Author: Edgy Source: X, @thedefiedge Translator: Shan Ouba, Jinse Finance
Currently, it is unlikely that a full-blown Altcoin season, where all Altcoins rise together, will occur.
The Altcoin market is fragmented - there are too many tokens in the market now, and liquidity is insufficient to support the rise of all tokens (even if retail investors return en masse).
Only certain tracks will see a full-blown Altcoin season. Currently, the AI Agent track is in this stage.
While the broader market is generally declining, this area is continuously setting new historical highs. AI Agents are attracting market liquidity and attention.
If you missed the DeFi summer of 2020 or the Alt Layer 1 rally of 2021... the market is giving you another chance!
Don't try to go against the trend, don't be too clever.
If you've been on the sidelines lately, you may wonder if it's "too late" to get in now? I don't think so:
• The overall market is currently depressed, but AI infrastructure and agent projects are setting new historical highs. Imagine what will happen when the market rebounds?
• The entire AI Agent track has just started about 2 months ago (starting with the launch of GOAT). These AI Agents are currently in a "very early" stage. As time goes on, we are unlocking new features every day.
Imagine the changes when AI Agents can optimize your portfolio and trade directly on-chain?
• Treat AI infrastructure as Alt Layer 1, and AI Agents as DApps. Some AI infrastructure protocols may reach Alt Layer 1 valuation levels ($1 billion+).
The key difference is: most AI infrastructure projects do not have VC backing, so you don't have to worry about selling pressure from large-scale token unlocks.
• The top AI protocols have not yet been listed on CEXs (centralized exchanges). I have a few friends who are interested in this track but have encountered some difficulties in getting in.
I will mainly discuss AI infrastructure projects, as they currently have the best risk/reward ratio.
Rather than picking individual AI Agents, it's better to invest in the entire infrastructure, like "owning the whole casino". In a gold rush, the richest people are those who sell shovels and pickaxes.
Let's dive deeper into this track.
The Two AI Infrastructure Kings
1. Virtuals (@virtuals_io) - The Reigning King
• Market Cap: $4.6 billion
• Position: The leading Launchpad platform for AI Agent projects
Virtuals has become the go-to Launchpad for AI Agent projects, currently dominating the Base ecosystem, but may expand to other chains in the future. Its co-founder @ethermage hinted that they are preparing to launch the platform on Solana, and plan to expand to @HyperliquidX and @AbstractChain by 2025.
Virtuals has recently released a series of strategic goals, such as attracting top AI Agent developers and improving technology to make Virtuals the fairest stage for agent investors. Will Virtuals introduce a revenue-sharing mechanism for AI Agents?
Three Core Advantages:
1. Token Economic Model: Creating an Agent requires $VIRTUALS, and all new Agent liquidity pairs also need $VIRTUALS. Almost all new Agent Launchpads are mimicking their token economic model.
2. Network Effect: Virtuals currently attracts tremendous market attention, and launching an Agent on this platform means enjoying the tailwind of the Base ecosystem.
3. Flagship Agents: @AIXBT is the first Agent to reach a market cap of over $500 million.
Although I don't like to make price predictions, Virtuals may enter the top 10 market cap this cycle.
2. Ai16z ($AI16Z) - The Challenger King
• Market Cap: $2.2 billion
• Position: The origin of the ElizaOS Agent framework, one of the most popular code repositories on Github.
Ai16z has gone through some ups and downs since its inception, but its technical and community influence continues to grow.
The biggest issue was that its token economic model failed to effectively capture value. Now, they are undergoing a major overhaul and will launch a new Launchpad in Q1 2025.
To launch a project on the Eliza framework will require:
• Paying a launch fee (a portion of which will be used for $AI16Z buyback and burning)
• Holding $AI16Z to gain access and liquidity pairing opportunities
• The project token needs to be paired with $AI16Z to stabilize the market
It's like owning a share of a casino, where every new Agent has to pay an "entry fee".
