The A.I. Agent Supercycle: A Guide to the Best Infrastructure Plays

Intermediate1/15/2025, 7:38:25 AM
The article provides a detailed overview of market leaders Virtuals and Ai16z, exploring their project backgrounds, core strengths, and future potential. It also analyzes other promising competitors like ZereBro, ARC, and Griffain.

There’s a strong possibility that there won’t be a full-blown alt season where everything goes up.

Altcoin dispersion - there are too many tokens now, and there’s not enough liquidity to pump them all (even if retail’s back in full force).

Only some sectors get a full-blown altcoin season. Right now, you’re in that season now with A.I. Agents.

While the market is bleeding red, this sector is breaking all-time highs. It’s a blackhole that’s sucking liquidity and attention. If you missed DeFi Summer in 2020, or the Alt Layer 1 trade in 2021…

THE UNIVERSE IS GIVING YOU ANOTHER CHANCE. Don’t try to be contrarian. Don’t try to be cute.

If you’ve been sidelined for a while, you may be wondering if it’s “too late” to enter. I doubt it:

  • The markets are bleeding now, yet A.I. infrastructure and agents are printing all-time highs. What do you think happens once the market rebounds?
  • The whole meta is a little over 2 months old (starting with the launch of GOAT). These A.I. agents are the worst they’re ever going to be. Day by day we’re unlocking new utilities. Imagine what happens once they’re able to optimize your portfolio and trade on-chain for you.
  • Value the the Infrastructures as Alt Layer 1s, and A.I. agents as Dapps. We could see some of these protocols valued as Alt Layer 1s ($10B+). The big difference? Most of them aren’t backed by VCs so you don’t have to deal with massive unlocks.
  • None of the top protocols have CEX listings yet. A few of my IRL friends are interested in this play, but they’re having trouble onboarding.

I’m mainly going to talk about A.I. Infrastructure plays because they’re the best risk / reward now.

Instead of picking individual agents, you’re pretty much owning the whole casino. In a gold rush, the richest people were the ones selling picks and shovels.

(I’ll do a separate thread on A.I. agents in the future)

Let’s dive in.

The Two Kings of A.I. Infrastructure

@virtuals_io - ($4.6B) - The Current King

Virtuals has positioned itself as the go to launchpad for A.I. Agents.

It’s dominating Base, but it could expand to other chains. Co-founder, @ethermage, has teased a ready-to-launch platform on Solana. They’re also exploring expanding to @HyperliquidX and @AbstractChain in 2025.

They recently posted about their high-level strategy initiatives such as attracting top agent builders, improving the tech, and making Virtuals the fairest place for agent investment. Pay close to attention to how they plan on making this a reality.

Could we see A.I. agent revenue share soon?

The 3 biggest strengths:

  1. Tokenomics. Virtuals is needed to create agents, and virtuals is needed for the Liquidity pairing of all tokens. Every new agent launchpad is imitating their tokenomics.
  2. Network Effect. All eyes are on Virtuals now. To launch an agent there means you have the tailwinds of Virtuals and Base behind you.
  3. The biggest Agent. @AIXBT - is its 1st agent to cross the $500mc barrier.

I hate making price predictions, but it’s possible Virtuals can enter the top 10 market cap this cycle.

  1. Ai16z (2.2B MC) - The Kingslayer?

Ai16z has had quite the rollercoaster ride since its inception. Ai16z is the home of the leading agent framework, ElizaOS. It’s the #1 trending repository on Github.

The biggest complain before was their tokenomics, and how they weren’t capturing value. They’re about to do an overhaul which unveil a new Launchpad, set for Q1 2025.

Projects launching on the Eliza framework will:

  • Pay launch fees (a % flows back to $AI16z via buybacks & burns).
  • Require $AI16z holdings for access and liquidity pairings.
  • Pair tokens with $AI16z in liquidity pools to stabilize markets.

It’s like owning part of the casino where every new agent pays to play.

  • Treasury Strategy w/ Power Moves: Plans to diversify across major L1 tokens, stablecoins, and project tokens while providing liquidity for $AI16z pairs.

Think long-term sustainability without fire sales.

