Final Stretch Before Mainnet
Core developers have begun full-scale testing of a fork that bundles each Ethereum Improvement Proposal (EIP) earmarked for Glamsterdam. The work marks the last stretch before the codebase is hardened and shipped to public testnets, with mainnet activation now expected in the second half of 2026. Ethereum Foundation core developer Parithosh Jayanthi postured the release as a turning point for the network, adding:
“Probably the largest fork we’ve had since the Merge. It could change a lot of assumptions about Ethereum and set us up for much more scaling.”
The upgrade’s name stitches together two simultaneous changes, i.e. the Amsterdam upgrade on Ethereum’s execution layer and the Gloas upgrade on its consensus layer. Both are due to ship in the same hard fork, a pattern the network has used to coordinate execution- and consensus-layer work since its move to proof-of-stake.
What ePBS and Access Lists Change
At the center of Glamsterdam sit two headline proposals. The first being EIP-7732, which introduces Enshrined Proposer-Builder Separation (ePBS) and pulls the work of building and proposing blocks into Ethereum’s core protocol. Developers say the change narrows the room for maximal extractable value (MEV) manipulation, which today leans on offchain relays that carry centralization concerns.
For validators, ePBS is meant to level the field by guaranteeing payment for proposed blocks even when a separate builder assembles them, trimming the leverage that sophisticated players hold over ordinary stakers (dovetailing Ethereum’s wider push toward distributed block building).
The second proposal, EIP-7928, adds Block-Level Access Lists, which as the name suggests, allows each block to declare in advance which accounts and contract data it will touch, so client software can preload the information and process transactions in parallel. The payoff, developers say, is block execution that is faster, cheaper to optimize, and more predictable.
Bigger Blocks, Cheaper Calls
Glamsterdam also carries a sweeping gas repricing designed to make fees track the real cost of computation more closely and to clear a path for zero-knowledge proving systems that could verify blocks far faster. The gas limit is set to climb from 60 million to 200 million per block, while throughput targets reach toward 10,000 transactions per second (TPS), roughly 10 times what the network clears today.
Backers project a 78.6% drop in fees across both simple transfers and complex smart contract calls, with high-level compute operations getting cheaper and raw state access getting more expensive.
Those targets would mark a step change for an execution layer that has long traded raw throughput for decentralization. Cheaper, more abundant block space tends to flow through to lower costs on L2 rollups, which settle their data on Ethereum, the same dynamic that followed last December’s Fusaka upgrade (which widened data capacity for rollups).
Developers have already named the next milestone after Glamsterdam, i.e. Hegota, a sign of how quickly the roadmap is moving. However, it does bear mentioning that Glamsterdam was first penciled in for H1 2026 before developers pushed activation, citing the scale of the changes. No firm date has been locked in, and the testnet phase will set the final schedule.
