Ethereum's next hard fork, Glamsterdam, has entered Devnet-5 testing as of April 2026 with a target deployment in H1 2026, though the Ethereum Foundation acknowledged on April 10 that "Glamsterdam in Q2 seems unlikely" due to implementation complexity in its headliner feature: enshrined Proposer-...
"ePBS only prevents builder centralization from spilling over into the staking layer, but the issue of builder centralization itself remains." — Vitalik Buterin, Ethereum Co-Founder
Ethereum's next hard fork, Glamsterdam, has entered Devnet-5 testing as of April 2026 with a target deployment in H1 2026, though the Ethereum Foundation acknowledged on April 10 that "Glamsterdam in Q2 seems unlikely" due to implementation complexity in its headliner feature: enshrined Proposer-Builder Separation (ePBS). The upgrade, anchored by EIP-7732 (ePBS) and EIP-7928 (Block-Level Access Lists), would raise the gas limit from 60 million to 200 million per block and enable parallel transaction execution — targeting approximately 10,000 transactions per second, roughly 10x current throughput of 15–20 TPS.
The economic stakes are significant. Today, two block builders — Beaverbuild and Titan Builder — produce approximately 86% of Ethereum mainnet blocks, a concentration level that yields a Herfindahl-Hirschman Index (HHI) of 3,892, well above the 1,800 threshold that the U.S. Department of Justice considers "highly concentrated." Glamsterdam's ePBS would move block building from off-chain relay infrastructure into the protocol itself, eliminating Ethereum's 80–90% dependency on third-party relay services like MEV-Boost. Whether this restructuring reduces or amplifies builder concentration remains an open question: a January 2026 academic paper found that ePBS "significantly amplifies profit and content centralisation" among builders.
The upgrade carries a projected 78.6% reduction in gas fees for both simple transfers and complex smart contract calls, according to analysis of the gas repricing bundle (EIP-8007, EIP-7904, EIP-8037, EIP-8038). If delivered, Glamsterdam would represent the most aggressive structural overhaul of Ethereum's execution and consensus layers since The Merge in September 2022.
Glamsterdam bundles eight Ethereum Improvement Proposals outlined by Vitalik Buterin in late February 2026. Two headline EIPs anchor the release:
| EIP | Name | Layer | Function | |-----|------|-------|----------| | EIP-7732 | Enshrined PBS (ePBS) | Consensus | Moves block building into protocol rules | | EIP-7928 | Block-Level Access Lists | Execution | Enables parallel transaction processing | | EIP-7904 | General Repricing | Execution | Recalibrates gas costs for EVM opcodes | | EIP-8007 | Gas Repricing Bundle | Execution | Coordinates gas cost adjustments | | EIP-8037 | State Creation Gas Increase | Execution | Raises gas for new accounts/storage slots | | EIP-8038 | State-Access Gas Increase | Execution | Increases cold account/storage read costs | | EIP-7954 | Max Contract Size Increase | Execution | Raises contract size limits |
The scope narrowed through a structured CFI/DFI (Considered for Inclusion / Do Not Consider for Inclusion) process on All Core Developer calls. Memory repricing proposals EIP-7686 and EIP-7923 were given DFI status for lack of developer support. EIP-2780, which restructures intrinsic gas costs, advanced to CFI.
Target parameters for the upgrade:
| Parameter | Current | Glamsterdam Target | |-----------|---------|-------------------| | Gas limit per block | 60 million | 200 million (phased) | | L1 throughput | 15–20 TPS | ~10,000 TPS | | Gas fees | Baseline | ~78.6% reduction | | Block time | ~12 seconds | ~12 seconds (unchanged) | | Relay dependency | 80–90% | 0% (ePBS replaces relays) |
According to former Ethereum Foundation co-executive director Tomasz Stańczak, gas limit increases would be "phased, reaching 100 million per block initially and 200 million once ePBS is fully operational."
EIP-7732 integrates the proposer-builder separation mechanism directly into Ethereum's consensus layer. Under the current system, block production relies on an off-chain supply chain: builders assemble blocks, submit them through relay services (primarily MEV-Boost operated by Flashbots), and proposers (validators) select the most profitable bid.
Under ePBS, the process changes:
This removes the relay as a trust intermediary. Validators no longer need to trust Flashbots, Ultrasound, or other relay operators to faithfully transmit blocks. The protocol itself enforces the commitment.
The mechanism introduces a known tradeoff called the "free option problem." Builders who commit bids can observe market movements during the commitment window and choose not to reveal their block if conditions change. According to data cited in analysis of the upgrade, this affects approximately 0.82% of blocks on average and up to 6% during volatile periods. The Payload Timeliness Committee is designed to mitigate this by penalizing non-revealing builders.
Approximately 30% of Ethereum blocks currently comply with OFAC sanctions lists, according to relay data — a censorship concern that ePBS addresses by decoupling block content selection from the relay layer's compliance decisions.
EIP-7928 introduces Block-Level Access Lists (BALs), a structural change to how Ethereum processes transactions. Currently, transactions execute sequentially because the EVM cannot predict which storage slots each transaction will access. Two transactions touching the same contract state must be processed in order.
