Ethereum activated its Glamsterdam hard fork on May 4, 2026, raising the block gas limit from 60 million to 200 million and tripling on-chain processing capacity in the network's largest protocol change since the September 2022 Merge. ETH traded at $2,390 on the day of activation, with 24-hour tr...
"In 2026, no longer. Every compromise of values that Ethereum has made up to this point... we are making that compromise no longer." — Vitalik Buterin, Ethereum Co-Founder
Ethereum activated its Glamsterdam hard fork on May 4, 2026, raising the block gas limit from 60 million to 200 million and tripling on-chain processing capacity in the network's largest protocol change since the September 2022 Merge. ETH traded at $2,390 on the day of activation, with 24-hour trading volume surging 145% to $17.71 billion, according to CoinMarketCap data. The upgrade introduces two headline protocol changes — Enshrined Proposer-Builder Separation (ePBS) via EIP-7732 and Block-Level Access Lists (BALs) via EIP-7928 — alongside gas repricing, multidimensional gas metering, and inclusion list enforcement.
The activation arrives as Ethereum navigates a structural tension: network activity has reached record highs across active addresses and smart contract calls, yet Layer 1 fee revenue has collapsed more than 95% from 2021 peaks. Glamsterdam attempts to address both the capacity constraint and the value-accrual gap simultaneously. Whether it succeeds on the latter remains the open question.
Glamsterdam activated at the consensus layer on May 4, 2026. The block gas limit increased from 60 million to an initial 100 million, with a phased ramp to 200 million as ePBS reaches full operational status. The upgrade targets a throughput ceiling of approximately 10,000 transactions per second, roughly 10x the pre-upgrade capacity.
Market response was immediate. ETH pushed above $2,390 intraday on May 4, with daily volume hitting $19 billion according to CoinMarketCap. By May 6, ETH stabilized around $2,410–$2,415 on major exchanges including Binance and OKX. Market capitalization stood at approximately $284.77 billion, representing 10.4% of total crypto market dominance, per Spoted Crypto data.
On-chain data showed 140,000 ETH accumulated by whale addresses in the 96 hours surrounding activation, according to blockchain analyst Ali Martinez. Perpetual futures open interest sat at $5.0 billion with a negative funding rate of -0.0020%, indicating a slight directional tilt toward long positioning. The long/short ratio stood at 1.28 (56% long vs. 44% short).
For context, ETH remains 55% below its August 2025 all-time high of $4,954. The Fear and Greed Index read 50/100 — neutral — on May 5.
Glamsterdam bundles approximately ten EIPs into a single hard fork. The Ethereum Foundation's DevOps team tested three core EIPs through Devnet-4 and progressed to Devnet-5 before the first generalized devnet launched in late April 2026, merging all Glamsterdam components into a single testing environment for the first time.
The headline EIPs:
The net effect: gas fees across both simple transfers and complex smart contract interactions are projected to decline by approximately 78.6%, according to Ethereum Foundation benchmarks.
Prior to Glamsterdam, Ethereum's proposer-builder separation operated almost entirely through off-chain infrastructure — specifically MEV-Boost relays operated by Flashbots and other providers. By early 2025, the top two block builders captured over 90% of block auctions, according to MEV research data. Approximately 80–90% of Ethereum blocks were produced using external relays, creating trust assumptions and centralization vectors at the infrastructure layer.
EIP-7732 replaces this with a protocol-native mechanism. Under ePBS, builders assemble blocks and cryptographically seal their contents. Proposers select the highest-paying block without visibility into transaction ordering. Transactions are revealed only after finalization, reducing opportunities for sandwich attacks and frontrunning.
The data propagation window expands from 2 seconds to approximately 9 seconds under the new architecture, decoupling execution validation from consensus validation both logically and temporally.
On Ethereum's MEV landscape: 2025 MEV transaction volume reached approximately $561.9 million, with sandwich attacks accounting for 51.56% (approximately $289.8 million), according to research compiled by CoinLaw. Monthly sandwich extraction on Ethereum had already declined from approximately $10 million in late 2024 to roughly $2.5 million by October 2025, partly due to private transaction routing. ePBS aims to reduce total MEV extraction by up to 70%, per Phemex analysis, by eliminating the builder's ability to reorder transactions after a proposer commits.
EIP-7928's Block-Level Access Lists represent a structural shift in how Ethereum processes transactions. By requiring each block to pre-declare which accounts and storage slots it will access, the protocol can identify non-overlapping transactions and execute them simultaneously rather than sequentially.
This parallel execution model, combined with the gas limit increase to 200 million, is what produces the 10,000 TPS throughput target. The approach differs from Solana's parallel execution model (which uses predetermined account access at the transaction level) by operating at the block level, with the tradeoff being slightly less granular parallelism but compatibility with Ethereum's existing smart contract architecture.
EIP-7904's gas repricing adjusts costs that in some cases have not been recalibrated since Ethereum's early years. The repricing decouples state creation costs from execution gas, meaning that creating new storage entries becomes significantly more expensive relative to computation — a deliberate design choice to slow state growth.
Glamsterdam does not directly solve Ethereum's most pressing economic challenge: the collapse of Layer 1 fee revenue.
