Ethereum 2026 Roadmap: Pectra, Fusaka, and Beyond

The Ethereum Development Roadmap Structure

Ethereum's development follows a named upgrade schedule coordinated between the execution layer (formerly Eth1) and consensus layer (formerly Eth2). Since the Merge, each hard fork combines execution-layer EIPs (Prague releases) with consensus-layer EIPs (Electra/Fusaka releases).

The roadmap as of early 2026 has three active tracks: the Pectra upgrade (live on mainnet as of Q1 2026), the Fusaka upgrade (expected Q4 2026), and the longer-term Glamsterdam release (2027+).

Vitalik Buterin's roadmap pillars provide the strategic framing: The Merge (complete), The Surge (L2 scaling — ongoing), The Scourge (MEV and censorship resistance — ongoing), The Verge (Verkle trees and statelessness — in progress), The Purge (historical data reduction — in progress), and The Splurge (everything else — ongoing).

Each upgrade advances multiple pillars simultaneously. Understanding which EIPs belong to which pillar helps developers predict the long-term trajectory of the platform they are building on.

Pectra: What Is Live Now

Pectra activated on Ethereum mainnet in early 2026, combining the Prague execution layer upgrade with the Electra consensus layer upgrade.

Execution layer highlights:
EIP-7702: EOA account abstraction. Allows EOAs to delegate execution to contract code for a single transaction. This is the foundational change for the next generation of wallets.
EIP-2537: BLS12-381 precompile. Reduces ZK proof verification costs by 5-80x for BLS curve operations.
EIP-7685: General purpose execution layer requests. A new mechanism for consensus-layer-triggered execution layer operations, enabling more direct staking contract interactions.

Consensus layer highlights:
EIP-7251: Increased max effective balance to 2048 ETH. Allows validators to consolidate, reducing total validator count.
EIP-7002: Execution layer triggerable withdrawals. Staking contracts can now trigger validator withdrawals autonomously without requiring validator key access.
EIP-7549: Reduced attestation data size for lighter consensus overhead.

EIP-7002 is particularly significant for liquid staking protocols — Lido, Rocket Pool, and EigenLayer can now build withdrawal flows that do not require trusting validator key holders.

Fusaka: What Is Coming Next

Fusaka (Frankfurt + Osaka) is Ethereum's next planned hard fork, targeting Q4 2026. The EIP set is still being finalized, but several proposals are near-certain inclusions.

EIP-7594 — PeerDAS (Peer Data Availability Sampling): upgrades how Ethereum nodes sample blob data. Instead of each node downloading all blobs, nodes sample small chunks and use erasure coding to reconstruct data. This increases the blob target from ~3-6 to potentially 64+ blobs per block, proportionally reducing L2 transaction costs again.

EIP-7692 — EOF (EVM Object Format) superset: a major restructuring of EVM bytecode format. EOF separates code from data, enables static analysis, and lays groundwork for more efficient JIT compilation. Contracts compiled to EOF bytecode gain intrinsic gas savings and improved security properties.

EIP-4444 — History expiry: full Ethereum nodes stop being required to store pre-Merge historical data. This reduces the storage cost of running a node by hundreds of gigabytes, supporting better decentralization.

EIP-7623 — Increase calldata cost: discourages using calldata as a cheaper blob alternative by increasing non-zero calldata gas costs. Pushes L2s fully toward blob data.

The Verge: Verkle Trees and Statelessness

One of Ethereum's long-term scalability challenges is state size. The current state (all account balances, contract storage, and code) exceeds 1 terabyte when including history. Running a full node requires significant hardware, which limits decentralization.

Verkle trees replace Ethereum's current Merkle Patricia trie with a more compact vector commitment scheme. The key property: Verkle witnesses (proofs of state inclusion) are orders of magnitude smaller than Merkle witnesses. This makes stateless clients viable.

A stateless client does not store the full state. Instead, it verifies transactions using witnesses provided alongside the transaction. Block producers provide witnesses; validators and light clients verify without storing state.

The developer impact is modest but real: Verkle tree migration changes how storage proofs work. Any contract that generates or verifies Merkle state proofs on-chain will need to be updated. ZK proof systems that prove Ethereum state inclusion will need to support Verkle witnesses.

Verkle tree migration is one of the most technically complex changes in Ethereum's history. Expect it to land in 2027 or later.

What Developers Should Prioritize in 2026

Given the Ethereum roadmap, here is a priority list for developers in 2026:

Priority 1 — EIP-7702 integration. This is live now. Update your wallet integrations to support type-4 transactions. Build gas sponsorship flows using EIP-7702 delegation. Audit any tx.origin usage in your contracts.

Priority 2 — EIP-7002 withdrawal integration. If you interact with staking protocols, the new withdrawal trigger mechanism changes how protocol-initiated withdrawals work. Update integrations accordingly.

Priority 3 — PeerDAS preparation. When Fusaka ships, blob capacity will expand significantly. L2s will reduce fees further. Design your protocol to benefit from L2 cost reductions rather than being locked to L1.

Priority 4 — EOF readiness. EOF is expected in Fusaka. While existing contracts are unaffected, new contracts compiled to EOF will have gas advantages. Begin evaluating compiler support in Solidity and Vyper.

Priority 5 — History expiry planning. If your application depends on querying historical Ethereum state from nodes (not indexers), prepare for nodes to stop serving pre-Merge data. Migrate historical queries to dedicated archive nodes or indexing services like The Graph or Goldsky.