Bitcoin's mining industry is undergoing its most severe structural transformation since the network launched in 2009. Hashprice — the standard metric for mining revenue per unit of computational power — fell to $27.89/PH/s/day in June 2026, a post-halving record low and roughly 66% below October ...
"Bitcoin mining is a tool to bring power capacity forward. Mining is the trojan horse that gets us interconnection rights so we can host AI compute." — Jason Les, CEO, Riot Platforms
Bitcoin's mining industry is undergoing its most severe structural transformation since the network launched in 2009. Hashprice — the standard metric for mining revenue per unit of computational power — fell to $27.89/PH/s/day in June 2026, a post-halving record low and roughly 66% below October 2025 peaks. The CoinShares Q1 2026 mining report found 15-20% of legacy hardware is now operating at a loss. Publicly traded miners sold more than 32,000 BTC in Q1 2026 alone, a single-quarter record exceeding combined 2025 sales.
The response from listed mining operators has been uniform: pivot to artificial intelligence. Over $70 billion in aggregate GPU co-location and cloud service contracts have been signed with hyperscalers since late 2025. CoinShares projects 70% of listed miner revenue will derive from AI hosting by end-2026. What was once a single-product industry — converting electricity to Bitcoin — is bifurcating into a dual-revenue infrastructure play where BTC mining serves as a flexible baseload that underwrites fixed power procurement costs, while AI compute generates the margin.
The April 2024 halving reduced the block subsidy from 6.25 BTC to 3.125 BTC. Two years later, the full economic impact has arrived. At approximately 144 blocks per day, the network produces roughly 450 BTC in daily subsidies. Combined with $3-4 million in daily transaction fees, total miner revenue runs approximately $33 million per day across the entire network.
Hashprice — daily revenue per petahash of deployed computational power — declined 17.8% over the 30 days ending in late June 2026, falling from $37.44/PH/s/day to $30.77/PH/s/day. It touched $27.89/PH/s/day in late June, a five-year low comparable to post-COVID levels in 2020, according to CoinShares.
The fully loaded cost to mine one Bitcoin — including depreciation and stock-based compensation — has surged to approximately $137,800, according to CryptoRank and public financial disclosures from listed miners. The average cash cost stands at roughly $74,600. With BTC trading near $62,800, cash-positive operations require sub-5 cent/kWh electricity and latest-generation hardware rated below 15 J/TH.
An Antminer S21 XP (13.5 J/TH) breaks even at $0.088/kWh. An S23 Hydro (9.5 J/TH) breaks even at $0.124/kWh. Operators running anything less efficient than mid-generation hardware are underwater at prevailing power rates.
The network breached 1 zettahash per second (ZH/s) in January 2026 — a milestone equivalent to 1,000 exahash (EH/s). That peak proved unsustainable. Approximately 252 EH/s of older hardware switched off as margins compressed, and the 30-day average hashrate fell close to 6% to roughly 1,004 EH/s by late June 2026.
Bitcoin completed its 11th-largest difficulty decline in history at block height 953,568, with difficulty falling 10.09% from 138.96T to 124.93T. This was the second-largest downward adjustment of 2026 and the steepest since the China mining ban in 2021. Current difficulty stands at approximately 133.87T, with the next retarget scheduled for July 11, 2026.
The Hash Ribbons indicator — which compares the 30-day and 60-day moving averages of hashrate — flagged miner capitulation through much of early 2026. CoinDesk reported in February that the three-month capitulation signal was one of the longest on record, with BTC trading below its average production cost for the first time since November 2022. According to CryptoQuant, the most recent Hash Ribbons data suggest the capitulation phase is approaching its end, a signal that has historically preceded price recoveries: January 2019 (+56% in 90 days), March 2020 (+82% in 90 days), and July 2023 (+25% in subsequent weeks).
This is the halving cycle working as designed. Difficulty adjusts downward, marginal producers exit, survivors capture a larger share of block rewards at lower competition levels, and the network rebalances.
The profitability squeeze forced aggressive treasury management. Publicly traded Bitcoin miners sold more than 32,000 BTC in Q1 2026, a single-quarter record.
Specific examples from public filings:
Collectively, public miners reduced their BTC treasuries by over 15,000 BTC from peak levels during the period, according to CoinShares.