  • Agent Swarms: The key focus? Clusters of AI agents collaborating. This is where decentralized AGI starts becoming real.
  • Ecosystem Incentives: Developers building on Eliza will get access to: Ecosystem funds and Fee-sharing opportunities.

The idea? Turn $AI16z into the glue binding the entire decentralized AI stack. Think $ETH or $SOL but for autonomous agents.

Is Virtuals or Ai16z Better?

Virtuals: I really their strategy. They think far ahead, and it does feels like everyone else is just playing catch up. They’re also riding the momentum of the Base ecosystem.

Ai16z: From what I’ve seen, ElizaOS does seem like better tech for developing agents. And their tokenomics overhaul will solve many of the issues they had.

I know this industry loves tribalism, but you don’t have to pick sides. I’m not sure who’s going to win in the long run - it’s simpler just to own both. (I am more allocated towards Virtuals and its ecosystem).

Many industries eventually consolidate to becoming a 2 horse race. Apple and Android. Coke and Pepsi. Could we see the same for this sector?

The Contenders

Virtuals and Ai16z are dominating, but it’s still early. If you want higher returns, then consider some beta plays.

ZereBro ($632mc) - The Pivot is here

ZereBro started as a generic A.I. agent that also made music. But now it’s pivoting into a full-stack framework for creating other agents.

Zerepy is an open sourced agent framework based on Python. Python’s a big deal.

  1. it’s the most popular programming language in the world

  2. Most of the A.I. tooling and library are written in Python

It’ll let you launch your own agent that’s similar to Zerebro.

The biggest catalyst is the Zentients Agent Platform that’s coming in early 2025.

This new launchpad ties value directly to $ZEREBRO. Every agent created will funnel value back through launch fees, LP bonding, and innovative mechanisms that lock liquidity.

It’s now part of the thriving Virtuals ecosystem, launching a ZereBro/Virtual pool on @Base to expand its reach.

They’re running a similar playbook to the current kings, but it’s trading at a much lower valuation now.

$ARC (@arcdotfun) - MC: $358m - AI Agents Framework Built on Rust

Arc .fun is building an open-source agents framework named Rig. This is the only agent framework built on Rust at the moment. This is a big deal because Rust is Solana’s native language.

What Make $ARC Interesting?

  • The Rust language helps optimize performance due to zero-cost abstractions.
  • Rig Framework has proved to outperform Langchain on memory and cold start times.
  • @0xCygaar is contributing to The Rig Github. He’s a core contributor to Abstract Chain.

Basically, Arc = Rust and Good codebase.

$GRIFFAIN - MC: $457m - The A.I. App store of Solana

Griffain allows you to search anything and do anything on-chain by using natural language with your own agent.

Key Notes

  • Griffain agents can autonomously execute on-chain transactions on DEX set beforehand, such as DCA every hour or Limit Orders to buy or sell. Or get emailed up-to-date reports about what tokens any address owns and what changes they’ve made to their portfolio
  • The project is publicly supported by @aeyakovenko and Solana Labs. They are aiming to build Griffain on Solana Mobile the same as Siri on iPhone.
  • Griffain is built by @tonyplasencia3, the previous builder of Underdog and Blink on Solana.

This project to be interesting.

I’m trying to look at where the puck is going. Everyone keeps criticizing agents as “chatbots” and lack utility. Griffain is setting up the infrastructure for agents that can be useful onchain.

REI ($153mc) - A Smarter Take on AI Agents

At its core, REI isn’t just another Ai agent bot. It’s a full-stack framework combining AI innovation with blockchain.

Think of it as AI that doesn’t just act but learns, reasons, and adapts

What does it do? REI primarily interacts with users through X, where her cognitive architecture enables natural, context-aware interactions.

How does she work? Her Oracle Bridge connects off-chain data, while ERCData Standard structures it for gas-efficient on-chain use.

Together with REI’s 4-Layer Cognitive Model (Thinking, Reasoning, Deciding, Acting), this allows her to learn, adapt, and deliver deterministic execution.