Strategies and Plans:
• Diversified Treasury Strategy: Diversifying investments into major L1 tokens, stablecoins, and project tokens, and providing liquidity for $AI16Z trading pairs to ensure long-term sustainability and avoid selling pressure.
• Agent Swarms: Agent swarms will be the future focus, where multiple Agents work collaboratively, which is the beginning of decentralized Artificial General Intelligence (AGI).
• Ecosystem Incentives: Developers building on Eliza will receive:
• Ecosystem fund support
• Fee-sharing opportunities
Positioning $AI16Z as the "glue" of the entire decentralized AI stack, similar to $ETH for the Ethereum ecosystem or $SOL for Solana.
Who is Stronger, Virtuals or Ai16z?
Virtuals' Advantages:
• Strategic foresight, leading the industry, with competitors still catching up.
• The boost from the Base ecosystem, providing strong momentum for its growth.
Ai16z's Advantages:
• The ElizaOS framework technology is leading, providing a better experience for Agent development.
• The token economic model reform will solve the value capture issue.
The crypto industry loves to take sides, but you don't have to choose one. I'm not sure who will ultimately win this race, so directly holding both might be a simpler choice. However, I'm more tilted towards Virtuals and its ecosystem.
Many industries eventually evolve into a duopoly, like Apple and Android, or Coca-Cola and Pepsi. Will this track also see a similar scenario? Let's wait and see.
The Competitors
While Virtuals and Ai16z are currently dominating the market, this game has just begun. If you are seeking higher returns, you can consider some more "elastic" options.
ZereBro (Market Cap $632 million) - Transforming
ZereBro was initially a general-purpose AI Agent that could also create music. But now it is transforming into a complete full-stack framework for creating other Agents.
ZerePy is an open-source Agent framework based on Python, and the status of Python is undeniable:
1. Python is the most popular programming language globally.
2. The vast majority of AI tools and libraries are written in Python.
You can use it to launch an Agent similar to ZereBro.
The Zentients proxy platform will be launched in early 2025. This new Launchpad will directly empower $ZEREBRO, with each created proxy flowing back value through launch fees, liquidity pairing (LP) binding, and innovative liquidity locking mechanisms.
ZereBro has already joined the thriving Virtuals ecosystem and launched the ZereBro/Virtual liquidity pool on @Base to expand its influence.
They are following the successful strategies of current industry leaders, but their current valuation is far lower than their competitors.
$ARC (market cap $358 million) - Rust-based AI agent framework
Arc.fun is building an open-source agent framework called Rig, which is currently the only agent framework built on Rust. Rust is the native language of Solana, which is a critical point.
Why is $ARC worth watching?
• Rust language can optimize performance through zero-cost abstraction.
• Rig framework outperforms Langchain in terms of memory and cold start time.
• Core contributor @0xCygaar has started contributing code to the Rig Github, and he is also one of the core developers of Abstract Chain.
In short: Arc = Rust + excellent code library.
$GRIFFAIN (market cap $457 million) - AI app store on Solana
Griffain allows you to search and execute any operation on-chain through natural language, and agents can autonomously complete on-chain transactions.
Core highlights:
• Griffain agents can automatically execute preset trades on DEXs, such as hourly DCA or limit orders.
• Agents can also send the latest address holdings and transaction reports via email.
• The project has received public support from @aeyakovenko and Solana Labs, and plans to build AI agents similar to Siri on iPhone on the Solana Mobile platform.
• Griffain was built by @tonyplasencia3, who was previously a developer on the Underdog and Blink projects on Solana.
Why is this important?
Many currently criticize AI agents as "just chatbots" without practical applications. Griffain is building the infrastructure for useful on-chain AI agents.
REI (market cap $153 million) - A smarter new take on AI agents
REI is not just a regular AI agent, it is a full-stack framework that combines AI innovation with blockchain. Think of REI as an intelligent entity that can not only act, but also learn, reason, and adapt.
Functional highlights:
• REI primarily interacts with users through X (Twitter), with natural and contextual conversational abilities.
