Ethereum’s 2026 roadmap matters less as a simple throughput story than as a market-structure test: the network is trying to add capacity without handing more power to a narrow set of builders, validators, or infrastructure operators. The two planned hard forks, Glamsterdam in the first half of 2026 and Hegota in the second half, pair scaling features with explicit anti-censorship and decentralization measures, which is the part traders, node operators, and institutions should verify rather than assume.
Glamsterdam’s throughput gain depends on how parallel execution lands
Glamsterdam is expected to introduce Block Access Lists through EIP-7928, a change designed to let Ethereum clients execute non-overlapping transactions in parallel. If implementation matches the proposal’s intent, the network could use modern multi-core hardware more efficiently and push block gas limits from roughly 60 million toward 200 million gas per block, a large jump that would directly affect congestion conditions and transaction inclusion during busy periods.
That does not make this a simple “faster Ethereum” upgrade. Parallel execution only helps where transactions do not contend for the same state, so realized gains will depend on transaction composition, client performance, and where gas-limit increases actually settle after launch. For DeFi and rollup-linked activity, the practical signal is not the headline gas target itself but whether higher block capacity arrives without creating new validator hardware concentration or unstable execution patterns.
ePBS is the clearest clue that Ethereum is solving for neutrality, not just speed
A second Glamsterdam feature, enshrined proposer-builder separation, would move MEV-Boost-like block construction logic closer to the protocol. That matters because Ethereum’s MEV market already sits at the intersection of liquidity, ordering power, and censorship risk. Formalizing the separation between block proposers and block builders is meant to reduce reliance on off-protocol coordination while making the block-building market more contestable.
The distinction is important for anyone evaluating Ethereum against rival layer-1s or for institutions assessing execution risk. If ePBS works as intended, it should lower the chance that a small set of actors can dominate order flow or apply transaction filtering under commercial or regulatory pressure. The relevant checkpoint after Glamsterdam will be observable market behavior: whether builder participation broadens, whether censorship concerns ease, and whether MEV extraction becomes less concentrated rather than merely better hidden.
Hegota targets node economics and inclusion policy
Hegota, planned for the second half of 2026, shifts from immediate throughput to the cost of participating in Ethereum’s base layer. Its headline item is Verkle trees, a long-discussed data structure change that would support near-stateless clients. In operational terms, the goal is much lower storage requirements and much faster syncing, making it easier to run validating or full-node infrastructure without enterprise-grade hardware.
That is a decentralization issue before it is a developer convenience issue. If node costs keep rising, neutrality weakens because fewer independent actors can verify the chain. Verkle-based statelessness is Ethereum’s attempt to prevent scaling from turning into a hardware arms race. Hegota also includes Fork-Choice Inclusion Lists, or FOCIL, which would let validators coordinate on transaction inclusion under the fork-choice process. Together with ePBS, that creates a layered response to censorship: one change addresses who builds blocks, the other addresses whether transactions can still reach blocks when filtering pressure appears.
| Upgrade stage | Named features | Primary mechanism | What to verify in practice |
|---|---|---|---|
| Dencun, 2024 | EIP-4844 proto-danksharding, blob transactions | Cheaper data for rollups | Whether lower rollup costs persist and support the next scaling layers |
| Glamsterdam, H1 2026 | EIP-7928 Block Access Lists, ePBS | Parallel execution plus protocol-level builder/proposer separation | Actual gas-limit increases, execution stability, MEV market concentration, censorship outcomes |
| Hegota, H2 2026 | Verkle trees, FOCIL | Near-stateless clients and validator-assisted inclusion guarantees | Finalized proposal scope, node cost reductions, inclusion behavior under stress or filtering attempts |
The roadmap only works if the anti-censorship pieces ship with the scaling pieces
The easy misread is to treat Glamsterdam and Hegota as one more sequence of fee and speed upgrades. The roadmap Ethereum has laid out is more strategic than that. Dencun in 2024, with proto-danksharding under EIP-4844, already reduced rollup costs through blob transactions; the 2026 plan extends that path while also trying to keep verification cheap and transaction ordering less centralized. Without that second half of the equation, more throughput could still leave Ethereum more fragile at the validator and builder layer.
That is why finalized feature scope matters. Hegota’s Verkle tree and FOCIL work is still the place where implementation details could change the risk profile materially, and Glamsterdam’s real-world effects will hinge on post-upgrade gas settings and MEV market outcomes rather than roadmap language. For infrastructure providers, exchanges, and institutional users, the operational question is not whether Ethereum is upgrading, but whether these upgrades reduce dependence on concentrated intermediaries while preserving predictable execution.
Immediate checks for traders, validators, and infrastructure teams
For market participants, the most useful lens is to separate narrative from measurable signals. A credible bullish interpretation would involve higher effective throughput after Glamsterdam, stable client performance, and no obvious worsening in validator or builder concentration. A weaker outcome would be a headline gas-limit increase that improves benchmarks but pushes more dependence onto specialized operators or leaves censorship pathways largely unchanged.
Validators and node operators should watch hardware assumptions and client readiness as closely as protocol announcements. Hegota’s promise of lighter nodes is meaningful only if Verkle implementation is mature enough to reduce sync and storage burden in practice. For funds, treasury teams, and compliance-sensitive institutions, the key checkpoint is whether ePBS and FOCIL materially improve neutrality under pressure, because that affects settlement confidence more than a nominal reduction in fees.
Quick checks
When does the new cadence start to matter? The shift is visible in 2026 because Ethereum is planning two distinct hard forks in one year: Glamsterdam in H1 and Hegota in H2.
What is the biggest misconception? Treating the roadmap as a pure speed upgrade. The core design goal is to add scale while reducing MEV centralization, censorship risk, and node-cost pressure.
What should readers monitor first after Glamsterdam? Actual gas-limit changes, client stability under parallel execution, and whether block building becomes less concentrated after ePBS.
Where is the main unresolved area? Hegota’s final feature set, especially the implementation details around Verkle trees and FOCIL, because those determine whether the promised decentralization benefits are real.
About the Author