Recent Developments shaping REI’s potential:

  • AI Stack SDKs: Rolling out in phases, making agent creation more accessible for devs.
  • Expanding Oracle Network: Critical for building real-world AI applications on-chain.
  • Phase 2 Catalysts: Full testnet launch, advanced validation protocols, and mainnet plans in 2025.

What’s the big picture? REI isn’t just about smarter bots. It’s shaping decentralized intelligence by combining AI x blockchain precision.

If you’re looking for the future of autonomous agents, REI is a framework worth watching.

$VAPOR - (MC: $154m) - First and Only AI Agent Launchpad on Hyperliquid

Vapor is a native AI Agent launchpad on Hyperliquid, and was built using the @ai16zdao Eliza framework.

Key Notes

  • While $AI16Z is about 1.2% of Solana’s FDV, $VAPOR is sitting at $132m, equivalent to about just 0.47% of Hyperliquid’s FDV.
  • Once live, users have to burn $VAPOR to launch an agent. This means the $VAPOR is deflationary and the total supply will keep decreasing.
  • All agent tokens launched on Hyper EVM will be tradable in a native DEX pool paired with $VAPOR. As more agents are launched, more $VAPOR will be locked in liquidity pools.
  • It’s already live on testnet. All agents are now approved without token gated.

MODE ($114mc) - Where A.I. Agents Meet DeFi

You can think of Mode Network as an ETH Layer 2 that focuses on scaling DeFi to billions of users through on-chain AI Agents.

“by 2026, more than 80% of on-chain TX will be via AI agents”. If you believe this to be the case, then Mode will get you exposure.

  • $500M+ TVL: The #3 chain in the OP Superchain ecosystem.
  • User Growth: 367k users with 24M+ transactions.
  • Undervalued: MC/TVL ratio signals potential upside.

What’s happening on Mode?

  • Synth Testnet launch: A Bittensor subnet for financial forecasting, unlocking DeFi intelligence.
  • AI Agent Hackathon: First-ever hackathon concluded with 667 registrations and 10+ Eliza plugins in just six days.

Developments on AI Agent App Store. Currently hosts few live agents, including:

  • Amplifi: Bitcoin Abstraction Layer. Scaling Bitcoin To Billions With AIFi.
  • Arma: Optimizes USDC yield farming.
  • Brian: Enables natural language execution for transactions and contracts.
  • Sturdy: Aggregates AI-powered yield vaults for higher returns.
  • QuillAI Network: Security AI agents powered by D-LLM & EigenLayer AVS

Mode’s vision? Build the infrastructure for an on-chain agent economy. With the Synth Testnet, BitcoinOS integration, and growing AI agents

$GAME (@game_virtuals) - MC: $390m - The Virtuals Framework

G.A.M.E was the first infrastructure project that allowed developers to integrate AI Agents into games.

From Virtuals:

“Every society has core infrastructure crucial to its growth and development. @game_virtuals plays that role as the Shopify for Virtuals, with @base acting as SWIFT.”

Notable Traction

  • Developed by Virtuals Protocol team themselves.
  • $GAME is the backbone of the Virtual economy. Whenever an AI agent operates, it effectively pays a fee in $GAME to keep the ecosystem running.
  • G.A.M.E already empowers more than 200 projects. It leads to more than 150,000 requests being sent on G.A.M.E every day.
  • Top #1 in 30D performance within the Virtuals ecosystem with nearly 400% growth.
  • G.A.M.E. has the highest Lifetime Inferences on Virtuals Protocol now at over 2.2m inferences.

G.A.M.E. still feels undervalued because people don’t really know what it does. And the G.A.M.E. team themselves don’t really shill it

(there’s like 5 tweets total from the official account. The bros haven’t even updated the twitter header lol)

A similiar and higher risk play is CONVO which powers the conversations on Virtuals.

Soul Graph (@soulgra_ph"">@soulgra_ph) $GRPH - ($36mc) - Making Agents more Human Like

Imagine if other frameworks shape the “hard” skill of the agents, Soul graph shapes its ‘soft” skill and makes it more human. Its framework aims to add personality to new or existing AI Agents.

