Ethereum is executing the most ambitious protocol overhaul since The Merge. In 2026, two back-to-back hard forks — Glamsterdam (H1) and Hegota (H2) — will fundamentally restructure how the network processes transactions, distributes MEV revenue, and resists censorship. This is not an incremental ...
"We want to push the gas limit toward and beyond 100 million in the first half of 2026… I predict it will double to 200 million following ePBS." — Tomasz Stańczak, Co-Executive Director, Ethereum Foundation
Ethereum is executing the most ambitious protocol overhaul since The Merge. In 2026, two back-to-back hard forks — Glamsterdam (H1) and Hegota (H2) — will fundamentally restructure how the network processes transactions, distributes MEV revenue, and resists censorship. This is not an incremental upgrade cycle. It is a structural reinvention of Ethereum's economic plumbing at a moment when the network faces an existential competitiveness crisis.
The stakes are enormous. Ethereum L1 fee revenue collapsed from peak daily levels of ~$23 million to roughly $6.3 million in 2025. The ETH token has been the worst-performing major cryptocurrency year-to-date in 2026. Solana has surpassed Ethereum in transaction volume. And the network's own Layer 2 ecosystem — led by Base, Arbitrum, and Optimism processing 90% of all L2 transactions — is cannibalizing the very fee revenue that secures the base layer. Glamsterdam and Hegota represent Ethereum's answer: scale the L1 by 3-10x, enshrine fair MEV distribution into the protocol itself, and hardwire censorship resistance so no government or relay operator can selectively exclude transactions.
The question is no longer whether Ethereum can ship these upgrades. Testnets are live. Client teams are building. The question is whether the upgrades arrive fast enough to reverse a narrative of structural decline — and whether the economic model that emerges on the other side is actually self-sustaining.
Ethereum enters 2026 under siege from multiple directions. The numbers tell a stark story:
Fee revenue erosion. The Dencun upgrade in March 2024 slashed L2 data posting costs by roughly 90%. This was celebrated as a user experience win — and it was. But it also cratered L1 fee revenue. Ethereum's daily gas revenue fell from peaks of ~$23 million to approximately $6.3 million in 2025, a structural decline that no amount of narrative engineering can mask. The L1 is generating roughly $2.3 billion annually in fees, down from over $8 billion at peak levels.
L2 cannibalization. The "Big Three" Layer 2s — Base, Arbitrum, and Optimism — now process nearly 90% of all L2 transactions, with Base alone commanding over 60% market share. Base was the only L2 to turn a net profit in 2025, earning approximately $55 million. The irony is brutal: Ethereum's most successful scaling strategy is also its biggest economic threat. Activity that once generated L1 fees now settles cheaply on rollups that capture the economic surplus for themselves.
Price underperformance. ETH has been the worst-performing major cryptocurrency in 2026. While Solana has surpassed Ethereum in daily transaction volume and Bitcoin maintains dominance above $1.8 trillion in market cap, Ethereum's ~$520 billion valuation reflects a market that increasingly prices in competitive decline rather than platform dominance.
MEV centralization. The builder market that determines transaction ordering has consolidated to alarming levels. As of January 2026, Titan controls 50.42% of all Ethereum blocks, BuilderNet controls 27.09%, and Quasar holds 14.04%. Three entities control over 91% of block construction. This is the kind of centralization that undermines Ethereum's core value proposition.
The Ethereum Foundation, under co-executive director Tomasz Stańczak (who announced he will step down at the end of February 2026), published a Protocol Priorities Update on February 18 that effectively declared a state of emergency. Three work tracks were defined: scale the L1, improve user experience, and harden the network. Glamsterdam and Hegota are the vehicles.
Glamsterdam, targeting deployment in the first half of 2026, is the largest execution-layer upgrade since The Merge. It comprises up to 22 Ethereum Improvement Proposals (EIPs) organized around two headline changes: Block Access Lists and Enshrined Proposer-Builder Separation.
Ethereum currently processes transactions sequentially — one at a time, in a single thread. This is the fundamental bottleneck that limits throughput. EIP-7928 introduces Block Access Lists, a mechanism that declares upfront which state each transaction will touch. This allows execution clients to preload necessary data into memory and process independent transactions simultaneously across multiple CPU cores.
The analogy is precise: Ethereum's current execution model is a single-lane road. Block Access Lists convert it into a multi-lane highway. Transactions that touch different parts of state can execute in parallel without conflict.
The practical impact is a gas limit increase from the current 60 million to approximately 100 million in H1 2026, with Stańczak projecting 200 million following ePBS deployment, and potentially 300 million by year-end. This represents a 3-5x throughput increase without raising hardware requirements proportionally, because the parallelism means existing hardware is used more efficiently.
For context, a 200 million gas limit translates to roughly 2,000-3,000 simple transactions per second on L1 — a dramatic improvement from today's ~15-30 TPS ceiling, though still below the 10,000 TPS long-term target that will require ZK-proof validation in later upgrades.
Development is progressing on schedule. The bals-devnet-2 testnet launched in early February 2026 with Lighthouse and Lodestar clients confirmed ready for testing. The epbs-devnet-0 testnet, incorporating all changes from consensus specification v1.7.0-alpha.2, is expected to go live by the end of February. Client teams including Erigon are conducting snapshot analysis and synchronization testing. Despite earlier concerns about Fusaka testing issues, developers have confirmed Glamsterdam remains on track.
The second headline change in Glamsterdam is EIP-7732: Enshrined Proposer-Builder Separation (ePBS). This is not a performance upgrade. It is a governance and economic restructuring of how Ethereum handles its most valuable — and most controversial — resource: Maximum Extractable Value.
