Ethereum Foundation's zkAPI Launches Privacy-Preserving API Payments on Ethereum Mainnet
Key Takeaways
- •The Ethereum Foundation and the Open Anonymity Project have deployed zkAPI, a privacy-focused payment system, live on Ethereum mainnet for paying AI and other API services.
- •Users deposit funds into an Ethereum vault and use zero-knowledge proofs to demonstrate they hold enough credit without revealing which deposit is theirs, an application of the same cryptography behind zk-rollups.
- •zkAPI issues short-lived API keys with predefined spending limits, while prompts go directly to AI providers and usage is settled separately through the payment layer, unlike conventional account-based billing.
- •The mainnet implementation follows a February proposal for ZK-based API usage credits from Ethereum Foundation researcher Davide Crapis and co-founder Vitalik Buterin, with a local client, SDK, and browser-based AI chat demo now publicly available.
- •Privacy protection covers only the payment layer, since prompt contents and network metadata remain visible to providers, potentially allowing users to be linked across sessions through IP addresses or timing.
The Ethereum Foundation announced on Thursday that zkAPI, a privacy-focused payment system built in collaboration with the Open Anonymity Project, has gone live on Ethereum mainnet. The system is designed to let users pay for AI and other API services without revealing their billing identities.
How zkAPI works
Users deposit funds into an Ethereum vault and then use zero-knowledge proofs to show they hold sufficient credit to pay for API requests, without revealing which of the deposits belong to them, according to a post from Vittorio Rivabella, AI coordinator at the Ethereum Foundation's dAI team. Zero-knowledge proofs are a well-established cryptographic technique that allows one party to prove a claim is true without the underlying information; on Ethereum, the technology also underpins scaling systems known as zk-rollups. In zkAPI's case, the same technique is applied to payments rather than to transaction scaling.
The design contrasts with conventional API billing, where usage is typically tied to an account backed by a payment method, giving providers a direct association between requests and a payer's identity. One of the system's immediate use cases is paying for AI usage without linking the payment to a user. zkAPI issues short-lived API keys with predefined spending limits. Prompts are sent directly to the AI provider, while usage is settled separately through the payment layer.
Alongside the mainnet deployment, the project has released a local client and a software development kit (SDK), together with a browser-based AI chat implementation.
From proposal to mainnet
The launch follows a February proposal for ZK-based API usage credits by Ethereum Foundation researcher Davide Crapis and Ethereum co-founder Vitalik Buterin. The concept was presented as a way to give users private, prepaid access to AI and other metered APIs. The new release turns that design into a working mainnet implementation. With the client, SDK, and browser demo now publicly available, how developers and API providers take up the tooling is the near-term development to watch.
Privacy limitations
zkAPI does not hide prompt contents or network metadata from providers. As a result, users can still potentially be linked across sessions through IP addresses, timing, or information contained in their requests — meaning the privacy guarantee covers the payment layer rather than the AI interactions themselves.
Source: Cointelegraph