“we’re building open source tools & provide infra that let developers give their agents an evolving personality, persistent memory and real time comms”

Key Notes

  • It already crossed 2,000 souls created and 15,000 voice-to-voice minutes served in soulgraph playground.
  • Soulgraph has recently integrated with AI16Z plugins.
  • 85% total supply is allocated to the community. Dev allocation is locked and vesting for 1 year.

I think this is going to be big.

Thesis: Character ai is huge. Normies log into that website and spend hours talking to A.I. text. Google just bought them for $2.7B). So imagine if people could talk to A.I. agents in real-time. And these agents have human like personalities / memories.

That’s what Soul Graph does. Go to the website and see for yourself.

Others

This article’s getting quite long, and I’m unable to cover every single protocol. Here are a few other protocols to pay attention to.

• Bully (@dolos_diary) - Started as a roasting account. It now features @dolion_ai which is a framework for social media a.i. agents.

• TopHat (@tophat_one) - Solana agent launchpad.

• Swarm Node @Swarmnode - Think of it like AWS Lambda, but specifically for AI agents and frameworks.

• Cookie Fun @cookiedotfun - it’s like the DeFiLlama / Kaito of A.I. agents - it tracks data such as mindshare & smart engagement / marketcap Advanced features are token gated.

Which to Choose?

First, everyone should have exposure to both Virtuals and AI16z. They’re clearly dominating and I don’t see signs of that slowing down. Don’t underestimate network effects.

If you’re considering some beta plays, here’s what I’m looking at:

• LaunchPad on a new Hot Chain. This is a playbook straight from 2021. Virtuals dominates Base. So, bet on the LaunchPad for other hot chains. This is why Vapor has skyrocketed in price because Hyperliquid’s so hot right now.

• Is it Niche’ing down? It’s hard to launch a general purpose Layer 1 or Layer 2 now. Instead, you gotta focus on something like being the “gaming chain.” This is why I find Mode interesting as it’s laser focusing on A.I. agents x DeFi, or how Griffain is focused on natural language.

• Programming Language. Infrastructure projects need agents. Agents need developers. Developers have their preferred programming language. This is where Zerebro (Python) and Arc (Rust) Shine.

A.I. Portfolio Ideas

  1. The Chill Portfolio - 50% Virtuals and 50% ai16z. This is simple and will probably outperform 80% of people.
  2. Alpha / Beta - Virtuals as the alpha play, and you pick a competing framework as a beta such as Arc or Griffain
  3. The Barbell - Virtuals and ai16z as your “safe” bets. Then you go for higher risk play with various A.I. agents.
  4. A little of Everything - Infrastructure Alpha Plays and some beta plays. Then you throw in various A.I. agents across the risk curve. Some alpha ones like @aixbt, and go into some sub $10mc trenches.

Criticisms of A.I. Agent Sector

“Too many launchpads and frameworks” - Bros, do you see how many layer 1s and and layer 2s there are? We’re doing fine. This entire space is new and moving fast. If you want to build a new city, it makes sense to build the roads, bridges, and pipes first.

“Most of this are garbage and will go to zero” - This is crypto. Most things go to shit, but the stuff that survive changes the game.

“A.I. sector has topped” - you’re not gonna make it. Don’t midcurve it.

“All the a.i. agents suck” - there was a time when people were too scare to order things online. I also remember when the first iPhones couldn’t copy and paste. We’re early. This is a bet on the future.

Final Thoughts

I knew there would be a bull run this cycle. What I wasn’t sure of his what new game-changing primitive would be introduced.

What would be the “DeFi” or “NFT” moment for this cycle. The one sector that would re-invigorate all the developers. The sector that would make retail go “ah ha” and ape in.

I wasn’t sure if there was going to be one. But the good news is…it’s here.

This is your chance to make it - don’t fuck it up.

If you enjoyed this article, please help me by engaging with this. Took a ton of time to research and write.

Disclaimer: Nothing here is sponsored. I have no free allocation to anything. I have exposure to several protocols listed here.

Disclaimer:

  1. This article is reprinted from [X]. All copyrights belong to the original author [@thedefiedge]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
  2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
  3. The Gate Learn team does translations of the article into other languages. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.

