Ethereum Foundation Launches zkAPI for Private AI Model Payments

The Ethereum Foundation, in collaboration with the Open Anonymity Project, has launched zkAPI, a system designed to let users pay for AI models and other metered APIs without revealing their identities. According to TheBlock, the protocol is live on the Ethereum mainnet.

Zero-Knowledge Privacy Architecture

To use the system, users deposit tokens such as ETH or USDC into a vault contract on Ethereum, which records the balance as a private note. Software on the user's device generates a zero-knowledge proof showing the request is backed by a funded note, without revealing the specific deposit. A zkAPI server verifies the proof and issues a temporary API key with a spending cap. The user then sends prompts directly to the AI model provider alongside the key, with usage deducted from their private balance.

Each payment publishes a serial number known as a nullifier to prevent double-spending. The client exposes standard OpenAI and Ollama APIs locally, allowing existing chat clients and editors to connect through localhost.

Implementation and Future Use Cases

According to TheBlock, the system implements "ZK API Usage Credits," a design published in February by Ethereum co-founder Vitalik Buterin and Ethereum Foundation dAI lead Davide Crapis. The dAI team was formed in September 2025 to enhance Ethereum's capabilities as a settlement and coordination layer for AI.

Initial applications for zkAPI include AI chat and agents, blockchain RPC queries, image and video generation, VPN bandwidth, and machine-to-machine payments. The project's GitHub repository notes that the protocol remains experimental, and the system does not inherently provide network anonymity unless traffic is routed through external tools like Tor.

Key facts

• The Ethereum Foundation launched zkAPI with the Open Anonymity Project on the Ethereum mainnet.
• Users deposit ETH or USDC into an Ethereum vault to fund API usage privately via zero-knowledge proofs.
• The system implements a design published in February by Vitalik Buterin and Davide Crapis.
• Initial use cases cover AI chat and agents, blockchain RPC queries, and machine-to-machine payments.