Platåberget launched on August 13 and was formally announced by the Ethereum Foundation on August 17. Unlike the short-lived development networks used during earlier Glamsterdam work, it is intended to remain available for months and accepts public validators.

It is still a laboratory network. That distinction matters because Glamsterdam has not reached Sepolia, Hoodi or Ethereum mainnet yet. The point of Platåberget is to expose client software, staking infrastructure, builders and downstream applications to the new rules early enough for something to fail without involving real assets.

Gas repricing reaches far beyond node operators

Glamsterdam includes a coordinated repricing of Ethereum operations as developers prepare the network for substantially higher L1 capacity. State creation is an important part of that work: creating accounts, deploying code and writing new storage imposes a permanent database cost on every node, so those actions are being accounted for differently.

That immediately reaches software that most users never think about. Wallets estimate transaction costs. Indexers reproduce assumptions about block capacity. Bundlers, RPC middleware and internal transaction calculators may carry constants that were perfectly reasonable when they were written.

The Foundation says tools relying on a hardcoded maximum gas limit will break and need updating. It also specifically asks developers to revisit assumptions around fixed transfer costs and single-dimension gas estimation. The exact repricing table is still something developers should take from the current testnet specifications rather than an old implementation.

Contract deployment gets more room as well. The maximum deployed contract size in the Glamsterdam test configuration rises from 24 KiB to 64 KiB, while the initcode limit moves from 48 KiB to 128 KiB.

The bigger protocol change happens in block production

Enshrined Proposer-Builder Separation, EIP-7732, moves the proposer-builder handoff into Ethereum itself. Today that relationship relies heavily on external infrastructure and middleware. Glamsterdam introduces protocol-level builder mechanics, a new builder API flow and payload-timeliness checks.

Ethereum's roadmap describes another practical benefit: the propagation window available around the execution payload can grow from roughly two seconds to about nine seconds. That gives the network more room to move larger payloads without forcing every part of the validation path through the same narrow timing window.

Block-Level Access Lists attack a different bottleneck. EIP-7928 records which accounts and storage locations a block touches, allowing clients to identify independent work before executing it. That prepares Ethereum for parallel disk reads and, eventually, more parallel transaction processing.

Platåberget is currently active with an open validator set and is expected to survive until the Glamsterdam mainnet transition. Before that happens, the plan still includes further devnet testing followed by upgrades on Sepolia and Hoodi. Ethereum.org currently places Glamsterdam in Q4 2026; no exact mainnet activation date has been locked.