecting the departure of unprofitable mining capacity:
The current difficulty stands at approximately 133.87 trillion as of early July 2026. The next retarget is estimated for July 10.
The self-correcting mechanism functions as designed: when unprofitable miners shut down, block production slows temporarily, difficulty adjusts downward, and the surviving operators earn proportionally more bitcoin per unit of hashpower. Remaining miners now earn approximately 11% more bitcoin per unit of active computing power following the June adjustment. However, the dollar value of that additional bitcoin output remains below production cost for many operators — the adjustment reduces the hashrate required to mine, but it does not increase the USD price of the output.
The industry transformation raises a question about Bitcoin's long-term security model. The network's security budget — the total economic incentive for miners to honestly validate transactions — depends on block rewards and transaction fees. With block rewards scheduled to halve again in 2028 (to 1.5625 BTC per block), and the current subsidy already insufficient to cover production costs at present prices, the fee market must eventually compensate for declining block subsidies.
The AI pivot introduces an additional variable. As miners diversify revenue streams, their dependency on bitcoin mining revenue decreases. Operators with profitable AI business lines may continue to mine at marginal or negative returns on the mining operation alone, subsidized by AI revenue. Alternatively, they may further reduce mining capacity in favor of higher-return AI workloads.
The concentration dynamic also matters. As smaller and less efficient miners exit, hashrate concentrates among larger, better-capitalized operators. While this may improve average fleet efficiency, it reduces the number of independent entities securing the network. Whether Bitcoin's security model adapts to this new structure — through increased transaction fees, protocol-level changes, or simply higher bitcoin prices — remains an open question.
The Bitcoin mining industry is undergoing a structural transformation that no difficulty adjustment can reverse. The economics are unambiguous: AI workloads pay 3–5x more per megawatt-hour than bitcoin mining, with contractual revenue certainty that block rewards do not provide. The $70 billion in signed AI contracts represents a permanent reallocation of power infrastructure away from proof-of-work computation.
For Bitcoin's network, the near-term impact is manageable. Difficulty adjustments ensure block production continues regardless of how many miners exit. The longer-term question is whether sufficient economic incentive remains to attract the hashrate necessary to secure a trillion-dollar network as block subsidies continue to decline. The 2028 halving will provide the next stress test. The mining industry that faces it will look fundamentally different from the one that entered 2026.