Ethereum's zkAPI has gone live on mainnet, letting users pay for AI and other metered APIs through zero-knowledge proofs instead of linked accounts.
Ethereum News
The Ethereum Foundation has launched zkAPI, a payment system that lets people pay for AI models and other metered application programming interfaces without linking the payment to their identity. The system, built with the Open Anonymity Project, is now live on Ethereum mainnet.
Vittorio Rivabella of the Ethereum Foundation's dAI team described the problem that zkAPI aims to solve in a post published Thursday. Rivabella wrote that every AI API call today carries an identity, since an API key points to an account, which in turn points to a payment method. He said this setup means every prompt a user sends joins a record tied to both.
Vault Deposits Replace Linked Accounts
Under zkAPI, a user deposits credits in Ethereum (ETH), USD Coin (USDC), or other assets into a vault contract on Ethereum in a single ordinary transaction. From that point, the balance exists as a private note that only the depositor can spend and that cannot be traced back to the original deposit. Ethereum functions as the settlement layer for the system, recording deposits, balance closures, and withdrawals while keeping what those balances paid for hidden from public view.
To authorize spending, software running on the user's own device generates a small zero-knowledge proof showing that a funded note covers the request and that it has not been spent before. The zkAPI server checks the proof and issues a short-lived API key with a capped dollar limit, without learning which note, deposit, or person the request came from.
The system relies on two cryptographic components to keep the process both private and secure. Deposits exist as commitments inside a Merkle tree, allowing a proof to confirm that a note is valid without revealing which one it is. Every spend also publishes what Rivabella called a nullifier, a one-way serial number tied to the note's secret, which exposes any attempt to spend the same balance twice without revealing anything else about the user.
Privacy Has Limits on Network and Content
Prompts travel directly from the user's device to the AI provider using the temporary key, while the zkAPI server separately settles the metered usage against the user's private balance once the key expires. This means the party that handles payment never sees the content of a request, and the party that sees the content never learns who paid for it. The result is a system in which users stay pseudonymous at the payment layer rather than anonymous at every layer, since the AI provider can still see the content of what is being asked even though it cannot see who is paying for it.
Rivabella wrote that zkAPI does not provide network-level anonymity. A provider can still see a request's network metadata, including IP address, and could attempt to correlate sessions through timing patterns. Users seeking stronger protection against this kind of tracking may route traffic through Tor and use a fresh circuit for each session, according to the post.
Related Article: Ethereum To Undergo Major Shift After Hegota Fork, Buterin Says
Content-level privacy carries a separate limitation. Rivabella said an inference provider could still re-link sessions if a user repeats the same personal details, writing style, or project documents across requests, since shared prompt content can act as a fingerprint for anyone reading the prompts.
The release turns a proposal first published in February on the Ethereum Research forum by dAI team lead Davide Crapis and Ethereum co-founder Vitalik Buterin into a working mainnet system. Along with the core protocol, the project has released a local client compatible with OpenAI and Ollama-style APIs, plus a browser-based AI chat implementation called OA Chat, built for use cases including AI inference, blockchain RPC calls, image and video generation, and machine-to-machine payments.
This article contains links to third-party websites or other content for information purposes only ("Third-Party Sites"). The Third-Party Sites are not under the control of CoinMarketCap, and CoinMarketCap is not responsible for the content of any Third-Party Site, including without limitation any link contained in a Third-Party Site, or any changes or updates to a Third-Party Site. CoinMarketCap is providing these links to you only as a convenience, and the inclusion of any link does not imply endorsement, approval or recommendation by CoinMarketCap of the site or any association with its operators. This article is intended to be used and must be used for informational purposes only. It is important to do your own research and analysis before making any material decisions related to any of the products or services described. This article is not intended as, and shall not be construed as, financial advice. The views and opinions expressed in this article are the author's [company's] own and do not necessarily reflect those of CoinMarketCap.
