Ethereum's Glamsterdam hard fork, targeted for May–June 2026, represents the network's most structurally significant upgrade since The Merge. The fork centers on two headliner Ethereum Improvement Proposals: EIP-7732 (Enshrined Proposer-Builder Separation) and EIP-7928 (Block-Level Access Lists)....
"2026 is the year that we take back lost ground in terms of self-sovereignty and trustlessness." — Vitalik Buterin, Ethereum Co-Founder
Ethereum's Glamsterdam hard fork, targeted for May–June 2026, represents the network's most structurally significant upgrade since The Merge. The fork centers on two headliner Ethereum Improvement Proposals: EIP-7732 (Enshrined Proposer-Builder Separation) and EIP-7928 (Block-Level Access Lists). Together, they aim to move block building into the protocol, enable parallel transaction execution, and reduce gas fees by an estimated 78.6% across both simple transfers and complex smart contract calls.
Devnet testing is already underway. The bals-devnet-2 testnet launched in February; epbs-devnet-0 went live in late February. Public testnet deployments and dual audit phases are scheduled for spring 2026. Scope freeze occurred in late Q1, locking the technical ingredients. If testing proceeds without major setbacks, mainnet activation could arrive by mid-year — though core developers have stressed the timeline remains aspirational pending validation of critical components.
The upgrade arrives as Ethereum processes over 200 million transactions per quarter, maintains 1.1 million active validators staking 35.9 million ETH (28.9% of total supply), and supports an L2 ecosystem handling 8.2 million daily transactions at fees that fell 40–90% following the December 2025 Fusaka upgrade.
The Fusaka upgrade activated on December 3, 2025, followed by BPO-2 (Blob Parameter Only) on January 7, 2026. Fusaka introduced PeerDAS (Peer Data Availability Sampling), scaling blob capacity by approximately 8x while reducing the data load on home validators by 87.5%. According to Fidelity Digital Assets, the upgrade generated an estimated 8x increase in burnt blob fees in the first three weeks of January 2026 compared to the same period in 2025.
Post-Fusaka, L2 transaction costs fell sharply. Average fees on Base dropped to approximately $0.000116, according to MEXC research, while Optimism and Arbitrum users reported costs of $0.005–$0.02. Ethereum mainnet gas fees fell to $0.14–$0.15 on average — the lowest since 2017. Routine token swaps cost as little as $0.04, per Blocklist reporting.
On the Ethereum Foundation's February 2026 blog post outlining protocol priorities, three development tracks were defined for the year: Scale (pushing the gas limit toward and beyond 100 million, up from 60 million), Improve UX (native account abstraction and interoperability), and Harden the L1 (post-quantum readiness, censorship resistance, expanded testing infrastructure). Glamsterdam addresses the first and third tracks directly.
The centerpiece of Glamsterdam's consensus-layer changes is EIP-7732, which moves proposer-builder separation (PBS) from an external relay system into the Ethereum protocol itself.
The problem it solves: Today, MEV-Boost relays operate as trusted intermediaries between block proposers and builders. According to Relayscan data from April 2, 2026, a single builder — Titan — produced 52.16% of blocks in one 24-hour period and earned 72.99 ETH in block profits, compared to 9.55 ETH for the second-largest builder, BuilderNet. The top relay, relay.ultrasound.money, handled 33.92% of payloads, followed by Titan's relay at 24.19% and bloXroute at 14.67%. The European Securities and Markets Authority (ESMA) has flagged centralization at the relay and builder levels as presenting censorship and operational risk.
How ePBS works: EIP-7732 separates each Ethereum block into consensus and execution components, introducing an in-protocol commit-reveal flow. Builders commit to execution payloads; the protocol enforces those commitments at the consensus layer. Explicit deadlines, payload commitments, and fallback behavior ensure that builder non-delivery does not halt chain liveness. No external relay infrastructure is required.
What changes: Block building becomes protocol-native. Relay operators — currently an unregulated chokepoint — are removed from the critical path. MEV handoff rules are standardized. Builder commitments become enforceable rather than trust-based.
The execution-layer headliner is EIP-7928, which introduces Block-Level Access Lists (BALs) — a mechanism that records all accounts and storage locations accessed during block execution, along with their post-execution values.
The problem it solves: Transaction execution on Ethereum is sequential. Without knowing in advance which addresses and storage slots a transaction will access, nodes cannot safely process transactions in parallel. While EIP-2930 (Berlin, 2021) introduced optional transaction-level access lists, they are not enforced and see limited adoption.
How BALs work: Each block includes a pre-declared map of state access. A new field, block_access_list_hash, is added to the block header containing the Keccak-256 hash of the RLP-encoded access list. According to research published on Ethereum Research, historical data shows approximately 70 KiB average BAL size. Critically, 60–80% of Ethereum transactions access disjoint storage slots, meaning they can be parallelized without conflict. The remaining 20–40% can still benefit from parallelization through post-transaction state diffs.
