TechFlow, July 23 — Walrus Protocol, a blockchain storage protocol built on Sui, announced a partnership with Gata, an initiative focused on developing decentralized large model inference, training, and data technologies, becoming a core component of the Gata AI open execution infrastructure.
Leveraging Walrus, Gata will address the high data access costs that have long plagued decentralized AI. This integration directly empowers Gata's DataAgent, enabling anyone to contribute idle computing power with one click via a browser to collaboratively generate synthetic datasets for AI training. Walrus delivers performance comparable to centralized cloud services along with low-cost storage, but more importantly, completely eliminates download fees—allowing the Gata community to conduct large-scale data interactions permissionlessly without bearing unpredictable and unsustainable costs.
In the future, Gata plans to utilize Walrus’s unique on-chain data programmability to build an open and auditable pricing model. By storing verifiable computation proofs on Walrus, Gata can transparently validate computational tasks and settle payments on-chain using Move-based smart contracts, creating a fairer pricing mechanism—a significant improvement over the opaque models used by current centralized AI providers.
Notably, Walrus is a next-generation storage protocol designed for AI Agents, on-chain applications, and enterprises handling large-scale unstructured data. Whether images, videos, audio files, or dynamic metadata, Walrus efficiently, securely, and economically stores them through its distributed node network.
