Ethereum sets October 26 for Glamsterdam on the Hoodi testnet, with the gas limit rising to 200 million

What happened

Ethereum's next network upgrade now has its second public rehearsal dated. The Ethereum Foundation announced on October 16 that the Glamsterdam upgrade is scheduled to activate on the Hoodi testnet on October 26, 2026, at 17:42:48 UTC, at epoch 132,352 and slot 4,235,264. At the same epoch, Hoodi's gas limit will be raised to 200 million, matching Sepolia. Glamsterdam activated on the Sepolia testnet on October 6, at epoch 353,024. The mainnet activation date has not been decided, and the Foundation was explicit that this announcement does not schedule a mainnet upgrade. A separate announcement will cover mainnet when client teams decide it.

Glamsterdam combines two upgrades: Amsterdam on the execution layer and Gloas on the consensus layer. Its two headline changes restructure how blocks are built and checked. The first, enshrined proposer builder separation, known as ePBS and specified in EIP-7732, brings the separation between block proposers and builders into the consensus protocol itself. A proposer includes a builder's commitment to an execution payload, and the builder reveals the payload afterwards. The protocol handles payment to the proposer, reducing reliance on trusted middleware for that exchange. The change also separates consensus validation from execution validation, giving validators more time to validate execution payloads, and a payload timeliness committee attests to whether the builder revealed its payload and the corresponding blob data on time.

The second headline change, block level access lists in EIP-7928, requires each block to carry an enforced record of the accounts and storage locations accessed during the block, together with post transaction state changes. Those lists let clients read state from disk in parallel, validate transactions in parallel and compute state roots more efficiently. Together, the Foundation says, the two changes provide a foundation for scaling execution capacity while keeping block validation practical for node operators.

The third strand is gas pricing. EIP-8037 increases the cost of creating state and meters it separately, and EIP-8038 updates state access costs, alongside further changes to intrinsic transaction gas, calldata, access lists and block gas accounting, so that prices better reflect the cost of execution and state growth. The rest of the scheduled package, 18 EIPs in the tracking list, includes ETH transfer logs, a slot number opcode, a higher maximum contract size, a deterministic factory contract, new stack manipulation instructions, excluding slashed validators from proposing, and increased exit and consolidation churn. The upgrade's bug bounty program is active for its specifications, and client implementations become eligible as their releases are confirmed.

The announcement is also an operator deadline. Node operators must update both their execution and consensus layer clients before activation, and validators must update both the beacon node and the validator client, then review new duties, builder tooling compatibility and gas limit settings. The confirmed release table is still filling in: on the consensus side, Lodestar 1.50.0, Nimbus 26.10.1, Prysm v7.3.0 and Teku 26.10.1 are listed, with two other clients not yet listed. On the execution side, Besu 26.10.0, Ethrex 30.0.0 as a release candidate and Nethermind 2.2.0 are listed, with three other clients not yet listed. The Foundation's instruction is to use only a release whose notes explicitly confirm support for the Hoodi activation. Holders of ether and mainnet operators have nothing to do in response to this announcement.

Why it matters

Glamsterdam is Ethereum's layer 1 scaling roadmap expressed as a single upgrade, and its design choice is visible in the pairing of its parts. The network is not simply raising the gas limit and asking every node to do more work. It is changing the structure of block production first, moving the builder market inside the protocol and giving clients a map of everything a block touches so work can be parallelized, and then raising the per block gas budget to 200 million on the testnets where that structure runs. Whether capacity bought with structure stays cheap to validate is the question the whole design answers, and Hoodi is where the answer gets measured under validator heavy, mainnet like conditions. Sepolia, which activated on October 6, is the developers' rehearsal. Hoodi is the staking infrastructure rehearsal, the last public testnet step in the schedule before a mainnet date can be set.

The builder change carries a trust argument as well as a capacity one. Today, the exchange between proposers and builders runs largely through middleware outside the protocol, which proposers and builders have to trust to relay payloads and payments honestly. Enshrining that exchange means the protocol itself handles the commitment, the reveal and the payment, with a committee attesting to timeliness. That removes a layer of trusted intermediaries from the most economically sensitive moment in block production, the moment where the right to order transactions is sold. It also creates new duties and new failure modes, which is why validator operators are being told to review builder tooling rather than simply upgrade a binary.

The gas repricing is the part most likely to touch people who never run a node. Contracts that rely on fixed gas stipends, hardcoded gas limits or assumptions about how much gas remains during execution may behave differently under prices that charge separately for creating state and more for accessing it. The Foundation publishes a repricing impact guide and an affected contracts search precisely because real, deployed contracts carry those assumptions. A scaling upgrade that quietly breaks a class of deployed contracts would trade one problem for another, so the repricing has to be tested application by application, and the October 26 date is the deadline that organizes that testing on Hoodi.

What to watch

Watch the activation itself on October 26 at 17:42:48 UTC. A clean Hoodi fork, with clients following the new rules and finality uninterrupted, is the baseline result. Client diversity in the first epochs after activation, and any payload timeliness problems under ePBS, are the details that will decide how the rehearsal is judged.

Watch the release table. Several major clients on both layers were still unlisted when the announcement was published, and operators are told to use only releases that explicitly confirm Hoodi support. Whether every major client ships a confirmed release with time to spare, or whether some operators are upgrading at the deadline, is a practical measure of how ready the ecosystem is.

Watch the bug bounty and the repricing reports. Findings against the Glamsterdam specifications are eligible for the bounty program as client releases are listed, and application teams testing against the new gas rules on Sepolia and Hoodi will surface the contracts that fixed gas assumptions break. Either category of finding, if serious, is the kind of evidence that moves a mainnet date.

And watch for the mainnet decision itself. The Foundation has left it explicitly open, to be announced once client teams decide. Hoodi's performance is the last scheduled input into that decision, so the first mainnet timing discussion after October 26, and whether the 200 million gas level is proposed for mainnet or held as a testnet setting, will show how much of Glamsterdam's ambition survived contact with a live validator set.