The Ethereum Foundation announced on Thursday that zkAPI has officially gone live on the Ethereum mainnet. Developed in collaboration with the Open Anonymity Project, the newly launched system transforms an earlier zero-knowledge API payment proposal into a fully operational implementation designed to grant users private, prepaid access to artificial intelligence and other metered digital services.
According to a public announcement shared by Vittorio Rivabella, an artificial intelligence coordinator within the Ethereum Foundation’s dAI team, the system allows users to deposit funds directly into an Ethereum-based vault. Once the funds are deposited, users can utilize zero-knowledge proofs to mathematically demonstrate that they possess sufficient credit to pay for subsequent API requests without revealing which specific deposits belong to them, thereby decoupling their billing identities from their ongoing service consumption.
Bringing ZK-Based API Credits From Theory to Mainnet Reality
The deployment of zkAPI on the mainnet marks a significant milestone for decentralized privacy infrastructure, following a conceptual proposal introduced in February by Ethereum Foundation researcher Davide Crapis and Ethereum co-founder Vitalik Buterin. That initial proposal outlined a theoretical framework for utilizing zero-knowledge proofs to manage API usage credits privately. The new mainnet release successfully translates that design into working software capable of handling real-world transactions and service interactions.
One of the primary, immediate use cases championed by the development team is paying for AI utilities without establishing a direct link between the payment source and the individual utilizing the service. To achieve this, zkAPI issues short-lived, ephemeral API keys equipped with predefined spending limits. Users can then route their prompts directly to participating AI providers, while the underlying financial usage is settled separately and anonymously through the decentralized payment layer. Accompanying the mainnet rollout, the project has also released a local client and a comprehensive software development kit, alongside a browser-based AI chat implementation designed to demonstrate the system’s practical capabilities.
Understanding the Privacy Boundaries of zkAPI
While zkAPI introduces a robust cryptographic layer for financial anonymity, the development team and project documentation emphasize specific operational boundaries regarding what the system protects. Although the protocol effectively severs the link between a user’s wallet identity and their API spending habits, zkAPI does not inherently mask the contents of prompts or underlying network metadata from the API providers themselves.
Consequently, users interacting with services through zkAPI must still remain mindful of external privacy vectors. Because the system does not automatically obscure Internet Protocol addresses, connection timing, or identifying patterns contained directly within the text of the requests, users could potentially be linked across disparate sessions by service providers if proper auxiliary precautions—such as virtual private networks or alternative routing methods—are not simultaneously employed.
The launch of zkAPI on the Ethereum mainnet adds a new dimension to the ongoing discussion surrounding decentralized artificial intelligence and user privacy, offering developers and consumers a practical tool to navigate metered digital ecosystems with enhanced financial confidentiality.






