Ethereum Foundation Rolls Out zkAPI for Private AI Inferences

The Ethereum Foundation has introduced a new tool called zkAPI, which enables private AI inferences using zero-knowledge proofs. The protocol allows users to decouple financial accounting from their identity during the consumption of commercial large language models. With zkAPI, users can deposit tokens such as USDC and ETH into a smart contract on the Ethereum mainnet, and then generate zero-knowledge proofs to certify available funds without exposing the funding origin or remaining account balance.

The system issues ephemeral keys with local balance caps, ensuring that payment servers never see user prompts and inference providers never learn the payer's identity. The protocol anchors its security in the Groth16 scheme over the BN254 curve, utilizing Poseidon hashes to verify commitments and nullifiers. This design prevents the double-spending of compute credits without compromising user anonymity on-chain.

The zkAPI tool is now live on the Ethereum mainnet under the ZkApiVault contract, supporting deposits in tokens such as USDC and ETH. The protocol's underlying framework builds on research into ZK credits published by Vitalik Buterin and Davide Crapis.

The Ethereum Foundation clarified that the protocol does not resolve network-level anonymization on its own, and users should integrate complementary mixnets such as Tor or dedicated VPNs to further protect their anonymity.