The defining strategic shift of the 2026 mining cycle is the mass conversion of mining capacity to AI and high-performance computing (HPC) infrastructure. Over $70 billion in aggregate GPU co-location and cloud service contracts have been signed between listed miners and hyperscale cloud providers.
The largest deals:
| Company | Counterparty | Contract Value | Duration | Status | |---------|-------------|---------------|----------|--------| | Core Scientific | CoreWeave | $10.2 billion | 12 years | Active; AI = 39% of revenue | | IREN (fmr. Iris Energy) | Microsoft | $9.7 billion | 5 years | GB300 GPU deployment at Childress, TX | | TeraWulf | Multiple | $12.8 billion+ | Long-term | HPC > BTC mining revenue in Q1 2026 |
CoinShares projects that by end-2026, the revenue mix for listed miners will invert: IREN's HPC revenue is expected to reach 71% of total (up from 3% in 2024), Core Scientific is projected at 71% (up from 5%), and TeraWulf at 70%. Across the listed sector, 70% of revenues are forecast to come from AI hosting.
According to S&P Global, the pivot is rational: Bitcoin mining margins are negative or thin at current hashprice levels, while AI hosting contracts offer fixed, long-term revenue with higher margins. The miners' core asset — permitted, grid-connected power capacity with existing cooling infrastructure — is precisely what AI compute buyers need. Mining becomes the "buyer of last resort" for excess capacity, absorbing power during off-peak periods while AI workloads consume the base load.
The mining industry's energy profile has shifted materially. According to research compiled by Spark, 52.4% of Bitcoin mining electricity now comes from zero-emission sources, up from 37.6% in 2022. The breakdown: hydropower 23.4%, wind 15.4%, nuclear 9.8%, and solar 3.2%.
Stranded gas operations — where mobile mining units convert otherwise-flared methane at oil extraction sites into electricity — continue to expand. Companies including Crusoe Energy and Vespene Energy have deployed containerized mining units that reduce methane emissions by up to 90% compared to venting or flaring, per Crusoe's published data.
Marathon has deployed smaller, localized ~10MW containerized sites at the edge of energy networks, targeting the lowest-cost and often intermittent energy sources. This distributed model operates independently of the AI pivot and serves a distinct economic function: absorbing stranded or curtailed energy that has no other buyer.
The dual model emerging is: centralized, grid-connected facilities serve AI workloads as the primary revenue source, while distributed, edge-of-grid mining operations monetize stranded energy with Bitcoin as the flexible load.
With the network consuming an estimated 128 TWh/year at current hashrate levels (approximately 350-420 GWh/day), Bitcoin mining accounts for less than 0.5% of global electricity consumption.
Transaction fees now represent 12-15% of total miner revenue, up from under 7% before the 2024 halving. This ratio matters because Bitcoin's long-term security depends on fees replacing the declining block subsidy.
The current block reward of 3.125 BTC will halve again to 1.5625 BTC around April 2028. Each successive halving makes fee revenue more critical. Industry consensus, per Bitdeer's research arm, holds that if fees consistently account for over 20% of miner revenue, the network can sustain economic security incentives even without meaningful block subsidies.
At current levels, the fee share remains below that threshold. Periodic spikes — driven by Ordinals inscriptions, BRC-20 activity, and Runes launches — have temporarily pushed fees above the subsidy. Whether this becomes a sustained trend depends on Layer 2 adoption, which could either increase base-layer settlement demand or divert transaction volume.
The Bitcoin mining industry in mid-2026 is undergoing a forced evolution. The combination of the 2024 halving, depressed BTC prices, and rising operational costs has compressed margins below zero for a significant minority of operators. The response — mass liquidation of BTC treasuries and a $70 billion pivot to AI infrastructure — is transforming publicly listed miners from single-product Bitcoin producers into diversified energy infrastructure companies.
This transformation carries implications for Bitcoin's network security. As listed miners derive the majority of their revenue from AI hosting, their incentive to allocate marginal capacity to Bitcoin mining weakens. The network's hash security increasingly depends on distributed, low-cost operators mining at the edge of energy networks — a fundamentally different security model than the institutional mining that dominated 2021-2025.
The halving mechanism is functioning as intended: it forces marginal producers out and rewards those with the lowest cost structures. The unintended consequence is that the "lowest cost structure" now means using Bitcoin mining as a secondary load behind AI compute, rather than as a primary business. Whether this dual-use model produces sufficient hashrate to secure a network handling trillions in value remains the central unanswered question of the post-halving era.