The A.I. Agent Supercycle: A Guide to the Best Infrastructure Plays

Intermediate1/15/2025, 7:38:25 AM
The article provides a detailed overview of market leaders Virtuals and Ai16z, exploring their project backgrounds, core strengths, and future potential. It also analyzes other promising competitors like ZereBro, ARC, and Griffain.

There’s a strong possibility that there won’t be a full-blown alt season where everything goes up.

Altcoin dispersion - there are too many tokens now, and there’s not enough liquidity to pump them all (even if retail’s back in full force).

Only some sectors get a full-blown altcoin season. Right now, you’re in that season now with A.I. Agents.

While the market is bleeding red, this sector is breaking all-time highs. It’s a blackhole that’s sucking liquidity and attention. If you missed DeFi Summer in 2020, or the Alt Layer 1 trade in 2021…

THE UNIVERSE IS GIVING YOU ANOTHER CHANCE. Don’t try to be contrarian. Don’t try to be cute.

If you’ve been sidelined for a while, you may be wondering if it’s “too late” to enter. I doubt it:

  • The markets are bleeding now, yet A.I. infrastructure and agents are printing all-time highs. What do you think happens once the market rebounds?
  • The whole meta is a little over 2 months old (starting with the launch of GOAT). These A.I. agents are the worst they’re ever going to be. Day by day we’re unlocking new utilities. Imagine what happens once they’re able to optimize your portfolio and trade on-chain for you.
  • Value the the Infrastructures as Alt Layer 1s, and A.I. agents as Dapps. We could see some of these protocols valued as Alt Layer 1s ($10B+). The big difference? Most of them aren’t backed by VCs so you don’t have to deal with massive unlocks.
  • None of the top protocols have CEX listings yet. A few of my IRL friends are interested in this play, but they’re having trouble onboarding.

I’m mainly going to talk about A.I. Infrastructure plays because they’re the best risk / reward now.

Instead of picking individual agents, you’re pretty much owning the whole casino. In a gold rush, the richest people were the ones selling picks and shovels.

(I’ll do a separate thread on A.I. agents in the future)

Let’s dive in.

The Two Kings of A.I. Infrastructure

@virtuals_io - ($4.6B) - The Current King

Virtuals has positioned itself as the go to launchpad for A.I. Agents.

It’s dominating Base, but it could expand to other chains. Co-founder, @ethermage, has teased a ready-to-launch platform on Solana. They’re also exploring expanding to @HyperliquidX and @AbstractChain in 2025.

They recently posted about their high-level strategy initiatives such as attracting top agent builders, improving the tech, and making Virtuals the fairest place for agent investment. Pay close to attention to how they plan on making this a reality.

Could we see A.I. agent revenue share soon?

The 3 biggest strengths:

  1. Tokenomics. Virtuals is needed to create agents, and virtuals is needed for the Liquidity pairing of all tokens. Every new agent launchpad is imitating their tokenomics.
  2. Network Effect. All eyes are on Virtuals now. To launch an agent there means you have the tailwinds of Virtuals and Base behind you.
  3. The biggest Agent. @AIXBT - is its 1st agent to cross the $500mc barrier.

I hate making price predictions, but it’s possible Virtuals can enter the top 10 market cap this cycle.

  1. Ai16z (2.2B MC) - The Kingslayer?

Ai16z has had quite the rollercoaster ride since its inception. Ai16z is the home of the leading agent framework, ElizaOS. It’s the #1 trending repository on Github.

The biggest complain before was their tokenomics, and how they weren’t capturing value. They’re about to do an overhaul which unveil a new Launchpad, set for Q1 2025.

Projects launching on the Eliza framework will:

  • Pay launch fees (a % flows back to $AI16z via buybacks & burns).
  • Require $AI16z holdings for access and liquidity pairings.
  • Pair tokens with $AI16z in liquidity pools to stabilize markets.

It’s like owning part of the casino where every new agent pays to play.

  • Treasury Strategy w/ Power Moves: Plans to diversify across major L1 tokens, stablecoins, and project tokens while providing liquidity for $AI16z pairs.

Think long-term sustainability without fire sales.

