ROME - Swarm Protocol
ROME is the Swarm Protocol of the Rivalz World Abstraction Layer.
Last updated
ROME is the Swarm Protocol of the Rivalz World Abstraction Layer.
Last updated
AI is evolving beyond models—it’s becoming autonomous agents that need resources, coordination, and financial systems to operate at scale. Rome is building the world’s first AI-native economy, where any resource-backed tokenized AI agents (rAgents) that can be staked, borrowed, and participate in swarms.
These rAgents can be staked towards projects to form Swarms. Swarms utilize the staked rAgents, effectively operating as agent collectives. Swarms therefore become operators of mass-agent collectives, backed-by real-world social capital, compute power and more.
ROME also introduces Agentic DeFi creating an economic layer where users can lend, stake, and earn from rAgents.
With ROME, Humans become the backbone of a new ecosystem, where Agents and Swarms grow towards a self-sustaining AI economy. ROME is built as one of the modules of the Rivalz World Abstraction Layer, utilizing its other pieces such as and .
ELI5 Explain: ROME is building the future of decentralized AI economies. Today, resources like compute, data, and social networks exist in silos—unused, fragmented, and difficult to access. ROME turns these resources into tokenized AI agents called rAgents, allowing users to launch, stake, and monetize them within self-organizing AI Swarms. This creates a liquid market for decentralized infrastructure, enabling on-demand intelligence at scale. Think a whole new decentralized economy where rAgent and Swarm tokens are core primitives. By integrating AI, DePIN, and Web3, ROME is laying the foundation for an autonomous, agent-driven economy—one where machines transact, optimize, and evolve without human intervention.
Swarm Tokens as a financial primitive, natively integrated into ROME
Further tokenization of agentic services from existing Agent and DePIN projects
evolution into mini-swarms
More rAgents
Release of MASO [Mass Agent Swarm Orchestration] and RAS [Resource-based Agent System] Frameworks
ROME is currently on its v1 beta stage. If you encounter any issues, feel free to visit our for assistance or . At present, ROME operates exclusively on the Base mainnet. Later this year, we plan to integrate it with Rivalz L2 – the World Abstraction Layer Mainnet.
The launch of ROME v2 is scheduled for Q2 2025.