Ethereum's Glamsterdam hard fork, originally targeting H1 2026, has been delayed to Q3 2026 — with an end-of-August activation now the working estimate. The upgrade bundles two headline proposals, EIP-7732 (enshrined Proposer-Builder Separation) and EIP-7928 (Block-Level Access Lists), alongside ...
"2026 is the year that we take back lost ground in terms of self-sovereignty and trustlessness." — Vitalik Buterin, Co-founder, Ethereum
Ethereum's Glamsterdam hard fork, originally targeting H1 2026, has been delayed to Q3 2026 — with an end-of-August activation now the working estimate. The upgrade bundles two headline proposals, EIP-7732 (enshrined Proposer-Builder Separation) and EIP-7928 (Block-Level Access Lists), alongside a package of gas repricing EIPs that collectively target a 200-million gas limit floor — a 3.3x increase from the current ~60 million baseline. If delivered, the upgrade would push Layer 1 throughput toward 10,000 transactions per second, reduce gas fees by an estimated 78%, and eliminate Ethereum's dependence on off-chain relay infrastructure that currently routes approximately 88% of all blocks.
The delay, confirmed in the Ethereum Foundation's May 2026 protocol update following a week-long interop event in Svalbard, Norway, reflects the technical difficulty of enshrining block-building logic into the consensus layer. Devnets are live and multi-client testing is underway, but ePBS implementation has proven more complex than anticipated. Simultaneously, the Foundation is undergoing a leadership transition, with veteran coordinators Tim Beiko, Barnabé Monnot, and Alex Stokes departing or taking sabbaticals, replaced by Will Corcoran, Kev Wedderburn, and Fredrik.
Ethereum's block production pipeline has a centralization problem. Approximately 88% of Ethereum blocks are currently built off-chain via MEV-Boost, according to data from Everstake. The top three builders control over 80% of those blocks. The largest builder, beaverbuild, holds close to 50% market share on its own. Both beaverbuild and rsync-builder censor OFAC-sanctioned transactions, meaning roughly 60% of Ethereum blocks are subject to some form of transaction censorship.
This concentration exists because of Maximal Extractable Value (MEV) — the profit that specialized actors extract by reordering, inserting, or censoring transactions within a block. Over 50% of high-value Ethereum transactions were routed through private channels by mid-2025 to avoid MEV extraction, according to Bitfinex Research. The current system relies on trusted relays — off-protocol intermediaries that sit between block builders and validators. These relays operate without on-chain accountability, creating a trust assumption that exists entirely outside Ethereum's protocol rules.
Enshrined Proposer-Builder Separation aims to eliminate this dependency.
EIP-7732 moves the builder market from off-chain relay infrastructure into Ethereum's consensus layer. Under the current system, validators trust relays not to manipulate or reveal block contents before finalization. Under ePBS, builders cryptographically seal blocks and commit to bids without revealing contents. Validators select the highest bid. A new mechanism, the Payload Timeliness Committee (PTC), introduces a validator duty to attest that block payloads arrive on time.
The practical effect: relay operators like Flashbots, who currently serve as critical but unregulated intermediaries, would no longer be required for block production. Transaction ordering moves on-chain, subject to protocol rules rather than off-chain trust assumptions.
However, ePBS does not eliminate MEV itself. Bitfinex Research notes that enshrinement "significantly amplifies profit and content centralisation" among builders. The "free option problem" persists: builders can withhold block payloads after committing if late-arriving MEV makes abandonment profitable. This occurs in approximately 0.82% of blocks under normal conditions and up to 6% during periods of high volatility. Addressing this will require follow-on proposals — notably FOCIL (Fork-Choice enforced Inclusion Lists), which has been moved to the subsequent Hegotá upgrade.
Additionally, ePBS extends the data propagation window from 2 seconds to approximately 9 seconds, giving execution clients more time to process blocks — a prerequisite for handling higher gas limits.
EIP-7928 introduces a mechanism for declaring which accounts and storage slots a block will access before execution begins. This pre-declaration enables three capabilities:
BALs are the technical foundation that makes a 200-million gas limit credible. Without pre-declared state access, processing 3.3x more gas per block would overwhelm node operators' I/O capacity. According to QuickNode's technical analysis, "BALs open three downstream capabilities: parallel execution, stateless client reconstruction, and a credible path to 200M gas limit."
