Ethereum's Platåberget public testnet went live on August 17, 2026, opening community testing for the Glamsterdam hard fork — the largest protocol overhaul since the September 2022 Merge. The testnet transitioned to the Glamsterdam fork on August 20. Mainnet activation targets Q4 2026, though cor...
"Ethereum needs to do more to meet its own stated goals. Not the quest of 'winning the next meta' regardless of whether it's tokenized dollars or political memecoins." — Vitalik Buterin, Co-Founder, Ethereum
Ethereum's Platåberget public testnet went live on August 17, 2026, opening community testing for the Glamsterdam hard fork — the largest protocol overhaul since the September 2022 Merge. The testnet transitioned to the Glamsterdam fork on August 20. Mainnet activation targets Q4 2026, though core developers have stated that correctness outranks any fixed date.
Glamsterdam bundles 13 EIPs across two architectural changes and a gas-repricing package. The headline proposals — Enshrined Proposer-Builder Separation (EIP-7732) and Block-Level Access Lists (EIP-7928) — aim to move block-building auctions on-chain, enable parallel transaction execution, and set a floor for the gas limit at 200 million, more than triple the current 60 million ceiling. A third package of gas-cost recalibrations (EIP-7904) is projected to cut L1 transaction fees by approximately 78.6%.
The upgrade carries direct implications for the MEV supply chain, validator economics, Layer 2 data availability costs, and the 24 KiB smart contract size ceiling that has constrained developers since 2016. According to Tomasz Stańczak, co-director of the Ethereum Foundation, the gas limit could reach 300 million before year-end.
Glamsterdam is a combined consensus-layer (CL) and execution-layer (EL) hard fork. It addresses three structural bottlenecks simultaneously:
Throughput. The current 60 million gas limit constrains Ethereum L1 to roughly 15–30 standard transactions per second. Glamsterdam's parallel-execution framework targets a credible path to 200 million gas and approximately 10,000 TPS.
MEV extraction. Today, over 90% of Ethereum blocks are built through off-chain MEV-Boost relays operated by a handful of entities including Flashbots, BloXroute, and Ultrasound. This creates a trust dependency external to the protocol. Glamsterdam moves the proposer-builder handoff on-chain.
Fee misalignment. Many EVM opcode gas costs were calibrated years ago and no longer reflect actual computational overhead on modern hardware. EIP-7904 recalibrates these costs, with the net effect of reducing fees for simple ETH transfers and complex smart contract interactions by an estimated 78.6%.
As of August 23, 2026, Ethereum L1 gas sits at 0.074 gwei — already near historic lows following prior upgrades (Dencun, Pectra). Glamsterdam's repricing would push costs lower still. According to multiple estimates, a Uniswap trade currently costing $3–$8 in gas could fall below $1 post-upgrade.
EIP-7732 formally separates the proposer role (selecting the consensus block) from the builder role (assembling the execution payload) at the protocol level. Under the current system, this separation exists only through MEV-Boost, an off-chain middleware maintained by Flashbots. Validators trust relays to honestly broker the exchange of block payloads for builder payments.
ePBS removes the relay dependency. A builder commits to an execution payload and a payment on-chain; the proposer selects the highest valid bid. The protocol enforces the exchange trustlessly.
The upgrade also introduces the Payload Timeliness Committee (PTC), a new validator duty. A subset of validators attests to whether the builder's payload arrived within the expanded propagation window. This window increases from approximately 2 seconds to 9 seconds, providing builders more time to construct optimal blocks.
According to multiple analyses, ePBS is projected to reduce MEV extraction by up to 70% by eliminating the asymmetric information advantage that off-chain relays currently enable.
Parallel transaction execution is constrained by the fact that the EVM does not know which transactions will access the same state. EIP-7928 addresses this by requiring pre-declaration of state access patterns at the block level.
When validators and clients know in advance which storage slots, accounts, and contracts a block will touch, they can identify non-conflicting transactions and execute them concurrently. This is the technical prerequisite for raising the gas limit from 60 million toward 200 million without proportionally increasing node hardware requirements.
The target: approximately 10,000 TPS on L1, up from the current effective rate.
EIP-7904 realigns gas costs with actual computational resource consumption. The recalibration produces an estimated 78.6% reduction in gas costs for both simple ETH transfers and multi-step smart contract interactions.
This repricing interacts with two companion EIPs:
The Glamsterdam meta-EIP (EIP-7773) schedules the following 13 proposals for inclusion:
| EIP | Description | Layer | |-----|-------------|-------| | EIP-7732 | Enshrined Proposer-Builder Separation | CL | | EIP-7928 | Block-Level Access Lists | EL | | EIP-7904 | Gas Repricing | EL | | EIP-2780 | Reduce Intrinsic Transaction Gas | EL | | EIP-7954 | Increase Maximum Contract Size (24→64 KiB) | EL | | EIP-7976 | Increase Calldata Floor Cost | EL | | EIP-7981 | Increase Access List Cost | EL | | EIP-8037 | State Creation Gas Cost Increase | EL | | EIP-7778 | Block Gas Accounting Without Refunds | EL | | EIP-8024 | Backward-Compatible SWAPN, DUPN, EXCHANGE | EL | | EIP-7843 | SLOTNUM Opcode | EL | | EIP-7708 | ETH Transfers Emit a Log | EL | | EIP-8061 | Increase Exit and Consolidation Churn | CL |
EIP-7610 (collision resistance for CREATE/CREATE2) is also referenced in some listings.
Completed milestones:
Upcoming milestones:
Core developers have been explicit that the testing timeline is not compressed. Prior Ethereum forks required two to four months of public testnet seasoning. Applying this pattern to the August 17 Platåberget launch suggests a mainnet window of October–December 2026.