According to CoinDesk reporting from March 2026, Ethereum's network activity surged to all-time highs — daily active addresses approached 2 million in February 2026, exceeding peaks from the 2021 cycle, per CryptoQuant data. Yet fee generation at the base layer has not kept pace. Weekly fee revenue dropped below $500,000 as of early May 2026, down from $1.77 million the prior week, according to Spoted Crypto.
The core dynamic: Ethereum's success in scaling through rollups has suppressed L1 fee burn. After the Pectra upgrade in May 2025, average daily ETH burn fell to approximately 3.26 ETH per day — a 71% decrease from pre-Pectra levels. ETH is currently slightly inflationary at roughly 0.23% annually, with blob utilization sitting at only 20–30% of capacity.
The value leakage is quantifiable. Coinbase's Base chain earned over $94 million in profit but returned only $4.9 million to Ethereum in blob fees, according to Blockworks reporting. Over 65% of new smart contracts now deploy to L2 networks rather than Ethereum mainnet.
Glamsterdam's gas limit increase will further reduce L1 transaction costs, potentially accelerating this dynamic in the short term. The counter-argument: by making L1 competitive enough for more transaction types, Glamsterdam may pull some activity back from L2s. The data is inconclusive on which effect dominates.
With the mainnet's processing capacity tripling, settlement costs for Layer 2 rollup projects are estimated to have fallen by approximately 70%, according to industry estimates cited by multiple sources. This intensifies fee competition among major L2s.
As of late 2025, Ethereum Layer 2 networks collectively held approximately $39.4 billion in TVL. Base and Arbitrum dominate with a combined 77% market share — Base at 46.58% and Arbitrum at 30.86%, per L2BEAT data. Arbitrum led with $16.63 billion TVL, Base held $10 billion, and Optimism secured $6 billion.
DeFi TVL on Ethereum mainnet stood at $45.74 billion as of May 2026, representing 68% of global DeFi TVL, according to Spoted Crypto. That figure recovered from a $44.67 billion low, suggesting capital is not fleeing the ecosystem — it is redistributing within it.
The L2 consolidation trend is pronounced. Over 1.9 million daily transactions now process on L2 networks. DeFi transaction activity surged 38% year-over-year even as TVL growth remained in single digits, indicating that capital efficiency is improving — the same pool of locked value is generating more on-chain economic activity per dollar deployed.
ePBS resolves one centralization vector — reliance on off-chain relay infrastructure — but may introduce another. Vitalik Buterin himself outlined concerns that the Glamsterdam upgrade "could unintentionally concentrate power among a small group of sophisticated block builders," according to Unchained Crypto reporting.
The sophistication required to build blocks under ePBS with BALs may favor large-scale builders with low-latency infrastructure. The block-building market was already concentrated pre-Glamsterdam, with two builders producing nearly 80% of blocks. Whether ePBS disperses or entrenches that concentration is a function of how quickly smaller builders adapt to the new protocol mechanics.
Buterin proposed mitigations including Forward Obligatory Commitment to Inclusion Lists (FOCIL), where randomly selected participants ensure certain transactions must be included, and encrypted mempools to prevent frontrunning and sandwich attacks. FOCIL is partially addressed by EIP-7805, which is included in Glamsterdam.
The broader observation, per Buterin: "As Ethereum scales, decentralization challenges are shifting from validators to the infrastructure that determines what actually lands onchain."
Ethereum's staking statistics provide context: approximately 37 million ETH (30.6% of circulating supply) is staked, with over 3 million ETH pending in the validator queue. BlackRock's ETHA ETF alone holds approximately $11 billion in ETH exposure. The institutional staking footprint is large and growing.
Ethereum's 2026 roadmap does not end with Glamsterdam. The Hegotá upgrade, targeted for the second half of 2026, focuses on Verkle Trees implementation — a data structure change targeting a 90% reduction in node storage requirements.
The Verkle transition addresses a core concern Buterin raised in January 2026: that "nodes went from easy to run to hard to run" over the past decade. Reducing storage requirements directly supports the network's decentralization goals by lowering the hardware bar for running a full node.
Combined, Glamsterdam and Hegotá represent Ethereum's most aggressive protocol development year since 2022. Whether the pace of change introduces execution risk is an open question. Developers have repeatedly stated that timeline adherence is secondary to upgrade safety.
Glamsterdam delivers what Ethereum's technical roadmap promised: a substantial capacity increase, protocol-native block building, and repriced gas economics. The upgrade's execution was clean. The market response was positive but measured — a 145% volume spike followed by price stabilization in the $2,370–$2,415 range, not the euphoric surge that preceded the Merge.
The harder problem — Layer 1 value accrual — remains unresolved. Tripling block capacity and cutting gas fees by 78% makes Ethereum mainnet more useful, but also makes it cheaper to use, which in the short term suppresses the fee burn that supports ETH's deflationary narrative. The network is processing more activity than ever while generating less revenue per transaction than ever. That tension does not have a protocol-level fix; it requires sustained demand growth that overwhelms the capacity expansion.
For now, Glamsterdam represents competent engineering execution on a complex multi-EIP hard fork. The economic thesis — that scaling L1 capacity eventually attracts enough activity to restore fee pressure — is a bet on demand elasticity that will take quarters, not days, to validate.