Seven of the largest Bitcoin mining pools — Foundry USA, AntPool, F2Pool, SpiderPool, MARA Pool, Block Inc., and DMND — joined the Stratum V2 Working Group on May 7, 2026. Together they represent approximately 75% of Bitcoin's ~901 EH/s network hashrate. The move marks the most significant struct...
Seven of the largest Bitcoin mining pools — Foundry USA, AntPool, F2Pool, SpiderPool, MARA Pool, Block Inc., and DMND — joined the Stratum V2 Working Group on May 7, 2026. Together they represent approximately 75% of Bitcoin's ~901 EH/s network hashrate. The move marks the most significant structural change to Bitcoin's block construction process since the original Stratum protocol was introduced in 2012.
Stratum V2 includes a Job Declaration sub-protocol that permits individual miners, rather than pool operators, to select which transactions are included in new blocks. On June 25, 2026, DMND and GoMining mined block 955,318 — the first known production block where a miner constructed its own template via Job Declaration. The protocol also reduces bandwidth usage by 60-70% and cuts block-switching latency from 325 milliseconds to 1.42 milliseconds. However, as of mid-July 2026, only two pools — Braiins Pool and DMND — run Stratum V2 natively in production. The gap between pool-level endorsement and actual miner deployment remains the critical bottleneck.
Bitcoin's mining pool landscape in mid-2026 is concentrated. According to hashrate data from July 2026, Foundry USA holds 27.6%, F2Pool 17.8%, AntPool 17.3%, ViaBTC 9.5%, and SpiderPool 5.7%. Four pools control roughly 72% of all blocks. Two pools — Foundry and AntPool — together command approximately 45% of network hashrate.
This matters because under the original Stratum V1 protocol, pool operators — not individual miners — decide which transactions go into each block. Miners contribute hash power; the pool constructs the block template. The result: a small number of operators exert disproportionate control over transaction inclusion and ordering.
The risks are not theoretical. If a government compels a pool operator domiciled in its jurisdiction to exclude specific transactions, those transactions face delayed or uncertain confirmation. Foundry USA's miners are heavily concentrated in the United States, meaning geographic regulatory action against a single entity could directly affect more than a quarter of Bitcoin's block production.
According to analysis from D-Central, a single pool controlling 30% of hashrate is "dangerously close to the threshold where collusion or compromise could threaten network integrity." The current arrangement also creates MEV extraction opportunities: pool operators with template control can front-run, reorder, or selectively include transactions for profit.
Stratum V2, originally developed by Braiins (formerly Slush Pool), replaces the JSON-based V1 protocol with a binary-encoded, encrypted communication layer. The protocol includes three sub-protocols:
Job Declaration is the component with the most significant implications for decentralization. Under V1, the pool picks every transaction in every block. Under V2 with Job Declaration enabled, the miner proposes a template, the pool validates it, and the miner's version goes to chain.
The protocol also adds mandatory encryption between miners and pools, eliminating the hashrate hijacking vulnerability present in V1 where attackers on untrusted networks could silently redirect mining output.
The Stratum V2 Reference Implementation (SRI) project has been in development for several years, but pool-level adoption remained limited to Braiins Pool until late 2025.
Key milestones:
According to CoinDesk's May 11, 2026 report, the Working Group formation represents "the biggest decentralization shift in mining in years." However, joining a working group is distinct from deploying the protocol in production.
On June 25, 2026, GoMining used DMND's pool infrastructure to mine block 955,318 with a miner-constructed template. According to Bitcoin Magazine's reporting, GoMining selected its own transactions, built its own block template, submitted it to DMND for validation, and the miner's version went on-chain. No pool operator touched the transaction selection.
GoMining used the mechanism to include transactions from GoBTC Pay, an open-source, non-custodial Bitcoin instant payments protocol the company announced at Consensus Miami in May 2026.
The block demonstrated that Job Declaration works in production — a miner can exercise sovereign control over block contents while still participating in pooled mining for consistent revenue. This is the arrangement Stratum V2 was designed to enable: decentralized template construction with pooled hashrate economics.
One block does not constitute adoption. But it does establish a working reference implementation that other miners and pools can replicate.
Beyond decentralization, Stratum V2 offers measurable operational improvements that create economic incentives for adoption:
Latency: Block-switching latency drops from 325 milliseconds under V1 to 1.42 milliseconds under V2 — a 229x improvement. According to Braiins, this saves approximately 4.9 hours of wasted hash power per year for the average mining operation. In a post-halving environment where margins are compressed, eliminating stale shares directly impacts revenue.
