Ethereum’s Glamsterdam upgrade has activated on the Sepolia testnet, moving one of the network’s most consequential scaling packages into a long-lived public testing environment. The fork went live at epoch 353,024, slot 11,296,768, at 13:53:36 UTC on October 6. According to the Ethereum Foundation’s official Glamsterdam testnet announcement, Sepolia is now testing enshrined proposer-builder separation, block-level access lists and a substantially higher recommended gas-limit target together. Hoodi and mainnet activation dates remain undecided.
Public Sepolia beacon data checked after the transition showed the network continuing to finalize blocks, with finalized epoch 353,031 already seven epochs beyond the fork. That provides an early indication that Sepolia crossed the Glamsterdam boundary without an immediate consensus stall, although longer observation is still needed before drawing conclusions about client stability or performance under heavier workloads.
Prysm Moves Validators Toward 200M Gas
Prysm v7.2.1, released one day before the activation, added Sepolia’s new GAS_LIMIT_SCHEDULE. Validators using the client without a custom override now default to a 200 million gas preference beginning at the Gloas fork. The release removes the manual configuration requirement that existed in Prysm v7.2.0, which supported Glamsterdam but continued defaulting to 60 million gas. Operators can still specify another value through proposer settings, the keymanager API or the relevant client option.
The 200 million figure should not be interpreted as an instantaneous protocol-enforced block limit. EIP-8261 defines the schedule as optional client guidance, not a new consensus validity rule. Individual validator preferences remain configurable, while Ethereum’s existing gas-limit adjustment mechanism restricts changes to ±1/1024 from one block to the next. Sepolia’s realized block gas limit therefore moves progressively toward the new target rather than jumping directly from roughly 60 million to 200 million at the fork.
That distinction makes Sepolia useful as a capacity test. Glamsterdam’s architecture is intended to make larger execution workloads more practical through EIP-7732’s enshrined proposer-builder separation and EIP-7928’s block-level access lists. The latter exposes state-access information that can support parallel disk reads, transaction validation and state-root computation. These mechanisms have already been central to Glamsterdam’s earlier multi-client devnet testing. The higher gas target is being tested alongside structural changes intended to make heavier blocks easier for clients to process.
Sepolia Tests Capacity Before Mainnet
Glamsterdam has reached Sepolia only after multiple earlier development environments exposed client and networking problems. Devnet testing previously surfaced peer-recovery and connectivity weaknesses in Lodestar, illustrating why public-testnet deployment is a separate milestone from successful internal benchmarking. A clean fork transition does not yet establish that clients can sustain 200 million gas blocks reliably under adverse or production-like conditions.
The test also sits within a broader shift toward expanding Ethereum Layer 1 capacity. Work on ePBS, block-level access lists and gas repricing has increasingly moved scaling attention back toward the base layer, alongside the network’s continued rollup roadmap. That evolution is reflected in Ethereum’s renewed emphasis on scaling Layer 1 execution rather than treating rollups as the only path to higher throughput. Sepolia now provides the first long-lived public environment where those Glamsterdam changes can be evaluated together at substantially higher targeted capacity.
For operators, the immediate requirement was software compatibility rather than manually selecting 200 million gas. The Ethereum Foundation required Sepolia nodes to update both execution and consensus clients before the fork, while Prysm v7.2.1 incorporated the gas schedule automatically. Its release also included Gloas builder-configuration fixes and enabled partial data-column dissemination by default. Those client changes matter because Glamsterdam alters block construction and validation workflows, not merely a single network parameter.
The activation therefore marks the start of the more useful part of the experiment. Ethereum developers can now observe validator preferences, realized gas-limit growth, builder behavior, propagation performance and client resource requirements on Sepolia before setting parameters for later networks. Nothing in the current rollout commits Ethereum mainnet to a 200 million gas limit, and the Foundation has not yet scheduled Glamsterdam for either Hoodi or mainnet.