The Ethereum Foundation announced on Thursday that zkAPI, a system built with the Open Anonymity Project to let users pay for AI and other API services without revealing their billing identities, has gone live on the Ethereum mainnet. The deployment marks a significant milestone in bringing privacy-preserving financial tools to everyday web services, particularly as decentralized applications and artificial intelligence integrations continue to converge.
Users deposit funds into an Ethereum vault, then use zero-knowledge proofs to prove they have enough credit to pay for API requests without revealing which deposits are theirs, according to a post from Vittorio Rivabella, AI coordinator at the Ethereum Foundation’s decentralized AI (dAI) team. This cryptographic mechanism ensures that while the service provider can verify that a valid payment has been made, they cannot link the transaction back to the user’s original deposit address or broader financial profile.

One of the immediate and most prominent use cases of the system is paying for the use of artificial intelligence without linking the payment to a user’s personal identity or financial history. zkAPI issues short-lived API keys with predefined spending limits to facilitate this process. Prompts are sent directly to the AI provider for processing, while usage is settled separately and anonymously through the decentralized payment layer. To help developers and early adopters test the system, the project has also released a local client and software development kit, alongside a browser-based AI chat implementation that showcases how the architecture functions in a real-world environment.
The mainnet launch follows a February proposal by Ethereum Foundation researcher Davide Crapis and Ethereum co-founder Vitalik Buterin for zero-knowledge-based API usage credits. The newly released system successfully translates that initial theoretical design into a working mainnet implementation, demonstrating the rapid pace at which cryptographic research is moving from whitepapers to operational smart contracts and infrastructure.
The concept was originally presented to the community as a way to give users private, prepaid access to AI and other metered APIs without sacrificing the convenience of modern web services. However, technical observers note that zkAPI comes with specific threat models and privacy boundaries that users must understand. Specifically, zkAPI does not hide prompt contents or network metadata from the API providers themselves. This means that users can still potentially be linked across different sessions through persistent identifiers such as IP addresses, connection timing, or unique stylistic and contextual information contained directly within their requests.

As the Ethereum ecosystem continues to expand its footprint in decentralized AI and privacy infrastructure, the deployment of zkAPI provides a foundational layer for developers seeking to build more confidential applications. While challenges regarding network-level metadata and prompt privacy remain areas for future cryptographic and architectural research, the mainnet debut represents a concrete step forward in decoupling digital payments from personal surveillance on the decentralized web.