• Oracle Bridge connects off-chain data, while the ERCData standard optimizes on-chain data gas costs, making on-chain AI applications more efficient.
• REI's 4-layer cognitive model (thinking, reasoning, decision-making, execution) gives the agent the ability to learn and self-adjust, enabling deterministic execution.
Upcoming catalysts:
• AI Stack SDKs: Rolled out in phases, lowering the barrier for developers to create agents.
• Expanding oracle network: Laying the foundation for building real-world on-chain AI applications.
• Phase 2 plans: Comprehensive testnet, advanced verification protocols, and 2025 mainnet plans.
What's the big picture? REI's goal is not just to create smarter robots, but to shape the future of decentralized intelligent entities through the precise integration of AI and blockchain. If you're interested in the future of autonomous agents, REI is a project worth long-term tracking.
$VAPOR - (market cap: $154 million) - The first and only AI agent Launchpad on Hyperliquid
Vapor is a native-built AI agent Launchpad on Hyperliquid, using the Eliza framework from @ai16zdao.
Key points:
• Although $AI16Z accounts for 1.2% of Solana's fully diluted valuation (FDV), $VAPOR's market cap is only $132 million, which is 0.47% of Hyperliquid's FDV.
• After launch, users must burn $VAPOR to activate agents, meaning $VAPOR is a deflationary token with a continuously decreasing total supply.
• All agent tokens launched on Hyper EVM will be traded in a local DEX pool paired with $VAPOR. As more agents are launched, more $VAPOR will be locked in the liquidity pool.
• Already launched on the testnet, where all agents can be approved without holding tokens.
MODE (market cap: $114 million) - The intersection of AI agents and DeFi
The Mode Network can be seen as an Ethereum Layer 2 focused on extending DeFi to billions of users through on-chain AI agents.
"By 2026, over 80% of on-chain transactions will be completed by AI agents." If you believe this prediction, Mode will let you get ahead of the curve.
• TVL: Over $500 million, the third largest chain in the OP Superchain ecosystem.
• User growth: 367,000 users and 24 million+ transactions.
• Undervalued: MC/TVL ratio indicates potential upside.
Mode's latest updates:
• Synth testnet launch: Bittensor subnet for financial forecasting, unlocking DeFi automation.
• AI agent hackathon: First hackathon, attracting 667 registrants and developing 10+ Eliza plugins in just 6 days.
• AI agent app store: Multiple agents already launched, including:
• Amplifi: Bitcoin abstraction layer, using AI to enable Bitcoin expansion.
• Arma: Optimizing USDC yield farming.
• Brian: Allowing natural language execution of trades and contracts.
• Sturdy: Aggregating AI-driven yield vaults to boost yields.
• QuillAI Network: Secure AI agent supported by D-LLM and EigenLayer AVS.
Mode's vision: Building the infrastructure for the on-chain agent economy. With the Synth testnet, BitcoinOS integration, and a growing AI agent ecosystem, Mode will play a key role.
$GAME (market cap: $390 million) - The pillar of the Virtuals framework
GAME is the first infrastructure project that allows developers to integrate AI agents into games.
From Virtuals' description:
"Every society has core infrastructure to drive its development and growth. @game_virtuals plays the role of Shopify in the Virtuals ecosystem, while @base is like SWIFT."
Significant progress:
• Directly developed by the Virtuals Protocol team.
• $GAME is the pillar of the virtual economy, with each AI agent run paying $GAME as an operating fee.
• GAME has supported over 200 projects and processes over 150,000 requests per day.
• Within the Virtuals ecosystem, G.A.M.E's 30-day performance ranks first, with growth approaching 400%.
• GAME has the most Lifetime Inferences, exceeding 2.2 million.
Why is $GAME undervalued? People don't understand its functionality much, even the G.A.M.E team itself hasn't promoted it heavily (their official account only has 5 tweets, and they haven't even changed their Twitter header).
A similar but riskier choice is CONVO, which provides the driving force for AI agent dialogues on Virtuals.
