Recursive Proof (Proof Aggregation)
A zero-knowledge proof that verifies the correctness of one or more other proofs, letting many computations be compressed into a single, final proof.
A recursive proof is a zero-knowledge proof whose statement is, itself, that one or more other zero-knowledge proofs are valid — rather than proving a computation's correctness directly, it proves that a previous proof (or several previous proofs) correctly proved something. Chaining this recursively lets a very large number of individual computations or proofs ultimately be compressed into a single, compact final proof that attests to the validity of everything that came before it.
Proof aggregation, achieved through recursive proofs, is especially valuable for a zkEVM (see Kaavon's "zkEVM" glossary definition) or zero-knowledge rollup processing many individual transactions: rather than verifying each transaction's proof separately on the base chain, a recursive proof can combine many individual transaction proofs into one aggregated proof, significantly reducing the total on-chain verification cost regardless of how many underlying computations the aggregated proof actually represents.
Examples
- A zk-rollup processing thousands of transactions in a batch can use recursive proofs to aggregate all their individual proofs into a single proof verified once on the base chain.
- Recursive proof aggregation means the on-chain verification cost doesn't scale directly with the number of underlying transactions or computations being proven.