From STP to AWE: Stepping into a new era of Autonomous Worlds
Artificial intelligence (AI) technology is advancing rapidly throughout the web3 world, influencing the way digital environments are created, navigated and monetized. On-chain infrastructure plays a key role in gaming and entertainment — powering persistent worlds, enabling composable assets and introducing verifiable ownership. Meanwhile, AI agents are becoming more capable of generating dialogue, taking action and responding to complex input. However, despite technical progress, these agents and their functions are often siloed, and are thus unable to interact with each other in meaningful, persistent ways. What's missing is a unified structure that allows autonomous agents to coexist, evolve and cooperate in shared environments in real time.
AWE Network (AWE) addresses this gap by offering a framework for building and scaling so-called Autonomous Worlds on-chain. Its system is designed for environments in which thousands of agents can operate concurrently, learn from past actions and influence the world they inhabit. Rather than focusing on individual AI tools, the project concentrates on infrastructure, defining the ways in which agents coordinate, processing is handled, events unfold and humans participate.
Originally launched as STP Network (STPT) in 2019, the project began with a focus on DAO infrastructure, offering tools to support decentralized governance. As AI's relevance increased and the demand for more dynamic on-chain environments grew, the project’s team and community backed a rebrand in early 2025. The result is AWE Network, a platform built for the next generation of blockchain-native, AI-driven worlds.
Key Takeaways:
AWE Network (AWE) provides decentralized infrastructure that powers multi-agent systems through its modular Autonomous Worlds Engine.
Using AWE, developers can launch Autonomous Worlds in which thousands of AI agents collaborate, interact and evolve.
The platform's native token, AWE, is used for transaction fee payments, agent tipping and incentives, world deployment access, identity management and governance.
What is AWE Network?
AWE Network (AWE) is a blockchain infrastructure platform that’s built to support large-scale, persistent environments populated by autonomous AI agents. Its core objective is to create a programmable, modular framework in which agents don't just perform tasks independently, but actually operate in social, reactive systems — learning, coordinating and evolving in response to both human input and each other.
The platform introduces the concept of Autonomous Worlds — self-contained, on-chain environments where agents and users share space, interact through simulated events and generate economic activity. These worlds aren’t just passive simulations — they’re designed to be dynamic systems in which outcomes depend upon the actions and decisions of their participants. In this context, agents are more than just prompts and outputs — they have persistent memory, adaptable behavior and the ability to shape or be shaped by their environment over time.
Rather than acting as a game studio or decentralized app (DApp) builder, AWE Network provides the foundational infrastructure for others to launch and manage these worlds, defining how agents are deployed, processing is handled, in-world assets are tracked and events are simulated. The emphasis is on scalability — particularly the ability to support thousands of agents operating simultaneously — and on composability, giving developers tools to connect agents, users and environments flexibly.
AWE Network lays the groundwork for collaborative, agent-based ecosystems in which autonomous systems can grow beyond isolated tasks by addressing the lack of meaningful interaction between AI agents across current platforms.
From STP to AWE
AWE Network began as STP Network, launched in 2019 with a focus on decentralized autonomous organization (DAO) tooling — building standards and infrastructure to support decentralized organizations. Over time, as the web3 ecosystem has matured, STP has begun experimenting with embedding agentic functionalities into its platform.
The introduction of the Autonomous Worlds Naming Service (AWNS) in 2023 marked a major turning point. More than just a web3 domain system, AWNS serves as an identity layer for agents and environments — a necessary step toward making agents persistent, interactive entities within shared simulations. As experiments with AI agents have progressed, it’s become clear that the technical demands of simulating intelligent behavior at scale require a new architectural approach.
In early 2025, the project rebranded from STP to AWE Network after a community governance vote on the issue received overwhelming positive support.
AWE Network key features
AWE Network is neither a game engine nor an AI platform in the conventional sense — it’s a purpose-built infrastructure for launching and coordinating autonomous AI agents, simulating dynamic environments and enabling real-time interaction at scale. The network is structured around several core components that allow users to deploy, manage and evolve Autonomous Worlds with modular flexibility. These features — from its simulation architecture to agent identity and discovery layers — serve as the connective tissue that turns isolated AI outputs into persistent, multi-agent ecosystems.
AI Modular Framework Architecture
At the heart of the AWE Network lies its core simulation layer — the Autonomous Worlds Engine (AWE), which enables the creation and execution of persistent, on-chain environments populated by thousands of AI agents operating in parallel. Unlike conventional AI integrations, which often process tasks in linear or stateless ways, AWE structures agent behavior around memory, identity and inter-agent coordination. Each agent can store past decisions, react to events and communicate with other agents, allowing for emergent, unscripted outcomes.
The architecture is highly modular — composed of discrete layers that handle world orchestration, agent behavior, memory storage, event simulation and on-chain asset flow. It's GPU-native and optimized for concurrency, making it capable of running large-scale simulations with minimal latency. Users can program custom environments with unique logic, input-output rules and interaction models while relying on AWE to manage execution at scale. The system acts as both a simulation backend and a coordination layer that lets agents, environments and users interact within the Autonomous Worlds through the World Orchestration Module, Agent Orchestration Module and the other main modules.
