The roadmap for Ethereum for 2026 outlines a strategic focus on two key areas: enhancing rollup data capacity through blobs and increasing base-layer execution through gas limit adjustments. Central to these gas limit changes is the transition of validators from re-executing blocks to verifying zero-knowledge (ZK) execution proofs. The initial phase is underpinned by the launch of Fusaka, which was implemented on December 3, 2025.
Fusaka introduces PeerDAS along with blob parameter only (BPO) changes aimed at incrementally improving blob throughput, as detailed on ethereum.org. The secondary aspect of the roadmap is less defined, dependent on the progression of draft Ethereum Improvement Proposals (EIPs), client implementations, and validator operations that must adhere to decentralization constraints such as bandwidth, block propagation, and market structure for proving.
PeerDAS is considered a vital mechanism to enhance rollup data availability, as it allows nodes to avoid downloading every blob. Ethereum”s blob targets are designed to gradually increase, not immediately, doubling every few weeks to reach a cap of 48 blobs per block, contingent on network health monitoring by developers. The Optimism team has characterized this upper limit as “at least 48 blob target per block,” which could potentially elevate rollup-side throughput from approximately 220 to around 3,500 user operations per second (UOPS) under optimal conditions.
A pressing question for 2026 is whether demand will manifest through blob usage rather than escalating Layer 1 execution fees. Another concern is if peer-to-peer stability and node bandwidth will remain within acceptable limits as BPO deployment progresses. On the execution front, Ethereum is currently testing higher throughput via coordinated efforts instead of undergoing a hard fork. Recent data from GasLimit.pics indicates a gas limit of 60,000,000, with a 24-hour average around 59,990,755. This figure is crucial as it reflects validators” accepted practices and highlights the upper threshold of “social scaling” before issues like latency, validation load, and the MEV pipeline strain become significant.
In translating gas limits to throughput, using Ethereum”s 12-second slot time helps clarify gas per second, which is derived from dividing the gas limit by 12. The scheduled upgrade for 2026, termed Glamsterdam, encompasses various execution-focused proposals including enshrined proposer-builder separation (ePBS, EIP-7732), Block-Level Access Lists (BALs, EIP-7928), and general repricing (EIP-7904). Currently, each of these proposals remains in draft status, as outlined on their respective EIP pages.
The repricing initiative aims to address long-standing gas schedule discrepancies, asserting that rectifying mispriced computations could enhance usable throughput while acknowledging potential denial-of-service risks and the implications of contracts that embed gas assumptions. BALs are viewed as essential infrastructure for achieving parallelism, citing enhancements such as concurrent disk reads and transaction validation, with EIP-7928 estimating an average overhead of about 70 to 72 KiB for compressed BAL size.
However, the realization of these benefits is contingent upon client adoption of concurrency to address real bottlenecks, which raises concerns about whether additional data and verification processes will inadvertently introduce their own latency costs. The ePBS concept lies at the intersection of discussions on miner extractable value (MEV) and throughput, as it seeks to decouple execution validation from consensus validation temporally, potentially introducing new failure modes.
Research on the “free option problem” related to ePBS indicates option exercise rates of approximately 0.82% of blocks on average, escalating to around 6% during periods of high volatility. As Ethereum plans for 2026, this research underscores the need to focus on network liveness under stress rather than merely steady-state fee conditions. A critical underpinning for raising gas limits is the widespread adoption of ZK-proof technology by validators.
The Ethereum Foundation”s “Realtime Proving” roadmap outlines a phased approach where a select group of validators will initially operate ZK clients in a production environment. Following this, only when a supermajority of stake is comfortable, can gas limits be escalated to levels where proof verification supplants re-execution for practical validation on standard hardware. Constraints highlighted in the foundation”s July 10, 2025 post include achieving 128-bit security (with a temporary acceptance of 100-bit), keeping proof sizes under 300 KiB, and not relying on recursive wrappers with trusted setups.
The implications for scaling are closely linked to the proving market, necessitating a robust and credible supply of real-time proofs without consolidating into a narrow prover group that mimics existing relay dependencies. Following Glamsterdam, the upcoming phase, dubbed Hegota, is set to occur later in 2026 and will focus more on procedural aspects than specific scope. The Ethereum Foundation has released a timeline marking significant decision points, allowing investors and developers to track progress without misinterpreting commitments from code names.
The first deadline is for Hegota headliner proposals, which will close on February 4, providing a structured framework for future developments.
