TLDR
- The Ethereum Foundation partnered with the Open Anonymity Project to deploy zkAPI on mainnet.
- The protocol enables private payments for AI services and metered APIs by separating payment from identity.
- Deposits in ETH or USDC are verified through zero-knowledge cryptographic proofs.
- Short-lived API keys provide service access while maintaining deposit anonymity.
- Compatible services include AI conversations, autonomous agents, blockchain RPC nodes, content creation tools, VPN services, and automated machine transactions.
The Ethereum Foundation has deployed zkAPI, a protocol designed for anonymous payments when accessing artificial intelligence services and usage-based platforms. The system operates on Ethereum’s main network following collaboration with the Open Anonymity Project.
zkAPI decouples financial transactions from service usage patterns. This architecture allows individuals to purchase API access without creating persistent account identifiers that track their requests. The deployment builds on Ethereum’s ongoing development in privacy technology, cryptographic verification methods, and AI integration.
Participants fund a vault smart contract on Ethereum using ETH or USDC. The contract stores these balances as encrypted notes. Client software running locally generates a zero-knowledge proof whenever service payment becomes necessary.
Temporary Access Keys Replace Traditional Account Systems
The zkAPI server validates submitted proofs before generating time-limited API credentials with defined spending thresholds. These proofs demonstrate sufficient funds without revealing which specific deposit backs the transaction. Ethereum’s development of cryptographic proof systems suggests broader adoption of these verification methods across the ecosystem.
Users transmit requests directly to service providers using their temporary credentials. Once credentials expire, zkAPI calculates usage costs and reduces the encrypted balance accordingly. The system publishes a nullifier to prevent duplicate spending attempts from identical deposits.
The client software integrates with OpenAI and Ollama API standards on local machines. Applications and development environments can establish connections via localhost, eliminating the need for custom interfaces.
This release arrives as Ethereum network activity draws continued attention. Recent metrics indicated major stakeholders increasing their ETH positions while exchange reserves declined, offering broader context for emerging network utilities.
Ethereum Foundation Outlines Current Privacy Boundaries
zkAPI does not conceal network-level identifiers. Gateway services can potentially correlate multiple requests originating from identical IP addresses. Participants requiring network anonymity should route connections through Tor infrastructure.
Request content itself may compromise anonymity. Specific information, distinctive language patterns, or accumulated conversation data could reconnect supposedly independent sessions. The project documentation characterizes zkAPI as experimental technology subject to modification through ongoing evaluation.
The protocol implements concepts from a February design paper authored by Davide Crapis and Vitalik Buterin. Its arrival coincides with recent Ethereum price activity, while Buterin has explored applications combining cryptographic proofs with offchain computation.
The Ethereum Foundation identifies several target applications: AI conversation interfaces, autonomous AI agents, blockchain node RPC services, media generation platforms, VPN bandwidth allocation, and automated machine transactions. zkAPI provides mainnet infrastructure supporting these privacy-enabled use cases.





