zero-knowledge proofs.
Prove a computation without revealing its inputs.
A zero-knowledge proof is a short string that proves "this computation was done correctly" without revealing what went in or what came out. Void mints one for every job.
schemezk-snark
provesoutput = model(input)
revealsnothing else
verifierpublic · on chain
trust, but verify
You don't trust the enclave that ran your job — you verify the proof. If a node cheated, the proof won't clear, and the network's stake gets slashed. Verification is what makes "don't trust us" actually work.