The Ethereum Foundation's original internal target was end of August 2026 for mainnet. This is now widely considered aspirational. According to a CoinMarketCap report, the upgrade was pushed from Q2 to Q3 and may extend into Q4 depending on testnet stability.
ePBS renders MEV-Boost relays structurally unnecessary. The on-chain auction mechanism replaces the trusted relay model. Validators currently running MEV-Boost — which includes the vast majority of professional staking operations — must transition to the native ePBS mechanism.
This represents a direct economic disruption to relay operators. Flashbots, BloXroute, Ultrasound, Agnostic, and other relay providers face a protocol-level displacement of their core product.
The Payload Timeliness Committee (PTC) introduces a new attestation duty. Validators are randomly selected to attest to whether a builder's payload arrived within the propagation window. This adds operational complexity, particularly for solo stakers.
EIP-8061 increases the exit and consolidation churn limit. This is directly relevant to institutional stakers holding large positions. Current exit queue constraints have been identified as a systemic risk — particularly following cases where single entities accumulated millions of staked ETH. The increased churn limit allows faster unwinding of large staking positions.
All validators must update both consensus-layer clients (Lighthouse, Prysm, Teku, Nimbus, Lodestar) and execution-layer clients (Geth, Nethermind, Besu, Erigon) to Glamsterdam-compatible versions before mainnet activation. Failing to update will cause validators to fork off the canonical chain.
Glamsterdam's impact on Layer 2 networks operates through two channels:
Data availability budget. The expanded ePBS payload-propagation window (2 seconds → 9 seconds) creates room for higher blob capacity in future Blob Parameter Only (BPO) forks. While Glamsterdam itself does not directly increase blob count, it unblocks the infrastructure necessary for subsequent blob expansions, reducing data-posting costs for rollups over time.
Calldata cost pressure. EIP-7976's increase in calldata floor costs makes calldata-based data posting relatively more expensive. Rollups that have not yet migrated to EIP-4844 blob transactions face higher L1 settlement costs post-Glamsterdam, creating additional economic pressure to complete the migration.
The combined effect: rollups using blobs benefit from an improving data availability trajectory; rollups still on calldata face rising costs.
EIP-7954 raises the maximum deployed contract size from 24 KiB to 64 KiB and the maximum initcode size from 48 KiB to 128 KiB. The 24 KiB limit, in place since EIP-170 (2016), has forced developers to split contract logic across proxy patterns, delegatecall architectures, and the Diamond Standard (EIP-2535) — even when simpler single-contract deployments would suffice.
EIP-7907 introduces gas metering for code exceeding 24 KiB at 2 gas per 32-byte word, preventing DoS attacks while permitting larger contracts.
EIP-7843 exposes the current slot number to the EVM via a new SLOTNUM opcode. EIP-8024 adds backward-compatible SWAPN, DUPN, and EXCHANGE opcodes for cheaper stack manipulation.
EIP-7708 mandates that all non-zero ETH transfers and burns emit a standard log event. This eliminates the need for wallets, exchanges, and bridge operators to use custom tooling to track native ETH movements — a long-standing developer pain point.
The Ethereum Foundation has issued a specific warning: the 21,000 gas assumption for simple ETH transfers will no longer hold after Glamsterdam. Wallets, gas estimators, and dApps that hardcode this value will break. According to developer communications, this is expected to be the most disruptive user-facing change in the upgrade.
ePBS complexity. Enshrined PBS is the single most complex consensus-layer change since the Merge. The Ethereum Foundation has acknowledged that getting ePBS right outranks hitting any fixed date. Multi-party collusion between proposers and builders, and the risk of chain stagnation if builders fail to deliver payloads, remain areas of active research.
Gas limit governance. The 200 million gas target requires validator consensus on the gas limit parameter. The current 60 million level was itself the product of a contested validator signaling process. Scaling to 200 million — and potentially 300 million by year-end per Stańczak's comments — requires sustained validator coordination.
Node hardware requirements. Parallel execution reduces the per-transaction computational burden, but 200 million gas blocks still represent a 3.3x increase in total block capacity. Whether consumer-grade hardware can process these blocks without degrading sync times or increasing orphan rates remains to be validated on testnets.
Wallet and tooling breakage. The EIP-2780 gas changes and EIP-7708 logging changes will require updates across the ecosystem. Any wallet, exchange, or dApp that hardcodes gas assumptions faces potential breakage at fork activation.
EIP-8363 monetary policy conflict. Separately from Glamsterdam, the EIP-8363 tapered issuance burn proposal — published August 4 and authored by six researchers including the Ethereum Foundation's Justin Drake — has triggered significant opposition from DeFi protocols including Aave. While not part of Glamsterdam itself, the simultaneous debate over issuance policy adds governance friction during a critical testing period.
Glamsterdam is the most architecturally ambitious Ethereum upgrade since the Merge. It touches consensus, execution, gas economics, MEV infrastructure, and developer tooling simultaneously. The scope is broad by design — ePBS requires BALs to enable the gas limit increase, and gas repricing must accompany the capacity expansion to maintain fee accessibility.
The testing infrastructure is now in place. Platåberget provides the first public environment for stress-testing these changes. Sepolia and Hoodi forks will follow before any mainnet consideration.
The economic implications are material. MEV relay operators face displacement. Validators gain new duties. Institutional stakers get faster exit queues. Rollups on calldata face rising costs. And every wallet and gas tool in the ecosystem needs to update for the post-21,000-gas world.
Whether Glamsterdam ships in Q4 2026 or slips into early 2027 depends on testnet performance in the coming weeks. The Ethereum Foundation has made its position clear: correctness over speed. The data from Platåberget will determine the timeline.