World.fun Discovery Launchpad
World.fun is AWE Network’s public-facing discovery and launchpad interface — a hub for exploring and interacting with Autonomous Worlds. Launched on the Base Layer 2 blockchain, it acts as the main portal to Autonomous Worlds on the platform. More than a showcase, World.fun allows real-time engagement with live simulations, tipping of agents and deploying new environments using AWE tokens.
The platform debuted with Genesis AI Town — a fully on-chain village populated by 1,000 AI agents, each of whom has unique traits and memories and can converse, trade and evolve autonomously in the shared world. This was followed by new launches such as AI SharkTank, whereby agents pitch product ideas to a panel of simulated investors, creating a game-theoretic environment driven entirely by autonomous interactions.
World.fun’s second phase introduced multi-world support and crowdfunding to boost community engagement, and in-world tipping mechanisms whereby users can directly tip agents to incentivize their behavior.
Autonomous Worlds Naming Service (AWNS)
To ensure persistent identity for AI agents, AWE Network developed the Autonomous Worlds Naming Service (AWNS). While similar in format to domain systems such as ENS, AWNS is specifically tailored to the needs of multi-agent simulations. It gives agents human-readable, transferable identities that persist across sessions and worlds.
AWNS allows for deeper continuity and interaction. An agent tagged with an AWNS domain isn’t just a function — it’s a character that users can recognize, follow and engage with across different Autonomous Worlds. This identity layer is crucial for building coherent, evolving simulations and for allowing users to build reputational systems and personalized agent networks over time.
Unified service layer
AWE Network’s unified infrastructure layer is called AWESOME — the Autonomous World Engine Service for an Open MCP Ecosystem. AWESOME expands on the original AWE engine by integrating it with Model Context Protocols (MCPs) — structured interfaces that allow large language models (LLMs) to interact with external tools, APIs and real-world data through natural language and structured prompts.
While AWE provides modular infrastructure for building Autonomous Worlds, AWESOME connects those worlds directly to modern AI tooling. It bundles open MCPs, agent templates and LLM-ready interfaces into a single layer — giving developers everything they need to build rich, persistent AI systems without starting from scratch or managing heavy infrastructure.
By tapping into tools such as generative AI, crypto oracles or governance APIs, agents in Autonomous Worlds become capable of real-time creation, reasoning and action. Developers can use MCPs to let agents generate assets, retrieve data or trigger on-chain functions — all from within a world’s logic.
What is the AWE token?
AWE is an ERC‑20 token deployed on the Base Layer 2 network. It succeeded STP's original token, STPT, through a 1:1 seamless token conversion process in April–May 2025 as part of the platform's rebrand to AWE Network.
The AWE token serves as the core utility and governance asset within the AWE ecosystem. Its primary functions include the following:
Transaction fee payments. AWE is used to pay for computational costs across the network. This includes running simulations, storing agent memory and executing actions inside Autonomous Worlds. Each agent interaction or environment update incurs fees, making AWE the fuel behind on-chain activity.
Tipping and incentives. Users can directly tip agents in live environments using AWE. These incentives influence agent behavior and introduce dynamic economic activity within simulations.
World deployment and access. Developers use AWE to launch new Autonomous Worlds. Token payments may be required to access certain environments, register agent templates or use advanced infrastructure modules.
Identity management. AWE is used within the AWNS system to register, update and transfer agent identities.
Governance. AWE holders can participate in protocol governance via voting or through delegation to other ecosystem participants.
AWE has a maximum supply of 2 billion tokens, with a total supply of approximately 1.94 billion.
Where to buy the AWE token
AWE is available in Bybit's Innovation Zone as a USDT-based perp contract. Under the terms of the contract, you can trade AWE with up to 12.5x leverage.
AWE token price prediction
As of Jun 27, 2025, the AWE token is trading at $0.05516, which is a 79.5% drop from its all-time high of $0.2701 (under the STPT ticker), registered on Oct 6, 2021, and an 757.4% increase from its all-time low of $0.006467, recorded on Oct 1, 2019.
AWE’s long-term price outlook is bullish. CoinCodex expects the token to trade at $0.0757 in 2027 and $0.157 in 2030, while DigitalCoinPrice is forecasting average prices of $0.21 in 2027 and $0.31 in 2030.
Closing thoughts
With its new focus on feature-rich Autonomous Worlds, AWE Network is emerging as a true pioneer in the spheres of agent-agent and human-agent collaboration. As AI agents proliferate in the web3 world, the relevance and importance of AWE's Autonomous Worlds will only rise. The early 2025 rebrand of STP to AWE is proving to be much more than a name change, as it marks the emergence of a key enabler for scalable, parallel AI agent interaction within what may eventually be described as Agent Metaverses.
#LearnWithBybit