  • Agent Swarms: The key focus? Clusters of AI agents collaborating. This is where decentralized AGI starts becoming real.
  • Ecosystem Incentives: Developers building on Eliza will get access to: Ecosystem funds and Fee-sharing opportunities.

The idea? Turn $AI16z into the glue binding the entire decentralized AI stack. Think $ETH or $SOL but for autonomous agents.

Is Virtuals or Ai16z Better?

Virtuals: I really their strategy. They think far ahead, and it does feels like everyone else is just playing catch up. They’re also riding the momentum of the Base ecosystem.

Ai16z: From what I’ve seen, ElizaOS does seem like better tech for developing agents. And their tokenomics overhaul will solve many of the issues they had.

I know this industry loves tribalism, but you don’t have to pick sides. I’m not sure who’s going to win in the long run - it’s simpler just to own both. (I am more allocated towards Virtuals and its ecosystem).

Many industries eventually consolidate to becoming a 2 horse race. Apple and Android. Coke and Pepsi. Could we see the same for this sector?

The Contenders

Virtuals and Ai16z are dominating, but it’s still early. If you want higher returns, then consider some beta plays.

ZereBro ($632mc) - The Pivot is here

ZereBro started as a generic A.I. agent that also made music. But now it’s pivoting into a full-stack framework for creating other agents.

Zerepy is an open sourced agent framework based on Python. Python’s a big deal.

  1. it’s the most popular programming language in the world

  2. Most of the A.I. tooling and library are written in Python

It’ll let you launch your own agent that’s similar to Zerebro.

The biggest catalyst is the Zentients Agent Platform that’s coming in early 2025.

This new launchpad ties value directly to $ZEREBRO. Every agent created will funnel value back through launch fees, LP bonding, and innovative mechanisms that lock liquidity.

It’s now part of the thriving Virtuals ecosystem, launching a ZereBro/Virtual pool on @Base to expand its reach.

They’re running a similar playbook to the current kings, but it’s trading at a much lower valuation now.

$ARC (@arcdotfun) - MC: $358m - AI Agents Framework Built on Rust

Arc .fun is building an open-source agents framework named Rig. This is the only agent framework built on Rust at the moment. This is a big deal because Rust is Solana’s native language.

What Make $ARC Interesting?

  • The Rust language helps optimize performance due to zero-cost abstractions.
  • Rig Framework has proved to outperform Langchain on memory and cold start times.
  • @0xCygaar is contributing to The Rig Github. He’s a core contributor to Abstract Chain.

Basically, Arc = Rust and Good codebase.

$GRIFFAIN - MC: $457m - The A.I. App store of Solana

Griffain allows you to search anything and do anything on-chain by using natural language with your own agent.

Key Notes

  • Griffain agents can autonomously execute on-chain transactions on DEX set beforehand, such as DCA every hour or Limit Orders to buy or sell. Or get emailed up-to-date reports about what tokens any address owns and what changes they’ve made to their portfolio
  • The project is publicly supported by @aeyakovenko and Solana Labs. They are aiming to build Griffain on Solana Mobile the same as Siri on iPhone.
  • Griffain is built by @tonyplasencia3, the previous builder of Underdog and Blink on Solana.

This project to be interesting.

I’m trying to look at where the puck is going. Everyone keeps criticizing agents as “chatbots” and lack utility. Griffain is setting up the infrastructure for agents that can be useful onchain.

REI ($153mc) - A Smarter Take on AI Agents

At its core, REI isn’t just another Ai agent bot. It’s a full-stack framework combining AI innovation with blockchain.

Think of it as AI that doesn’t just act but learns, reasons, and adapts

What does it do? REI primarily interacts with users through X, where her cognitive architecture enables natural, context-aware interactions.

How does she work? Her Oracle Bridge connects off-chain data, while ERCData Standard structures it for gas-efficient on-chain use.

Together with REI’s 4-Layer Cognitive Model (Thinking, Reasoning, Deciding, Acting), this allows her to learn, adapt, and deliver deterministic execution.

