Bitcoin's network hash rate has fallen below 1 zetahash per second (ZH/s) for the first time since September 2025, settling at approximately 961 EH/s as of April 4, 2026. The decline — down 22% from the late-September 2025 peak of 1.28 ZH/s — marks the first quarterly hash rate contraction in six...
"Bitcoin miners are losing $19,000 on every BTC produced." — CoinDesk Markets Desk, reporting on CoinShares Q1 2026 data
Bitcoin's network hash rate has fallen below 1 zetahash per second (ZH/s) for the first time since September 2025, settling at approximately 961 EH/s as of April 4, 2026. The decline — down 22% from the late-September 2025 peak of 1.28 ZH/s — marks the first quarterly hash rate contraction in six years. Q1 2026 ended with hash rate down approximately 4% year-to-date, breaking a five-year streak of consecutive double-digit annual growth.
The April 3 difficulty adjustment brought a 3.87% increase to 138.97 T, but early block-interval data from the current epoch projects a 14.27% downward adjustment on April 19 — which would be the largest single negative adjustment since February 2026's 11.16% drop. Average block times are running at 11 minutes 39 seconds, well above the protocol's 10-minute target. With average production costs near $88,000 per BTC and the spot price at approximately $67,800, the median miner is operating at a 21% per-coin loss. CoinShares estimates 15–20% of legacy mining rigs are now cash-flow negative.
The Bitcoin network's total computational power peaked at approximately 1.28 ZH/s in late September 2025, according to data from CoinWarz and CloverPool. It has since contracted by roughly 320 EH/s — a 25% drawdown that industry analysts at CoinShares describe as "structural capitulation" rather than a transient weather event or regional disruption.
Key data points from the decline:
For context, over the prior five years, the network's hash rate rose from roughly 100 EH/s to over 1 ZH/s — a tenfold increase. Each preceding first quarter had ended with positive growth. Q1 2026 broke that pattern.
Bitcoin's difficulty retargets every 2,016 blocks (approximately two weeks) to maintain the 10-minute average block interval. When hash rate declines, blocks slow down, and the algorithm responds by reducing difficulty at the next epoch boundary.
The April 3 adjustment at block height 943,488 brought a modest 3.87% increase to 138.97 T. However, the block-time data accumulated since that adjustment tells a different story. As of April 4, 181 of the current epoch's 2,016 blocks had been mined, with average intervals running at 11 minutes 39 seconds — 16.5% above the 10-minute target.
Current projections from mining data aggregators estimate the April 19 adjustment will deliver a 14.27% reduction, potentially lowering difficulty from 138.97 T to approximately 118.44 T. If realized, this would be the largest single negative adjustment since the February 7, 2026 reset, which saw an 11.16% decline — itself the steepest cut since July 2021.
The February adjustment was triggered by a 20% hash rate collapse as miners with negative cash flow shut down equipment en masse.
The core problem is arithmetic. According to CoinDesk's reporting on CoinShares' Q1 2026 mining data, the average all-in production cost of one Bitcoin in early 2026 sits near $88,000. Bitcoin's spot price as of Q1 close was approximately $67,800 — a 24% shortfall. At an electricity-only breakeven of roughly $74,000, the majority of the network's mining fleet is underwater.
Hash price — daily revenue per petahash of deployed capacity — collapsed to approximately $28–30/PH/s/day by early March 2026, according to CoinShares. This is a five-year low and represents a critical threshold:
CoinShares estimates 15–20% of the global mining fleet is now operating at negative cash flow. Electricity accounts for 75–85% of monthly operating expenses. At $0.05/kWh, miners can sustain operations through the downturn; at $0.08/kWh or higher with mid-generation ASICs, the math collapses.
The breakeven price varies significantly by operator. Efficient large-scale miners with access to stranded energy can produce BTC for as low as $38,000. Retail miners or those paying residential electricity rates face production costs exceeding $92,000, per data compiled by CompareForexBrokers.
