Proving costs fell 40%, and the shortlist changed

Hardware acceleration and better circuit design cut the cost of generating a proof sharply over the past year. The reaction has mostly been to ask what becomes possible. The more useful question is what was already possible and simply cost too much.

Those applications share a shape: a verifier who cannot be trusted with the input, a computation that is expensive to repeat, and a result that must be checkable by a third party later.

What crossed the line

• Proof of correct inference — showing a model produced an output without revealing weights or prompt.
• Compliance attestation over private ledgers, where the auditor learns the conclusion and nothing else.
• Cross-chain state verification without a trusted committee in the middle.
• Identity claims that reveal a predicate rather than a document.

Where the cost still bites

Verification is cheap and proving is not. Any design that needs a proof per user action, rather than per batch, remains uneconomic outside narrow cases. Recursion helps, at the price of a considerably harder system to reason about.

We stopped asking how fast proving is and started asking how often we actually need to prove.

What to watch

Trusted setup is receding as a practical concern, which removes the objection procurement teams understood best. Expect the conversation to move to auditability of the circuit itself.