Author: Zack Abrams
Compiled by Deep潮 TechFlow
DeepChain Summary: The Ethereum Foundation and the OA project have turned the "ZK API Usage Credits" proposed by Vitalik and others in February into a mainnet-ready zkAPI: users first deposit ETH or USDC into a treasury, then use zero-knowledge proofs to obtain temporary keys with usage credits to invoke models, decoupling payment from authentication within the API flow. This adds value to narratives around ETH as a settlement layer and AI agent micropayments; however, gateways may still link requests via IP, and the protocol remains labeled as experimental.
The system was developed with input from the Open Anonymity Project and implements a design publicly released in February by Ethereum co-founder Vitalik Buterin and Ethereum Foundation dAI lead Davide Crapis.
According to a blog post released on Thursday, the Ethereum Foundation launched zkAPI—a system that allows users to pay for AI models and other pay-per-use APIs without revealing their identity.
The foundation stated that zkAPI was developed in collaboration with the open-source AI privacy project, Open Anonymity (OA) Project, and is now live on the Ethereum mainnet.
"Each AI API call today carries an identity," the blog writes. "Your API key points to an account, which in turn points to a payment method, and every prompt you send becomes part of a record tied to both. The service provider can piece together your usage history over the years into a single profile."
“It’s like a generalization of OA’s unlinkable reasoning, while also abstracting away payments,” wrote Ken Liu, a Stanford computer science PhD candidate participating in the Open Anonymity Project, in a post on X Thursday afternoon. He also wrote, “There should be more infrastructure that treats privacy and sovereignty as foundational, putting real control back into people’s hands.”
How does zkAPI work?
According to the blog, when using zkAPI, users must first deposit tokens such as ETH or USDC into a vault contract on Ethereum, which records the user’s balance as a private note. When spending these tokens, software on the user’s device generates a zero-knowledge proof demonstrating that the request is backed by a sufficiently funded note, without revealing which specific note is used or the corresponding deposit details.
After verifying the zero-knowledge proof, the zkAPI server issues a temporary API key with a spending limit. The user then sends the prompt directly to the AI model provider using this key; when the key expires, the server deducts the corresponding usage from the user’s private balance.
Each time a user pays for AI usage, a serial number called a "nullifier" is published. "If someone tries to spend the same balance twice, the duplicate nullifier will expose the attempt, without revealing any other information," the blog stated.
According to the blog, the client can work alongside existing AI tools. "The client exposes standard OpenAI and Ollama APIs on your local machine," the blog states. "Existing applications, editors, and chat clients can simply point to localhost to use it."
Use Cases and Limitations
The blog lists AI chat and agents as the primary use cases. Other explicitly mentioned applications include blockchain RPC queries, image and video generation, VPN bandwidth, and machine-to-machine payments between AI agents.
However, according to the blog, the zkAPI system does not provide network-layer anonymity. The gateway may correlate requests coming from a fixed IP address; if users require stronger privacy, the blog suggests routing traffic through Tor—a decentralized anonymous network that relays traffic through multiple servers and is also commonly used by so-called "dark web" sites.
The session may also be reassociated through prompt content, such as personal details, writing style, or conversation history. The GitHub repository for the zkAPI project describes this protocol as experimental.
Co-designed by Vitalik Buterin
zkAPI implements the "ZK API Usage Credits" design published by Davide Crapis and Ethereum co-founder Vitalik Buterin on the Ethereum Research forum on February 11.
Crapis leads the dAI team at the Ethereum Foundation; the team, established in September 2025, aims to enhance Ethereum’s capabilities as an AI settlement and coordination layer. The team also developed the AI agent identity standard ERC-8004, which went live on mainnet in January.
This blog announcing zkAPI was written by Vittorio Rivabella of the dAI team.
In a February X post, Buterin listed cryptographic payment mechanisms for AI services as one of Ethereum’s near- to medium-term visions regarding AI.