What changes: Nodes can execute non-conflicting transactions simultaneously. This enables parallel disk reads, parallel transaction validation, and parallel state root computation. Block processing time decreases; throughput capacity increases — a prerequisite for safely raising the gas limit.
Supporting the two headliners is EIP-7904, which recalibrates gas costs for EVM opcodes based on client benchmarks rather than historical estimates.
Many of Ethereum's current gas prices were set years ago and no longer reflect execution costs on modern hardware. EIP-7904 targets 18 opcodes that currently execute below 60 Mgas/s — meaning their gas costs overcharge relative to actual computational resources consumed. According to Etherworld reporting, the repricing focuses on computational complexity measured as execution time on bare CPU, excluding network-related costs like state persistency.
The combined effect of EIP-7928 (parallel execution) and EIP-7904 (accurate gas pricing) is projected to yield a 78.6% reduction in gas fees for both simple ETH transfers and complex smart contract interactions, according to analysis by Datawallet and Phemex.
The repricing is part of a broader package: 10 gas-related EIPs were presented at the initial Glamsterdam preparatory sessions, per Etherworld, forming the most comprehensive gas recalibration since EIP-1559 in August 2021.
| Phase | Milestone | Status | |-------|-----------|--------| | Q4 2025 | Glamsterdam EIP proposals submitted | Complete | | Late Q1 2026 | Scope freeze | Complete | | Feb 4, 2026 | bals-devnet-2 launch | Complete | | Late Feb 2026 | epbs-devnet-0 launch (consensus spec v1.7.0-alpha.2) | Complete | | Spring 2026 | Public testnets + dual audit phases | In progress | | May–Jun 2026 | Mainnet activation (tentative) | Pending |
According to QuickNode's reporting, up to 22 EIPs may be included in the final Glamsterdam scope. The scope freeze in late Q1 locked the headliners. Features not achieving interoperability by that deadline — such as any ePBS components requiring additional testing — could be deferred to the Hegota fork later in 2026.
The Ethereum Foundation's Checkpoint #8 blog post (January 2026) confirmed the accelerated upgrade cadence, with two hard forks planned for the calendar year for the first time since the pre-Merge era.
For validators: The 1.1 million active validators earning an average 3.3% APY face operational changes. ePBS alters the block production workflow: validators will no longer need MEV-Boost software to access builder markets. This lowers the technical barrier for solo stakers but may reduce MEV-related rewards for sophisticated operators who currently capture outsized returns through relay optimization.
For L2 operators: The gas limit trajectory — from 30 million (pre-2025) to 60 million (post-Fusaka) to a target of 100 million and beyond — directly affects L2 data costs. Each gas limit increase expands the data propagation window from approximately 2 seconds to roughly 9 seconds, per CryptoSlate analysis, unblocking Ethereum's ability to safely accommodate more data blobs. L2s already processing 8.2 million daily transactions stand to see further cost reductions.
For application developers: The 78.6% gas fee reduction, if realized, changes the economic calculus for on-chain computation. Complex smart contract operations that are currently prohibitively expensive on L1 become more viable, potentially shifting some activity back from L2s to mainnet — though this remains speculative.
For relay operators: ePBS renders the current relay infrastructure architecturally redundant. Relay operators like Flashbots, bloXroute, and Ultra Sound will need to pivot their business models or find new value-adds beyond the relay function.
Glamsterdam is the first of two planned 2026 hard forks. Hegota, targeted for H2 2026, has begun accepting EIP proposals as of January 8, with a February 4 submission deadline.
Hegota's confirmed headliner is FOCIL (Fork-Choice Enforced Inclusion Lists, EIP-7805), which randomly selects 17 participants per block slot who can force specific transactions into blocks, providing protocol-level censorship resistance. The other major target is Verkle Trees — a data structure replacement that could reduce node storage requirements by approximately 90%, enabling stateless clients.
Together, Glamsterdam and Hegota address the full scope of the Ethereum Foundation's 2026 priorities: parallel execution and gas scaling (Glamsterdam), censorship resistance and state reduction (Hegota).
Glamsterdam is an infrastructure upgrade, not a feature release. It restructures how Ethereum builds blocks, executes transactions, and prices gas. The two headliner EIPs address long-standing structural concerns — relay centralization and sequential execution — that have constrained Ethereum's throughput and decentralization properties since The Merge.
Whether the May–June timeline holds depends on testnet outcomes still in progress. The scope is ambitious: ePBS alone represents years of research being brought to production. But the engineering pipeline — devnets running, scope frozen, audits scheduled — suggests the upgrade is past the conceptual stage and into execution.
The economic question is whether lower fees and higher throughput generate proportional increases in on-chain activity, or whether L2s have already captured the marginal user. Glamsterdam makes the L1 more capable. Whether that capability translates to usage and value accrual depends on what gets built on top of it.