The Ethereum Foundation and the Open Anonymity Project have launched zkAPI on Ethereum Mainnet, a protocol for paying in advance for AI models and other metered APIs without tying a user’s billing identity to individual requests. It uses zero-knowledge proofs to authorise bounded usage, though it does not hide prompt contents or network details such as IP addresses.
Watch Desk analysis
What happened
With zkAPI, users deposit funds into an Ethereum smart-contract vault, then use proofs to create short-lived prepaid API keys. The design separates payment from the API provider, with the aim of preventing the provider from linking a request to the payer or the original deposit.
The project’s launch announcement and technical overview says the protocol uses Groth16 proofs, Baby-JubJub commitments and a 32-level Merkle tree. Its design was co-authored by Vitalik Buterin and Davide Crapis.
Why it matters
AI services that charge by usage can make payment details part of a user’s relationship with a provider. zkAPI proposes a different arrangement: prove that a request is authorised and paid for without giving the provider the billing identity behind it. The same approach is intended for metered APIs beyond AI.
That is a specific privacy boundary, not a cloak of invisibility. The project says zkAPI does not conceal what users put in prompts, or network metadata such as IP addresses. Those details can still matter to anyone deciding whether this setup suits their threat model.
Our read
Separating payment from an API call is a worthwhile idea, especially for users who want less linkage between identity and usage. The interesting test is whether this can be practical to use, not just technically elegant. Treat the launch as a protocol proposal with a useful stated privacy goal, not as proof that every part of an API interaction is anonymous.
What to watch
- Which AI providers and other metered API services support zkAPI.
- How users manage deposits, prepaid keys and any unused credit.
- Whether practical deployments preserve the claimed separation without making ordinary API use cumbersome.
Discussion spark: For paid AI services, is separating billing identity from API requests a meaningful privacy gain if prompts and IP addresses remain exposed?
Sources and evidence
- Introducing zkAPI: private usage credits for any API (1 October 2026, 19:12 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.