Originally reported by Jinse Finance on November 16, 2026, at 15:47 (UTC+8). The report describes Vitalik Buterin’s discussion of EIP-8288, a proposal aimed at reducing the on-chain storage needed for large cryptographic proofs. It is a proposed approach, not a change described as already deployed.
The idea is to aggregate work that needs to be checked once, such as signatures, STARK proofs, Merkle branches and fully homomorphic encryption computations. Under the mechanism described in the report, users would publish the original signatures or proofs, nodes would aggregate them in the transaction pool, and the final block would retain input and output hashes plus one aggregate STARK proof rather than the original large proofs.
Buterin compared three possible routes. leanSPHINCS targets hash-based post-quantum signatures for accounts and multisignatures, but supports a specific signature algorithm. leanSTARK could accommodate a wider range of uses, including privacy protocols, but would require agreement on an instruction set outside the EVM framework. leanPCS, a polynomial commitment approach, would require choosing finite-field parameters. Buterin favors it as a potentially faster route with more flexibility for off-chain proof systems, though the report says its on-chain cost would be somewhat higher. He suggested leanSTARK could be added later.
Buterin’s view that EIP-8288 and leanPCS could hasten a “cryptographic proof singularity” is a projection, not a reported deployment outcome.