Vitalik Buterin proposes ‘Extremely Lean’ Ethereum, shrinking the chain to near-zero state with ZK proofs

Vitalik Buterin has outlined a technical proposal to radically shrink Ethereum's consensus layer by shifting much of the state management burden to validators using zero-knowledge proofs, as part of a broader push toward a more efficient "Lean Ethereum."

In a recent Ethereum Research forum post titled "The Extremely Lean Chain," Buterin detailed how recursive STARK proofs could minimize onchain data while supporting upgrades like single-slot finality and quantum-resistant cryptography.

This past weekend, Buterin referred to "Lean Ethereum" as the network's "third major iteration" and said the series of updates could rival the Merge, when Ethereum pulled off the highly technically complex shift to proof-of-stake consensus.

Lean Ethereum is a long-term vision and roadmap first brought up around mid-2025 to overhaul Ethereum's core protocol to make it stronger, more secure, more decentralized, and quantum-resistant. This includes a redesign of Ethereum’s consensus system, called Lean Consensus, improved data handling with post-quantum features, called Lean Data, and a minimal, SNARK-friendly virtual machine, possibly RISC-V-based, called Lean Execution.

Buterin's latest "Extremely Lean" proposal is centered around a redesign of Ethereum's Beacon Chain to drastically minimize onchain state and processing. In the extremely technical research post, Buterin offers a two-step plan that will eventually reduce validator state to roughly 6 bytes, while maximizing privacy.

Phase 1 removes most validator data from the chain and replaces per-epoch balance updates with a single daily ZK-STARK proof submitted by each validator, shrinking onchain state to just a few bytes per validator while letting the chain stay almost stateless on rewards and penalties.

In other words, the proposal shifts more work to validators, who submit proofs, so full nodes and the chain itself can stay lightweight.

Phase 2 adds privacy by giving every validator a fresh anonymous key and identity each day, so they re-register and re-prove their balance privately — turning the validator set into a daily rotating, unlinkable list.

These changes "may allow consensus to scale to millions of validators if needed," Buterin wrote in the post early Monday morning.

According to Buterin, Lean Ethereum’s changes could take three to four years to develop as research begins to focus on the effort. This weekend, he noted that Ethereum's upcoming Hegota upgrade will likely be the network's last thematically "pre-Lean" fork.

The move comes amid recent governance changes in the Ethereum community, including significant downsizing and reorganization at the Ethereum Foundation. Two new nonprofits, backed by Bitmine, Sharplink, and Joseph Lubin, have also been launched, including Ethereum Institutional and EthLabs, the latter of which will contribute to protocol R&D.