Ethereum co‑founder Vitalik Buterin announced on X that he now views leanPCS as the most promising route to achieve quantum‑resistant accounts and efficient cryptographic proofs, positioning it as a faster alternative to the previously discussed leanSTARK approach. His comments come as part of ongoing work on the draft EIP‑8288, which aims to keep large verification data off‑chain while preserving security guarantees.

BTC / USD
Source: CoinGecko. Prices are indicative.

Update, Nov 16, 17:40 UTC: According to Unchained, leanPCS only needs the selection of a mathematical field, and leanSPHINCS would cost about 4,096 gas for a 64‑byte signature, while checking a SPHINCS+ variant on‑chain would require roughly 150,000 gas. The report also notes that Starknet is exploring a transition to a layer‑1 to achieve quantum resistance by 2027, with its CEO warning the quantum threat could be much nearer than expected.

Update, Nov 16, 10:01 UTC: According to Decrypt, Buterin said signatures would expand from 65 bytes to between 1 KB and 6 KB and zero‑knowledge proofs could grow to about 128 KB, underscoring the data‑size challenge the proposal aims to solve. He also reiterated his “total bull” stance on leanPCS in a November 16 tweet, describing the approach as quickly implementable, per Decrypt.

What is leanPCS and how it fits into EIP‑8288

leanPCS (lean Polynomial Commitment Scheme) is a proposal to aggregate signatures and STARK proofs before transactions enter Ethereum blocks. The draft, authored by Buterin and Thomas Coratger, suggests that nodes in the mempool combine these large proofs into a single recursive proof, allowing the blockchain to store only short descriptors—roughly 64 to 96 bytes per claim—rather than the full data, which can range from 1 KB to 128 KB per proof.

Comparison with leanSTARK and leanSPHINCS

Buterin outlined three approaches he evaluated: leanSPHINCS, leanSTARK, and leanPCS. leanSPHINCS focuses on hash‑based signatures suitable for quantum‑safe wallets, with a proposed verification cost of 3,000 gas. leanSTARK would enable arbitrary computation proofs but requires agreement on a new instruction‑set architecture, incurring a higher 30,000‑gas cost. leanPCS, by contrast, offers similar generality to leanSTARK but could be implemented more quickly, using existing execution pathways at an estimated 20,000‑40,000 gas per verification.

Roadmap and remaining uncertainties

EIP‑8288 builds on the earlier EIP‑8141 transaction format, which separates validation, payment, and execution into programmable frames slated for the upcoming Hegotá upgrade. However, the leanPCS components are not yet part of the formal specification, and no mainnet activation date has been set. The draft also limits each transaction to 16 signature dependencies and one STARK dependency, with one‑second aggregation intervals to curb potential abuse.