Virtuals Protocol is a decentralized platform built on the Ethereum layer-2 Base, designed as a launchpad where users can create their own AI agents. These agents come equipped with unique abilities, can be monetized, and even operate on different applications and platforms.
What has made Virtuals Protocol appealing so far is the ability to create a self-sustaining economy, mixing blockchain and AI in a much simpler way and thus cementing a unique product-market fit. Its user-friendliness is another plus, as people must only connect their crypto wallets and follow a few steps to create and launch their AI agents into the market.
This article takes an in-depth dive into what Virtuals Protocol is and how to create an AI agent following just a few steps here and there.
Quick Navigation
- The Rise of AI Agents
- How Are Crypto AI Agents Different from Traditional Bots?
- Virtuals Protocol: An Overview
- Virtuals Protocol Composer Stack
- The First AI Agent: LUNA
- Other Virtuals Protocol AI Agents
- VIRTUAL Tokenomics
- How to Create an AI Agent on Virtuals Protocol?
- Frequently Asked Questions
The Rise of AI Agents
Before going in-depth into Virtuals Protocol, it’s important to understand AI agents’ role in crypto and general. Put simply, AI agents are programs capable of perceiving their environment and thus making decisions to accomplish objectives.
In the crypto market, AI agents are tailored to interact with blockchain protocols and decentralized applications (dApps), becoming one of the hottest trends in the industry in late 2024 and early 2025.
Two popular examples of AI Agents used in crypto are:
- AIXBT: This AI agent monitors cryptocurrency discussions on social media and provides real-time insights to users.
- Truth Terminal: An AI chatbot designed to autonomously post and interact on social platforms, enhancing engagement and information dissemination.
- AI16Z: It works as a decentralized trading fund on the Solana network, which leverages AI agents to gather data, evaluate market sentiment, and execute trading orders both on-chain and off-chain.
How Are Crypto AI Agents Different from Traditional Bots?
AI agents are much more advanced than traditional bots because they can think and adapt to situations instead of simply following preset rules and providing fixed responses with limited knowledge.
These characteristics do not suggest traditional bots are unuseful. Quite the contrary. For instance, crypto trading bots are popular because they make a trader’s life easier by offering several key functionalities, such as:
- Portfolio management
- Algorithmic order execution
- Leveraging insights from market analysis to inform trading decision
- Analyze data and execute trades within milliseconds.
Traditional bots can also be trained on LLMs, yet their operational scope is quite confined. Moreover, another important characteristic of AI agents is their ability to work across many systems, including blockchains and other applications, and participate in social, economic, or even political activities.
Needless to say, AI agents are experimental but very flexible— they can handle many tasks, act as assistants, and even change their responses if something isn’t working. While they’re not perfect and can make mistakes (often known as “hallucinations”), they can self-correct themselves once the user has pointed out errors, though this capability is still in the early stages of development.
So, where does Virtuals Protocol fall in? Well, some users often view it as the Pump.fun of AI and meme coins (even the overlay is similar), in the sense that users can launch their own AI projects at a much cheaper price, considering that creating and launching an AI agent from scratch would cost a lot of money, especially if it includes various functionalities such as deep learning, reinforcement learning, real-time decision-making, and other capabilities.
With Virtuals Protocol, users can do it for 100 VIRTUALS.
Virtuals Protocol: An Overview
The platform allows users to create and tokenize AI agents, allowing for co-ownership within its community. These tokenized agents also provide governance opportunities, meaning co-owners can participate in the Virtuals Protocol decision-making processes.
AI agents developed on the platform can communicate via text, speech, and 3D animation, having a multimodal approach that’s engaging and spanning immersive interactions tailored to diverse user environments. They’re even equipped with actual on-chain wallets and can execute blockchain transactions, allowing them to do several things for users money-wise from real-time revenue accrual via ERC-6551 wallets.
Virtuals Protocol Composer Stack
Central to Virtuals Protocol is the composer stack, its technological backbone that supports AI operations within the ecosystem. It can be summarised as follows:
- Agent Prompting Interface: Acts as a communication bridge between AI agents and external platforms via APIs and SDKs.
- Agentic Behavior Framework: Defines core agent capabilities, including perception, planning, and learning.
- On-Chain Wallet Operator: Enables agents to manage digital assets and execute blockchain transactions.
- Parallel Hypersynchronicity: Ensures consistent agent performance across multiple platforms simultaneously.
- Modular Consensus Framework: Enhances collaboration among contributors and validators in the ecosystem.
Another important module within Virtuals Protocol is the Generative Autonomous Multimodal Entities (G.A.M.E.) framework which allows developers to design sophisticated AI agents for virtual settings.
All in all, one of Virtuals Protocol’s main goals is to establish a revenue-sharing system where users pay to interact with AI agents, creating value for the corresponding agent tokens. Revenue will be generated through monetized user engagement, referred to as “inference costs,” which represent the compensation for the computational resources consumed during interactions.
The First AI Agent: LUNA
Luna is the first AI agent created by the very own Virtuals team as an example of their vision of transforming AI-driven entertainment, serving as a testbed to explore and demonstrate the capabilities of their technology. Luna operates 24/7 as a livestreaming entity capable of answering questions from users thanks to LLMs.
