Ethereum Just Taught AI to Take Cash Without Making Eye Contact 🕶️

The Ethereum Foundation announced Thursday that zkAPI, a privacy-preserving system for paying AI and other API services on Ethereum mainnet, is now live. Built with the Open Anonymity Project, the system lets users deposit funds into an Ethereum vault and then use zero-knowledge proofs to pay for API requests without linking the payment to a specific deposit or identity, according to a post from Vittorio Rivabella, AI coordinator at the Foundation's dAI team.

Under the design, users deposit credits in ETH, USDC or other supported tokens into a vault contract in a single on-chain transaction. From that point, balances exist as private notes that only the user can spend. When a request is made, local software generates a zero-knowledge proof demonstrating the user holds sufficient funds and has not already spent them. The server verifies the proof and issues a short-lived API key capped at a predefined spending limit, held only in the device's memory. Prompts are sent directly to the AI provider, while settlement is handled separately through the payment layer. "Prompts are personal. People ask AI models about their health, their finances, their doubts. Under the current model, using AI means handing a running transcript of your thinking to whoever holds the billing relationship," the Ethereum Foundation said. "zkAPI separates payment from identity."

The launch ships alongside a local client, a software development kit, a browser-based AI chat implementation called OA Chat, and open-source code on GitHub. The documentation describes compatibility with OpenRouter, a service that routes requests to hundreds of AI models, and any application that speaks the OpenAI format can connect by pointing at a local address. The Foundation noted that zkAPI does not hide prompt contents or network metadata from providers, meaning users can still potentially be linked across sessions through IP addresses, timing or information contained in their requests.

The release turns a February proposal by Ethereum Foundation researcher Davide Crapis and Ethereum co-founder Vitalik Buterin into a working mainnet implementation. Crapis leads the dAI team, which has positioned Ethereum as a settlement and coordination layer for AI agents. In August 2025, Crapis and Marco De Rossi proposed the ERC-8004 standard, a shared rulebook for autonomous agents to discover, verify and transact with one another, and Crapis has predicted that most Ethereum traffic would come from machines within three to five years. The push to give users private, prepaid access to metered APIs traces back to a February 10 post in which Buterin laid out an Ethereum-AI roadmap that included zero-knowledge payments for private API usage.