CryptoDetector
Menu

Vitalik Buterin on leanPCS and a proposed path to lower-cost proof verification

For people holding assets on Ethereum, a useful question about any cryptographic proposal is what the chain would verify—and what it would no longer need to store. In a November 16, 2026 report, Vitalik Buterin said he had recently shifted from skepticism toward leanPCS and explained how it could fit with EIP-8288. These are proposed approaches, not changes described as already in effect.

Under EIP-8288 as Buterin described it, STARK proofs would be aggregated in the mempool. Data that needs to be computed only once—including signatures, STARKs, Merkle branches and FHE—could then scale without putting every full signature or proof on-chain. The chain would receive a statement hash and a main proof instead.

Buterin distinguished leanSPHINCS, a specific quantum-safe signature scheme; leanSTARK, a general-purpose proof system that needs an agreed instruction set; and leanPCS, a polynomial commitment scheme that needs only a specified field, such as a 128-bit binary field. He argued that leanPCS could offer a quicker route to similar generality, with STARK verification costing roughly 20,000–40,000 gas from the EVM’s perspective. He presented it as a possible faster route toward what he called a “cryptographic proof singularity,” with leanSTARK to follow later. The report does not establish that those costs or capabilities have been achieved in production.