Bandwidth: V2's binary format cuts bandwidth consumption by approximately 60% for pools and 70% for miners compared to V1's JSON messages. For miners in bandwidth-constrained environments — common in developing markets and remote industrial sites — this reduction improves connection reliability.
Profitability: Braiins estimates the combined efficiency gains — reduced stale shares, elimination of hashrate hijacking, and optimized transaction selection — can increase net mining profit by up to 7.4%. The breakdown: 1-2% from encryption (eliminating silent hijacking), several percent from transaction selection (running a local mempool to capture highest-fee transactions), and 1-2% from fewer stale shares. The 7.4% figure represents a ceiling, not a guarantee — actual gains depend on miner configuration, network conditions, and whether Job Declaration is actively used.
The gap between the 75% hashrate Working Group endorsement and actual production deployment is the central tension in the Stratum V2 narrative.
Production pools (as of July 2026):
Testing/planning phase:
Combined production V2 hashrate is estimated at under 5% of the network. The remaining ~70% represented by Working Group members is in various stages of integration.
The Stratum V2 project authors have acknowledged this gap directly. According to the SRI documentation, "the gap between pool-level endorsement and actual individual miner deployment of Job Declaration is the specific engineering bottleneck that determines whether Stratum V2's decentralization promise is realized or merely theoretical."
Pool operators face a coordination problem. Deploying V2 requires updating pool infrastructure, testing against multiple firmware versions, and ensuring backward compatibility with V1 miners who may not upgrade. Miners face a parallel problem: many ASIC devices ship with V1-only firmware, and upgrading requires flashing third-party firmware such as Braiins OS+.
SBI Crypto, the mining subsidiary of Japanese financial conglomerate SBI Group, will permanently close its Bitcoin mining pool on July 31, 2026. The pool's 7-day average hashrate as of June 30 stood at 20.9 EH/s — approximately 2.2% of the network.
SBI did not disclose a specific reason for the closure. According to Blockspace, a security breach in September 2025, followed by operational adjustments and an internal business review in April 2026, preceded the shutdown — though SBI has not confirmed a direct connection.
SBI named Luxor, Braiins, and NeoPool as reference options for migrating miners. The redistribution of 20.9 EH/s across remaining pools will modestly alter hashrate concentration ratios. If a significant portion migrates to V2-native Braiins Pool, it could accelerate production-level V2 adoption. If it flows to Foundry or AntPool, it marginally increases concentration.
The closure also highlights post-halving economics. With block rewards halved to 3.125 BTC in April 2024 and Bitcoin trading at approximately $63,900 in mid-July 2026, smaller pool operations face compressed margins. SBI's exit suggests that operating a mining pool as a secondary business line is increasingly uneconomic without scale.
ASIC firmware support is the supply-side constraint on V2 adoption.
Current status:
According to projections from the Stratum V2 Working Group, Q3 2026 is the target for V2 to become the default protocol in new ASIC firmware shipments, with both Bitmain and MicroBT expected to include V2 support in stock firmware for new models. End-of-2026 projections estimate 40-60% of network hashrate running on Stratum V2.
These projections should be treated with caution. MicroBT's current lack of V2 support means a substantial portion of deployed ASICs — Whatsminer is the second-largest ASIC manufacturer by market share — cannot run V2 without third-party firmware or hardware replacement.
Stratum V2 represents the first credible attempt to separate block template construction from hashrate pooling in Bitcoin mining. The protocol's Working Group now includes operators responsible for three-quarters of Bitcoin's hash power, and block 955,318 proved the mechanism works in production.
The economic case — bandwidth savings, latency reduction, stale share elimination — provides adoption incentives that do not depend on ideological commitment to decentralization. In a post-halving margin environment where every percentage point of efficiency matters, these gains are non-trivial.
The critical variable is deployment velocity. If V2 remains confined to Braiins Pool and DMND through 2026, the Working Group endorsement amounts to a statement of intent rather than a structural change. If Foundry, AntPool, and F2Pool ship production V2 implementations and Bitmain and MicroBT bundle V2 in stock firmware, the 40-60% hashrate target by year-end becomes plausible.
Until then, 75% is an aspiration. Under 5% is the reality.