With this integration, Walrus becomes the default memory layer for elizaOS V2, allowing AI agents to store, manage, and recall data in a secure and persistent way. This gives AI systems something they have long lacked: reliable memory that lasts beyond a single task or session.
elizaOS is an open-source platform designed to build and manage AI agents. By integrating Walrus, it now offers a trusted infrastructure for agent memory, shared datasets, and multi-agent workflows. New developers can access memory features from the start through a simple onboarding process, making adoption fast and straightforward.
Walrus addresses a major challenge in AI development by providing a unified data layer for models, datasets, and media. Every piece of data uploaded to Walrus is backed by a certificate recorded on the Sui blockchain. This makes stored data verifiable, traceable, and tamper-resistant. Developers can confidently store, retrieve, and share data in a decentralized way, making Walrus a strong foundation for agent memory.
Rebecca Simmonds, Head of the Walrus Foundation, explains that one of the biggest challenges for AI agents is building memory that is both long-lasting and trustworthy. Through this collaboration with elizaOS, developers can safely store, retrieve, and share data across intelligent agents while keeping everything auditable and secure.
The integration delivers several important benefits. Walrus is now the default data platform for elizaOS V2, giving developers full access to memory features from day one. The launch is supported by real-world case studies, developer feedback, and working project examples. Clear documentation, sample code, and production-ready templates make it easy to get started. The partnership also demonstrates how Walrus can operate across multiple blockchains, not just within the Sui ecosystem.
Shaw Walters, founder of elizaOS, says Walrus provides the scalable and secure memory layer needed for multi-agent systems. With data management handled by Walrus, developers can focus on building advanced AI workflows without dealing with fragmented or unreliable storage systems. This integration highlights the real potential of decentralized, verifiable memory in practical AI applications.
The collaboration also includes support for Seal and Nautilus, which enable encrypted access control and private computation. Together with Walrus, these tools allow developers to define how AI agents access, process, and even monetize data, all within a secure and transparent framework.
As elizaOS V2 agents and products roll out later this year, Walrus will serve as a universal data layer for AI agents. By combining storage, access control, and programmability, it provides a trusted foundation for the next generation of decentralized intelligence.
Sui Network, the blockchain behind Walrus, is a layer-one platform built to support scalable and efficient web3 applications. Using the Move smart contract language, Sui enables developers to build fast, low-cost applications designed for mass adoption.