Soul Graph ($GRPH, market cap: $36 million) - Making AI agents more human-like
Imagine if other frameworks shape the "hard skills" of agents, then Soul Graph shapes their "soft skills", making AI agents more human-like. Its framework aims to imbue new or existing AI agents with personality, giving them persistent memory and real-time communication capabilities.
"We are building open-source tools and infrastructure to enable developers to imbue agents with evolving personalities, persistent memory, and real-time communication capabilities."
Key points:
• Over 2,000 "souls" have been created, providing 15,000 minutes of voice interaction time in the Soul Graph Playground.
• Recently integrated the AI16Z plugin.
• 85% of the total token supply is allocated to the community, with the development team's allocation locked and vesting for 1 year.
Reasons I'm bullish on this project:
Character AI is booming. Ordinary users spend hours chatting with AI every day, and Google just acquired Character.ai for $2.7 billion. Imagine if users could interact in real-time with AI agents with human-like personalities? Soul Graph is making this happen.
Other projects worth watching
• Bully (@dolos_diary): Initially a "rant" account, now developing a framework @dolion_ai for social media AI agents.
• TopHat (@tophat_one): AI agent Launchpad on Solana.
• Swarm Node (@Swarmnode): Similar to AWS Lambda, but specialized for AI agents and frameworks.
• Cookie Fun (@cookiedotfun): Data tracking platform for AI agents, similar to DeFiLlama/Kaito, tracking agent influence, market share, etc. Advanced features require holding tokens to access.
How to choose?
First, Virtuals and AI16z are must-have configurations. They are clearly dominating the market, and I don't see this trend slowing down anytime soon, the network effects are not to be underestimated.
If considering other promising projects, I focus on the following:
• LaunchPads on new chains: This was the strategy template in 2021, with Virtuals dominating Base, so you can look for LaunchPads on other hot chains, like $VAPOR's amazing run on Hyperliquid.
• Niche market strategies: It's hard to launch a generic L1/L2 anymore, focusing on niche markets has a better chance. For example, Mode focuses on the intersection of AI agents and DeFi, while Griffain focuses on natural language interaction.
• Programming languages: Infrastructure needs agents, agents need developers, and developers have their own language preferences. Therefore, Zerebro (Python) and Arc (Rust) have strong competitiveness.
AI portfolio ideas
1. Passive portfolio - 50% in Virtuals, 50% in ai16z. This simple combination may outperform 80% of investors.
2.Alpha/Beta portfolio - Use Virtuals as the core Alpha investment, then choose competitive frameworks like Arc or Griffain as Beta options.
3.Barbell strategy - Virtuals and ai16z as the "safe" bets, then mix in various high-risk A.I. agent projects.
4.Comprehensive coverage - Alpha investments in infrastructure, then add some Beta projects, and then diversify into A.I. agent projects of different risk levels. Include leading agents like @aixbt, and also configure some small projects with a market cap under $10m.
Criticisms of the AI agent track
"Too many Launchpads and frameworks" - Guys, have you seen how many L1s and L2s there are now? We're doing great. This field is new and progressing rapidly. If you're building a new city, it's reasonable to first build roads, bridges, and lay pipes.
"Most of them are trash and will eventually go to zero" - This is the crypto world, most things will indeed go to zero. But the ones that survive will disrupt the game.
"The AI track has already peaked" - You're doomed to fail. Don't be too smart and take the middle ground.
"All AI agents suck" - Remember when people were afraid to shop online? I also remember the first iPhone couldn't even copy and paste. We're still in the early stages, this is a bet on the future.
Conclusion
I know this cycle will see a bull market, but I'm not sure what new, game-changing primitives will emerge.
Every cycle has a "DeFi" or "NFT" moment - a track that re-energizes developers, and one that makes retail suddenly wake up and go crazy.
I've doubted whether that moment would truly arrive. But the good news is... it has. This is your chance to achieve financial freedom - don't mess it up.
