Ethereum's Sepolia testnet activated the Glamsterdam upgrade on October 6, 2026, at 13:53:36 UTC, landing at epoch 353,024. The fork gives validators the option to raise the network's gas limit to 200 million, a significant increase from the previous ceiling of roughly 60 million.
Glamsterdam combines two upgrade tracks into a single package. The first, Amsterdam, covers the execution layer—the part of Ethereum that processes transactions and runs smart contracts. The second, Gloas, covers the consensus layer, where validators agree on which blocks are valid and their order.
Key Technical Changes
The headline feature is enshrined proposer-builder separation, or ePBS, defined in EIP-7732. In this structure, the builder assembles a block full of transactions, while the proposer is the validator who puts that block forward to the network. Being "enshrined" means the split now lives inside the protocol itself rather than in add-on software running alongside it.
The upgrade also includes block-level access lists and a round of gas repricings. Access lists help nodes know ahead of time what parts of Ethereum's state a block will touch. Gas repricings adjust what certain operations cost, ensuring fees better match the computational work demanded from the network.
Gas Limit Expansion
Gas is Ethereum's unit for measuring computational work, and the gas limit caps how much work can fit into a single block. The proposed increase from roughly 60 million to 200 million represents more than a threefold capacity boost.
Critically, validators can opt in to the 200 million target on Sepolia. Mainnet activation for Glamsterdam has not yet been scheduled, and the final gas limit parameters for mainnet have not been decided. There are no immediate changes to mainnet gas limits.
Testing and Deployment
Developers ran the upgrade through earlier test environments, including Platåberget and Devnet 8, to validate core functionality before reaching the public testnet. Consensus client teams including Prysm pushed last-minute patch releases to support the fork, with Prysm shipping version 7.2.1.
Implications
Testnet forks do not affect mainnet operations or prices. A gas limit of 200 million, paired with ePBS and access lists, demonstrates how developers aim to expand Layer 1 capacity by bundling structural changes to block production with raw capacity increases.
Bigger blocks require more data and computation for every node operator. If hardware demands climb too fast, smaller operators could face challenges maintaining participation, raising decentralization concerns. The opt-in structure on Sepolia allows developers to observe validator behavior and node performance under heavier blocks before committing to mainnet changes.
The primary open question remains the final mainnet gas limit. Sepolia's 200 million figure serves as a test target rather than a confirmed parameter for mainnet deployment.