BALs require each block to carry a header-embedded map — the "BAL root" — declaring all accounts and storage slots that each transaction will touch before execution begins. This enables:
Analysis projects 10–30x improvements in execution throughput from BALs alone, though this depends on the degree of state independence between transactions in a given block.
The implementation requires updates to developer tooling. Wallets, dApps, and contract development frameworks must generate access lists for transactions. Contracts with hardcoded gas estimates may break under the new gas repricing schedule, necessitating redeployment or proxy upgrades.
Glamsterdam bundles four gas-related EIPs that recalibrate Ethereum's fee structure based on modern hardware benchmarks:
EIP-7904 (General Repricing) adjusts gas costs for EVM opcodes that currently fall below the 60 Mgas/s floor — operations that are underpriced relative to their actual computational cost. The scope was "sharply reduced" during CFI review to focus only on the most underpriced operations.
EIP-8037 (State Creation Gas Increase) raises the gas cost of creating new accounts and storage slots, making state growth more expensive. This addresses Ethereum's state bloat problem, where the state trie has grown to over 200 GB.
EIP-8038 (State-Access Gas Increase) raises the cost of cold account and storage reads, aligning prices with actual disk I/O costs on commodity hardware.
EIP-2780 (Intrinsic Gas Restructuring) lowers baseline transaction fees while shifting state creation costs into execution, approved to CFI status.
The net effect: heavy state-creating transactions become more expensive; simple transfers and read-heavy operations become cheaper. The 78.6% fee reduction figure applies to the aggregate, weighted by typical transaction mix.
The block builder market that ePBS aims to restructure is highly concentrated. Data from early 2026 shows:
| Builder | Market Share | Profit Margin | |---------|-------------|---------------| | Beaverbuild | ~46% | ~9% | | Titan Builder | ~40% | ~17.75% | | Other builders | ~14% | Varies |
Titan's rapid rise — from less than 1% market share to over 40% — was driven by exclusive order flow agreements that give it privileged access to transaction bundles. This exclusive flow enables Titan to assemble blocks with higher MEV profits, allowing it to consistently win auctions.
Flashbots launched BuilderNet in response, jointly operated with Beaverbuild and Nethermind, aiming to return more MEV and gas fees to users.
The critical question for ePBS is whether moving block building on-chain changes this dynamic. A January 2026 academic paper found that ePBS "significantly amplifies profit and content centralisation" among builders. The mechanism gives efficient builders a structural advantage: by removing relay intermediation, the most profitable builders face fewer friction points to market dominance.
Vitalik Buterin acknowledged the limitation directly: ePBS "only prevents builder centralization from spilling over into the staking layer." Builder centralization itself — the fact that two entities build 86% of blocks — is not addressed by ePBS alone.
As of the Ethereum Foundation's Checkpoint #9 update on April 10, 2026:
The path to mainnet follows a sequential pipeline: stable devnet with all features → client releases → security reviews → testnet deployment (Holesky, Sepolia) → mainnet fork announcement.
The Foundation's own assessment: "Glamsterdam in Q2 seems to me to be unlikely." A Q3 2026 deployment appears more realistic given current development velocity. Ethereum's historical pattern supports this — Pectra, initially targeted for Q1 2025, shipped in May 2025.
The successor upgrade, Hegota (formerly referenced as Heze-Bogota), is planned for H2 2026. Its consensus-layer headliner is FOCIL (EIP-7805) — Fork-Choice Enforced Inclusion Lists — which directly targets the censorship resistance gap that ePBS leaves open.
Under FOCIL, 16 randomly selected validators form an inclusion list committee each slot. Each committee member broadcasts a list of transactions that must be included. Proposers who omit listed transactions face fork-choice penalties — the chain forks away from non-compliant blocks.
This mechanism raises the cost of transaction censorship from bribing a single proposer to neutralizing a rotating group of 16 validators across every slot. Combined with ePBS from Glamsterdam, the two upgrades form a paired defense: ePBS decouples building from proposing; FOCIL ensures that no builder can systematically exclude transactions.
Hegota also targets Verkle Trees for state management and initial post-quantum cryptography work.
Glamsterdam is the most structurally ambitious Ethereum upgrade since The Merge. It attempts to solve three problems simultaneously: builder centralization, sequential execution bottlenecks, and misaligned gas pricing. The scope is large, and the implementation challenges are acknowledged by the development team itself.
The economic value redistribution is the most consequential aspect. Moving block building on-chain removes relay operators as rent-extracting intermediaries but does not, by itself, reduce builder concentration. The two entities that dominate block building today will still compete in an ePBS-governed market — potentially with even greater structural advantages.
For Layer 2 networks that depend on Ethereum for settlement, a 200M gas limit and parallel execution substantially increase available blockspace and reduce settlement costs. For application developers, the gas repricing bundle creates both opportunity (lower fees for reads) and risk (broken gas assumptions). For validators, ePBS simplifies the trust model by eliminating relay dependencies, at the cost of absorbing the free option problem into protocol mechanics.
The upgrade's impact depends on whether it ships. Devnet-5 is running. The generalized devnet is imminent. But the distance between "devnet stable" and "mainnet live" has historically measured in months, not weeks. Q3 2026 appears the most likely deployment window.