Recent Developments shaping REI’s potential:

  • AI Stack SDKs: Rolling out in phases, making agent creation more accessible for devs.
  • Expanding Oracle Network: Critical for building real-world AI applications on-chain.
  • Phase 2 Catalysts: Full testnet launch, advanced validation protocols, and mainnet plans in 2025.

What’s the big picture? REI isn’t just about smarter bots. It’s shaping decentralized intelligence by combining AI x blockchain precision.

If you’re looking for the future of autonomous agents, REI is a framework worth watching.

$VAPOR - (MC: $154m) - First and Only AI Agent Launchpad on Hyperliquid

Vapor is a native AI Agent launchpad on Hyperliquid, and was built using the @ai16zdao Eliza framework.

Key Notes

  • While $AI16Z is about 1.2% of Solana’s FDV, $VAPOR is sitting at $132m, equivalent to about just 0.47% of Hyperliquid’s FDV.
  • Once live, users have to burn $VAPOR to launch an agent. This means the $VAPOR is deflationary and the total supply will keep decreasing.
  • All agent tokens launched on Hyper EVM will be tradable in a native DEX pool paired with $VAPOR. As more agents are launched, more $VAPOR will be locked in liquidity pools.
  • It’s already live on testnet. All agents are now approved without token gated.

MODE ($114mc) - Where A.I. Agents Meet DeFi

You can think of Mode Network as an ETH Layer 2 that focuses on scaling DeFi to billions of users through on-chain AI Agents.

“by 2026, more than 80% of on-chain TX will be via AI agents”. If you believe this to be the case, then Mode will get you exposure.

  • $500M+ TVL: The #3 chain in the OP Superchain ecosystem.
  • User Growth: 367k users with 24M+ transactions.
  • Undervalued: MC/TVL ratio signals potential upside.

What’s happening on Mode?

  • Synth Testnet launch: A Bittensor subnet for financial forecasting, unlocking DeFi intelligence.
  • AI Agent Hackathon: First-ever hackathon concluded with 667 registrations and 10+ Eliza plugins in just six days.

Developments on AI Agent App Store. Currently hosts few live agents, including:

  • Amplifi: Bitcoin Abstraction Layer. Scaling Bitcoin To Billions With AIFi.
  • Arma: Optimizes USDC yield farming.
  • Brian: Enables natural language execution for transactions and contracts.
  • Sturdy: Aggregates AI-powered yield vaults for higher returns.
  • QuillAI Network: Security AI agents powered by D-LLM & EigenLayer AVS

Mode’s vision? Build the infrastructure for an on-chain agent economy. With the Synth Testnet, BitcoinOS integration, and growing AI agents

$GAME (@game_virtuals) - MC: $390m - The Virtuals Framework

G.A.M.E was the first infrastructure project that allowed developers to integrate AI Agents into games.

From Virtuals:

“Every society has core infrastructure crucial to its growth and development. @game_virtuals plays that role as the Shopify for Virtuals, with @base acting as SWIFT.”

Notable Traction

  • Developed by Virtuals Protocol team themselves.
  • $GAME is the backbone of the Virtual economy. Whenever an AI agent operates, it effectively pays a fee in $GAME to keep the ecosystem running.
  • G.A.M.E already empowers more than 200 projects. It leads to more than 150,000 requests being sent on G.A.M.E every day.
  • Top #1 in 30D performance within the Virtuals ecosystem with nearly 400% growth.
  • G.A.M.E. has the highest Lifetime Inferences on Virtuals Protocol now at over 2.2m inferences.

G.A.M.E. still feels undervalued because people don’t really know what it does. And the G.A.M.E. team themselves don’t really shill it

(there’s like 5 tweets total from the official account. The bros haven’t even updated the twitter header lol)

A similiar and higher risk play is CONVO which powers the conversations on Virtuals.

Soul Graph (@soulgra_ph"">@soulgra_ph) $GRPH - ($36mc) - Making Agents more Human Like

Imagine if other frameworks shape the “hard” skill of the agents, Soul graph shapes its ‘soft” skill and makes it more human. Its framework aims to add personality to new or existing AI Agents.