Beyond the two headline EIPs, Glamsterdam bundles at least six additional proposals focused on gas economics:
| EIP | Description | |-----|-------------| | EIP-7976 | Raises calldata floor cost | | EIP-7981 | Increases access list costs | | EIP-8037 | Raises state creation gas costs (anti-state-bloat) | | EIP-7954 | Increases maximum contract size | | EIP-7778 | Removes gas refunds from block accounting | | EIP-8024 | New stack opcodes (SWAPN, DUPN, EXCHANGE) |
The net effect of these repricing changes is an estimated 78% reduction in gas fees across simple transfers and complex smart contract interactions, according to Phemex Research. The fee reduction is a function of higher throughput (more gas per block) combined with higher costs for state-expanding operations (discouraging state bloat).
EIP-8037 was finalized at the Svalbard interop with a fixed cost_per_state_byte structure, and full repricing figures were delivered on bal-devnet-6, according to the Ethereum Foundation's May protocol update.
Two proposals were removed from the Glamsterdam scope: EIP-7782 (6-second slot times) and FOCIL (inclusion lists). Both were moved to Hegotá, the subsequent upgrade targeting Q4 2026 or Q1 2027.
The original H1 2026 target, with June as the aspirational date, has slipped. The Ethereum Foundation's May 2026 protocol update confirmed Q3 2026 as the revised window, with end-of-August 2026 as the working target, according to Everstake's analysis.
Current development milestones:
The delay follows a pattern. Ethereum's upgrade cadence accelerated after The Merge (September 2022), with Pectra shipping in May 2025 and Fusaka in December 2025. Glamsterdam, if it lands in August, would represent an 8-month gap — the longest since the pre-Merge era.
The Glamsterdam delay coincides with a significant reshuffling at the Ethereum Foundation's Protocol cluster:
Departing:
Incoming:
The transition represents a shift in the Foundation's protocol coordination from consensus mechanism experts toward security-and-proving specialists — consistent with Ethereum's roadmap moving from consensus-layer changes toward execution-layer scaling and zero-knowledge proof integration.
MEV does not disappear. ePBS moves the builder market on-chain but does not eliminate the economic incentives for MEV extraction. Builder centralization may persist or worsen, as on-chain auctions formalize the advantage held by sophisticated, well-capitalized builders. FOCIL and encrypted mempools — the primary censorship resistance mechanisms — are deferred to Hegotá.
State growth pressure. A 3.3x increase in gas capacity will accelerate Ethereum's state growth. The gas repricing EIPs are designed to counteract this by making state-expanding operations more expensive, but the net effect on state size is an empirical question that mainnet data has not yet answered.
Node operator burden. The 200-million gas target requires node operators to ensure sufficient storage IOPS (input/output operations per second) and network latency. QuickNode flags this as a preparation requirement for operators. Validators must also prepare for the new PTC attestation duty introduced by ePBS.
Complexity risk. Glamsterdam introduces substantial consensus-layer complexity. The ePBS mechanism is unproven at mainnet scale. Bitfinex Research identifies "chain stagnation" as a concern — the risk that edge cases in the new block production pipeline could cause liveness failures.
Price context. ETH was trading around $1,920 at the time of the Glamsterdam delay announcement, according to Phemex — down approximately 45% from its October 2025 high of $3,600. Whether the upgrade can reverse ETH's price decline relative to Bitcoin and competing L1s remains uncertain.
Glamsterdam is Ethereum's most structurally significant upgrade since The Merge. It addresses a genuine problem: the off-chain relay infrastructure that builds 88% of Ethereum's blocks operates without protocol-level accountability, and three builders control the majority of block production. Moving this market on-chain is architecturally sound.
But the upgrade is late, the implementation is complex, and the MEV problem it partially addresses will require at least one more hard fork (Hegotá) to fully mitigate. The 200-million gas target is ambitious — a 3.3x capacity increase that depends on BALs working as designed at scale.
For the Ethereum network, Glamsterdam is a necessary step. For ETH holders, it is a technical upgrade whose price impact is uncertain. The data will speak after deployment.