snark

A SNARK, or succinct non-interactive argument of knowledge, is a type of cryptographic proof system. In a SNARK, a prover can convince a verifier that a stated computation was performed correctly for a given input, with a proof that is short and can be checked quickly. The term “non-interactive” means that after an initial setup, there is no need for back-and-forth communication between prover and verifier; the prover sends a single proof and the verifier performs a quick verification. “Succinct” refers to the small size of the proof and the fast verification, both of which do not grow with the complexity of the underlying computation.

Most SNARK constructions rely on a one-time or reusable trusted setup to generate a common reference string used by both parties. If the setup is compromised, the soundness of the system can be undermined. A SNARK can also be designed to be zero-knowledge (zk-SNARK), meaning the proof reveals nothing about the witness beyond the truth of the statement. However, not all SNARKs are zero-knowledge by default.

Notable SNARK families include Groth16 and PLONK. Groth16 offers very short proofs but typically requires a circuit-specific trusted setup. PLONK provides a more versatile approach with a universal, updatable setup, allowing a single setup to support many circuits. In practice, SNARKs are used to enable verifiable computation in environments such as blockchain and privacy-preserving protocols, where one party can prove that a computation was performed correctly without re-executing it.

An alternative class, zk-STARKs, aims to remove trusted setups and provide post-quantum security, though with generally larger proofs and verification times. SNARKs remain a core area in modern cryptography for scalable, private verification of complex computations.