Luna holds the ticker LUNA —this is in no way or form related to Terra Luna’s cryptocurrency— and can be purchased on the platform with trades routed through the protocol’s native token, VIRTUAL. While the long-term vision involves tying LUNA to utility functions linked to Luna’s services, the token has no actual use cases beyond governance. Nevertheless, it has achieved a market capitalization of over $130M.
Other Virtuals Protocol AI Agents
As of January 2025, its ecosystem spans over 2,200 AI agents, some reaching market caps of over $300M.
AIXBT, the largest AI agent on Virtuals Protocol, launched in November 2024 by RXBT. Its digital avatar, a purple Pepe frog, tracks discussions within the crypto community on Twitter (X), using a proprietary engine to analyze these conversations and provide actionable market insights.
The Virtuals team has partnered with several projects, including Music, Pond Hub, and Seraph, distributing these tokens as airdrops to traders in the platform’s early days. Bonus allocations have occasionally been granted to LUNA token holders, contributing to their profitability.
VIRTUAL Tokenomics
VIRTUAL is the governance token of the Virtuals platform, operating both on Ethereum and Base. Its role can be summarized in three functionalities:
- Liquidity Pairing: All agent tokens, including LUNA, are paired with VIRTUAL in their respective liquidity pools.
- Transaction Routing: Purchases of agent tokens, regardless of the currency used, must pass through VIRTUAL, generating buying demand.
- Agent Creation: A fixed amount of VIRTUAL is required to create a new AI agent
Its total supply is capped at 1 billion tokens, of this:
- 600 million (60%) are in public circulation.
- 50 million are reserved for liquidity provision.
- 350 million are allocated to the ecosystem treasury for community initiatives.
Note: According to CoinGecko, the entire supply is currently in circulation. However, keep in mind that the whitepaper reads:
Ecosystem: 35% (350,000,000 tokens) is dedicated to the ecosystem treasury. This allocation is earmarked for community incentives and initiatives that drive growth within the VIRTUAL protocol ecosystem. This will sit in a DAO-controlled multi-sig wallet and will not have more than 10% emission per year for the next 3 years, subject to deployment only after receiving governance approval.
How to Create an AI Agent on Virtuals Protocol?
The process is more simple than it looks. Users can start by visiting Virtuals Protocol, selecting “Create New Agent,” and connecting their wallet.
Once connected, just follow the steps below:
1. Fill out the form with the profile picture, agent name, ticker (short symbol, max 6 characters), and a description (agent’s purpose, personality, or story). Users can also add social media links (e.g., Twitter) and a website.
2. The second step is depositing VIRTUAL tokens; users need a minimum of 100 VIRTUAL tokens to create the agent, and they will be listed on a bonding curve, where users can trade their tokens.
3. Wait for the bonding curve. Once it reaches 42,000 VIRTUAL, the agent “graduates.” This means its liquidity pool is set up on Uniswap V2, with the agent’s ticker (for example, LUNA), paired against VIRTUAL.
4. After depositing the tokens, users must delegate their voting power, which is essential for validating updates and contributions to the agent.
After all is said and done, the system handles several on-chain tasks:
- Minting Agent NFT: The agents are created as NFTs with their own unique identification.
- Immutable Vault: This keeps a record of all contributions to the agent.
- Default AI Model: A default pre-defined cognitive model is applied to make the agent functional upon deployment (even with minimal customization, they can be modified later on to affect the agent’s capabilities and behaviors).
- Approval Process: A user’s delegated voting power approves the agent’s initial AI model.
Once ready, the agent’s AI model is downloaded from the InterPlanetary File System (IPFS) and added into the Agent Runner, a hosting environment designed to run the agent’s visual cores, voice, and cognition.
When creating your AI Agent, it will look more or less like this:
- ACTIVATING: This means the system is setting up the agent (usually takes around 5 minutes, so there is no reason to panic).
- AVAILABLE: The agent is fully functional and can interact on platforms like Telegram or be managed via a dashboard.
FAQ
What is Virtuals Protocol?
Virtuals Protocol is one of the largest AI launchpads, allowing users to launch their own agents at a much cheaper price and equipped with unique abilities, like interacting with social media communities, executing trades and transactions, and even livestream 24/7.
What Are AI Agents Capable Of?
AI agents are effective in many fields that require task planning and data gathering. They are particularly effective at promoting cryptocurrencies and influencing markets.
For example, Truth Terminal can act as a social media influencer, manage communities, and even start its own crypto projects. This is possible because these agents use advanced technologies like LLMs and deep learning, allowing them to create content and take actions independently, even though sometimes there may be a layer of human approval behind it (again, it’s quite experimental).
Is AI Crypto Risky?
There are a few risks to consider when creating AI agents or trading tokens associated with these projects. To mention a few:
- AI agents can be exploited for market manipulation schemes to destroy real values.
- AI Agents are not bulletproof: they can still be targets for hackers and malicious actors that can use them to mislead other users in the community
- AI systems could be influenced by biases in their data set, leading to wrongful, skewed decisions.