“we’re building open source tools & provide infra that let developers give their agents an evolving personality, persistent memory and real time comms”

Key Notes

  • It already crossed 2,000 souls created and 15,000 voice-to-voice minutes served in soulgraph playground.
  • Soulgraph has recently integrated with AI16Z plugins.
  • 85% total supply is allocated to the community. Dev allocation is locked and vesting for 1 year.

I think this is going to be big.

Thesis: Character ai is huge. Normies log into that website and spend hours talking to A.I. text. Google just bought them for $2.7B). So imagine if people could talk to A.I. agents in real-time. And these agents have human like personalities / memories.

That’s what Soul Graph does. Go to the website and see for yourself.

Others

This article’s getting quite long, and I’m unable to cover every single protocol. Here are a few other protocols to pay attention to.

• Bully (@dolos_diary) - Started as a roasting account. It now features @dolion_ai which is a framework for social media a.i. agents.

• TopHat (@tophat_one) - Solana agent launchpad.

• Swarm Node @Swarmnode - Think of it like AWS Lambda, but specifically for AI agents and frameworks.

• Cookie Fun @cookiedotfun - it’s like the DeFiLlama / Kaito of A.I. agents - it tracks data such as mindshare & smart engagement / marketcap Advanced features are token gated.

Which to Choose?

First, everyone should have exposure to both Virtuals and AI16z. They’re clearly dominating and I don’t see signs of that slowing down. Don’t underestimate network effects.

If you’re considering some beta plays, here’s what I’m looking at:

• LaunchPad on a new Hot Chain. This is a playbook straight from 2021. Virtuals dominates Base. So, bet on the LaunchPad for other hot chains. This is why Vapor has skyrocketed in price because Hyperliquid’s so hot right now.

• Is it Niche’ing down? It’s hard to launch a general purpose Layer 1 or Layer 2 now. Instead, you gotta focus on something like being the “gaming chain.” This is why I find Mode interesting as it’s laser focusing on A.I. agents x DeFi, or how Griffain is focused on natural language.

• Programming Language. Infrastructure projects need agents. Agents need developers. Developers have their preferred programming language. This is where Zerebro (Python) and Arc (Rust) Shine.

A.I. Portfolio Ideas

  1. The Chill Portfolio - 50% Virtuals and 50% ai16z. This is simple and will probably outperform 80% of people.
  2. Alpha / Beta - Virtuals as the alpha play, and you pick a competing framework as a beta such as Arc or Griffain
  3. The Barbell - Virtuals and ai16z as your “safe” bets. Then you go for higher risk play with various A.I. agents.
  4. A little of Everything - Infrastructure Alpha Plays and some beta plays. Then you throw in various A.I. agents across the risk curve. Some alpha ones like @aixbt, and go into some sub $10mc trenches.

Criticisms of A.I. Agent Sector

“Too many launchpads and frameworks” - Bros, do you see how many layer 1s and and layer 2s there are? We’re doing fine. This entire space is new and moving fast. If you want to build a new city, it makes sense to build the roads, bridges, and pipes first.

“Most of this are garbage and will go to zero” - This is crypto. Most things go to shit, but the stuff that survive changes the game.

“A.I. sector has topped” - you’re not gonna make it. Don’t midcurve it.

“All the a.i. agents suck” - there was a time when people were too scare to order things online. I also remember when the first iPhones couldn’t copy and paste. We’re early. This is a bet on the future.

Final Thoughts

I knew there would be a bull run this cycle. What I wasn’t sure of his what new game-changing primitive would be introduced.

What would be the “DeFi” or “NFT” moment for this cycle. The one sector that would re-invigorate all the developers. The sector that would make retail go “ah ha” and ape in.

I wasn’t sure if there was going to be one. But the good news is…it’s here.

This is your chance to make it - don’t fuck it up.

If you enjoyed this article, please help me by engaging with this. Took a ton of time to research and write.

Disclaimer: Nothing here is sponsored. I have no free allocation to anything. I have exposure to several protocols listed here.

Disclaimer:

  1. This article is reprinted from [X]. All copyrights belong to the original author [@thedefiedge]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
  2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
  3. The Gate Learn team does translations of the article into other languages. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.
Start Now
Sign up and get a
$100
Voucher!