Today, MEV is managed through an off-chain relay infrastructure, primarily MEV-Boost, built by Flashbots. Validators outsource block construction to specialized builders who compete to extract value from transaction ordering. The winning builder pays the validator a bid, and the relay intermediates the handoff.
This system works, but it concentrates power in a small number of trusted intermediaries. Of the seven major relays, only three do not censor transactions according to OFAC compliance requirements. The builder market itself is an oligopoly: three builders control 91% of blocks. And the relays themselves become single points of failure — if a relay goes down or acts maliciously, the entire system is compromised.
ePBS moves the proposer-builder split into the consensus protocol itself. Instead of relying on external relays, the protocol enforces a commit-reveal flow with explicit deadlines, payload commitments, and fallback behavior. Builders commit to block contents on-chain. Proposers select bids through a protocol-native auction. If a builder fails to deliver, the protocol handles the fallback without halting liveness.
The economic implications are significant:
Flashbots has already anticipated this shift, migrating its operations to BuilderNet in late 2024 and ceasing operation of centralized builders on Ethereum. ePBS essentially codifies what the market was already moving toward — but with protocol-level guarantees rather than trust assumptions.
The second fork of 2026, Hegota (a portmanteau of Heze and Bogotá, where Devcon was held), is scheduled for the second half of the year. Its headline feature is FOCIL — Fork-Choice Enforced Inclusion Lists — the most controversial proposal in Ethereum's current roadmap.
FOCIL addresses a problem that has haunted Ethereum since The Merge: transaction censorship. At peak levels in 2023, over 72% of MEV-Boost blocks were OFAC-compliant, meaning they excluded transactions involving sanctioned addresses. Even today, a significant percentage of blocks are built by relays that filter transactions based on regulatory compliance.
FOCIL works by randomly selecting 17 validators per block to submit inclusion lists of valid transactions they have observed in the mempool. Block producers must include these transactions or risk having their block rejected by the network. The mechanism ensures that any transaction deemed valid under Ethereum's rules will be included on-chain within a bounded timeframe — regardless of the builder's or relay's compliance preferences.
The decision to include FOCIL was confirmed at the February 19, 2026 All Core Devs call, with explicit backing from Vitalik Buterin. Ethereum Foundation researcher Alex Stokes presented the proposal as essential for maintaining Ethereum's credibility as a censorship-resistant settlement layer.
FOCIL is politically loaded. It effectively forces validators to include transactions from OFAC-sanctioned addresses, potentially exposing U.S.-based validators to regulatory risk. Critics argue this could drive compliant validators off the network or create legal liability for staking providers operating under U.S. jurisdiction.
Proponents counter that censorship resistance is not optional for a credibly neutral settlement layer. If Ethereum cannot guarantee that valid transactions will be processed, its value proposition as global financial infrastructure collapses. The tension between regulatory compliance and protocol neutrality is not new — but FOCIL forces the issue to a resolution.
Hegota is also expected to advance Verkle Trees, a data structure that allows Ethereum nodes to store and verify state more efficiently. This is a prerequisite for stateless clients — nodes that can validate blocks without maintaining the full state history — which would dramatically lower the hardware requirements for running a validator and further decentralize the network.
Viewed through the economic value distribution lens, Ethereum's 2026 upgrades represent a significant redistribution of rents:
Winners:
Losers:
The critical unknown is whether higher throughput translates to higher total fee revenue. If the gas limit triples but average gas prices drop proportionally, L1 revenue stays flat. Ethereum needs throughput increases to unlock new demand categories — not just redistribute existing demand at lower prices.
Glamsterdam (H1 2026) delivers parallel execution via Block Access Lists (EIP-7928) and protocol-native MEV auctions via ePBS (EIP-7732), targeting a 3-5x throughput increase from 60M to 200M gas limit.
Hegota (H2 2026) hardwires censorship resistance through FOCIL, requiring 17 randomly selected validators per block to enforce transaction inclusion regardless of sanctioned status.
The builder oligopoly — where three entities control 91% of Ethereum blocks — is directly challenged by ePBS, which replaces off-chain relay trust with protocol-level enforcement.
Ethereum L1 fee revenue has declined to ~$6.3M daily, and ETH is the worst-performing major crypto in 2026 YTD. These upgrades are a survival response, not optional improvements.
FOCIL forces a political resolution on censorship: validators must include OFAC-sanctioned transactions or risk block rejection, creating potential regulatory conflicts for U.S.-based operators.
The economic outcome depends on whether higher throughput generates new demand rather than simply repricing existing activity at lower margins — the same challenge that L2 fee compression already demonstrated.
Ethereum's 2026 upgrade cycle is the network's most consequential since the proof-of-stake transition. Glamsterdam and Hegota address the three vectors of existential risk simultaneously: throughput limitations that push users to competitors, MEV centralization that undermines credible neutrality, and censorship vulnerability that weakens the settlement-layer thesis.
The technical execution appears on track. Testnets are running. Client teams are delivering. The EIPs are finalized. But the harder question is economic: can a 3-5x throughput increase reverse the fee revenue decline caused by L2 cannibalization? Can protocol-native PBS actually break the builder oligopoly, or will the same players simply adapt? And will FOCIL's forced inclusion of sanctioned transactions trigger a regulatory backlash that fragments the validator set?
What is clear is that Ethereum has stopped pretending the status quo is acceptable. The network that built decentralized finance, launched the NFT economy, and pioneered rollup scaling is now rebuilding its own execution engine while the plane is in flight. The 2026 forks will determine whether Ethereum remains the settlement layer of Web3 — or becomes the MySpace of smart contract platforms, remembered for pioneering a category it ultimately lost.