The Hash Ribbon indicator — which tracks the crossover between the 30-day and 60-day moving averages of hash rate — has been flashing a capitulation signal. According to MEXC's research desk, this is the first sustained Hash Ribbon capitulation since the FTX-era signal in late 2022.
Historically, Hash Ribbon capitulation signals have preceded significant price recoveries:
The indicator's premise is straightforward: miner capitulation forces the weakest operators offline, reducing sell pressure from mining treasuries and resetting the cost structure for survivors. Whether 2026 follows historical precedent remains uncertain — the structural shift of miners toward AI infrastructure introduces a variable absent from prior cycles.
A declining hash rate reduces the cost of mounting a 51% attack, at least in theory. With hash power at 961 EH/s versus the September 2025 peak of 1.28 ZH/s, the economic barrier to an attack has decreased proportionally.
However, context matters. At 961 EH/s, Bitcoin's network remains orders of magnitude more secure than any other proof-of-work chain. The practical cost of sustaining a 51% attack for even one hour — accounting for hardware acquisition, electricity, and the self-defeating price impact — remains prohibitive. According to KuCoin's hashrate analysis, Bitcoin's difficulty adjustment algorithm is engineered for precisely this scenario: when miners exit and hash rate declines, difficulty adjusts downward, improving economics for remaining participants and drawing capacity back.
The concern is directional, not immediate. If the mining industry's structural pivot to AI continues to divert capital away from hash rate growth over multiple quarters, the long-term security premium that Bitcoin derives from massive proof-of-work expenditure could erode. CoinShares' base-case forecast still projects hash rate recovering to 1.8 ZH/s by year-end 2026, but conditions that estimate to a BTC price recovery toward $100,000.
Publicly listed miners have accelerated BTC treasury liquidation. Per CoinShares:
Companies including Core Scientific (CORZ), Riot Platforms, WULF, CIFR, and HUT have effectively become data center operators with residual Bitcoin mining operations. The capital reallocation is rational at current hash price levels: AI compute contracts offer higher and more predictable returns than mining BTC at a $19,000-per-coin loss.
JPMorgan noted in a January 2026 report that a declining network hash rate lifted revenues per exahash and margins for remaining miners, with gross mining margins improving by roughly 300 basis points from December to about 47%. Revenue per exahash, however, remains well below year-ago levels.
If the projected 14.27% difficulty cut on April 19 materializes, it will represent the algorithm working as intended — reducing the barrier to profitability for surviving miners. The mechanical effect:
This self-correcting mechanism has functioned reliably across Bitcoin's 16-year history. The question for 2026 is whether the correction can offset the structural outflow of mining capital to AI infrastructure, or whether the network settles into a lower-hash-rate equilibrium than the market has priced in.
CoinShares' forecast of 1.8 ZH/s by December 2026 requires BTC to recover toward $100,000. If price remains in the $65,000–$75,000 range, the hash rate may stabilize closer to 1.0–1.1 ZH/s — sufficient for network security but well below the growth trajectory the market assumed 12 months ago.
Bitcoin's difficulty adjustment algorithm — the protocol's most underappreciated feature — is performing its designed function: repricing the cost of participation as economic conditions shift. The projected 14.27% cut on April 19 would be the latest in a series of negative adjustments that have characterized 2026, with seven of the last ten adjustments trending downward.
The immediate network security concern is minimal. At 961 EH/s, Bitcoin remains the most computationally secured ledger in existence. The medium-term concern is structural: the mining industry's pivot to AI is not a temporary reallocation but a rational response to hash price levels that make BTC production uneconomic for a significant share of the fleet. Whether this pivot reverses depends almost entirely on the BTC spot price.
For the protocol itself, the adjustment mechanism is doing what Satoshi designed it to do. The question is whether the economic incentives that have historically drawn hash rate back after capitulation events will function the same way when the alternative use of that capital — AI infrastructure — offers returns that Bitcoin mining at sub-$70,000 prices cannot match.