> For the complete documentation index, see [llms.txt](https://zkagi.gitbook.io/introduction/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://zkagi.gitbook.io/introduction/introduction.md). # Introduction

\ ZkAGI (Zero-Knowledge Artificial General Intelligence) is a framework designed to advance decentralized physical infrastructure networks (DePIN) by promoting privacy-preserving AI technologies. Its primary objective is to incentivize GPU providers, model trainers, AI developers, and data owners to participate in a secure ecosystem. ZkAGI achieves this through the integration of federated learning and zero-knowledge proofs, ensuring that AI model inference and data remain private. ### Key features of ZkAGI include: * Privacy-preserving AI using cryptographic techniques for zero-leakage model inference. * Decentralized AI model training and inference using GPU networks across a global scale. * Zero-Knowledge Machine Learning (ZKML) for verifiable AI execution without revealing sensitive data. * Utilization of Transformer-based Large Language Models (LLMs) for accurate natural language processing.
### Architecture