Oasis Protocol Launches ROFL Mainnet for Private AI-Powered Web3 Apps

In Summary

  • Oasis Protocol debuts ROFL mainnet for private offchain AI compute
  •  Platform enables secure AI workloads with blockchain verification
  •  Zeph and WT3 among early adopters of ROFL tech
  •  ROFL dubbed “Trustless AWS” for Web3 and AI apps


Catenaa, Wednesday, July 09, 2025-The Oasis Protocol Foundation has launched the ROFL mainnet, a privacy-first offchain computing platform designed to power next-generation AI and blockchain applications.

Dubbed the “Trustless AWS,” ROFL allows developers to process high-volume computations such as AI model training, data analysis, and autonomous agents without compromising decentralization or data privacy.

ROFL, which stands for Runtime Offchain Logic, integrates with Oasis’ Trusted Execution Environment (TEE) Cloud, enabling cryptographically verifiable outputs from offchain processes to be securely linked back to onchain smart contracts.

This architecture addresses two key industry limitations, blockchain’s constrained compute capacity and AI’s need for verifiable prvacy.

Early adopters include Zeph, an AI companion app that uses ROFL to maintain secure data environments amid rising concerns over mental health app vulnerabilities.

Another project, WT3, applies the framework to decentralized finance, using it for trustless trading automation and key management, backed by a $100,000 grant from the foundation.

Oasis co-founder Jernej Kos said ROFL offers developers plug-and-play tools to build consumer and financial AI applications with privacy built into their infrastructure. The platform also supports future use cases like AI chatbots, price oracles, and game hosting, all handled offchain but trustlessly verified.

With this release, Oasis positions itself as a frontrunner in merging blockchain integrity with scalable AI solutions, aiming to reshape how decentralized applications handle compute-intensive tasks in an increasingly privacy-focused tech ecosystem.

Protected by Copyscape