take note

There’s a strong possibility we won’t see a full-blown alt season where everything pumps together. The era of rising tides lifting all ships might be over because the landscape is too crowded—too many altcoins, not enough liquidity to go around. But while the broader market bleeds, a particular corner of crypto is breaking ATHs: A.I. Agents and A.I. Infrastructure. It’s reminiscent of how DeFi changed the game in 2020 or how alt L1s took over in 2021.

Therefore, if you’ve been sidelined, you’re probably eyeing these moves and wondering if it’s “too late.” I doubt it. The bleeding overall market obscures the fact that A.I. protocols are on a tear. This surge is less than three months old, and the agents themselves are still in rudimentary form—yet they’re already unlocking new utilities daily. Imagine how they’ll evolve when they can trade on-chain autonomously and optimize your portfolio.

Setting the Stage

• Altcoin Dispersion – There’s a sea of tokens, and the days of everything going 5x or 10x across the board are likely behind us.

• Focus on A.I. Infrastructure – This is the “picks and shovels” play in a gold rush. Instead of trying to pick the next big A.I. agent, why not own the platforms they’re built on?

The Two Kings of A.I. Infrastructure

1. Virtuals (MC: ~$4.6B) – The Current King

Virtuals is your go-to for launching A.I. Agents. It’s dominated the Base ecosystem and could branch out to other chains like Solana, HyperliquidX, and AbstractChain as early as 2025. The big question: Will they move toward A.I. agent revenue sharing?

Three Biggest Strengths

1. Tokenomics: You need $VIRTUALS for creation of new agents, and for liquidity pairing. That means any new agent that launches has to integrate $VIRTUALS.

2. Network Effect: All eyes are on Virtuals as the #1 platform to launch. That momentum is hard to replicate.

3. Flagship Agent: @AIXBT is Virtuals’ first major success story, surpassing $500m in market cap and showcasing how powerful the Virtuals ecosystem can be.

It’s not crazy to think Virtuals could enter crypto’s top 10. Their strategic approach outpaces most competing platforms that are still playing catch-up.

2. Ai16z (MC: ~$2.2B) – The Kingslayer?

Ai16z is home to the leading open-source A.I. agent framework, ElizaOS, which sits atop Github’s trending repos. Tokenomics have been a sore spot, but they’re overhauling it soon with a brand-new Launchpad in Q1 2025.

What’s Changing?

• Launch Fees: A chunk of the fees flows back to $AI16Z through buybacks & burns.

• Collateral Requirements: Projects need to hold $AI16Z and pair their tokens with it to launch, stabilizing liquidity pools and capturing value.

• Treasury Strategy: They plan on diversifying across major L1 tokens, stablecoins, and project tokens for a more sustainable runway.

• Agent Swarms: The next leap toward decentralized AGI—clusters of A.I. agents collaborating in real time.

Ai16z aspires to be the “ETH or SOL of A.I. Agents.” But Virtuals is already miles ahead in brand recognition and market share. Therefore, you don’t have to pick sides. There’s room for at least two big winners in this new frontier.

The Contenders

Beyond the “two kings,” several up-and-comers might offer higher returns if you can stomach more risk.

ZereBro (MC: ~$632m)

Originally a music-making agent, ZereBro pivoted into a full-stack Python-based agent framework, Zerepy. Python is crucial because:

1. It’s the world’s most popular language.

2. A significant chunk of cutting-edge A.I. tooling is Python-based.

They’re launching Zentients Agent Platform in early 2025, funneling fees and liquidity pairings back to $ZEREBRO. They’re also tied to Virtuals via a ZereBro/Virtual pool. But because it’s at a lower market cap, the upside potential is bigger—so is the risk.

Arc ($ARC) (MC: ~$358m)

The only A.I. agent framework built on Rust, known as Rig. Why care?

• Rust is Solana’s native language and is favored for high-performance blockchain infrastructure.

• Rig outperforms Langchain in memory usage and cold start times.

• It’s got recognized contributors (e.g., @0xCygaar from Abstract Chain).

Therefore, Arc is appealing if you believe Rust-based A.I. frameworks will excel.

Griffain ($GRIFFAIN) (MC: ~$457m)

An “A.I. App Store” on Solana. Griffain aims to let users do on-chain tasks via natural language commands through an A.I. agent—like placing limit orders, dollar-cost averaging, or generating real-time portfolio updates.

• Publicly supported by Solana Labs.

• Integrating into Solana Mobile (think “Siri for Solana”).

• Built by the same team behind Underdog and Blink.

Normies often dismiss A.I. as “just chatbots,” but once these agents can be used to quickly and easily interact with DeFi, that skepticism fades.

REI (MC: ~$153m)

A full-stack approach to bridging advanced A.I. reasoning with on-chain data. REI’s “Oracle Bridge” pulls in off-chain info, while a 4-Layer Cognitive Model handles thinking, reasoning, deciding, and acting in a structured, deterministic way.

• AI Stack SDKs: A phased rollout to make agent creation more accessible.

• Expanding Oracle Network: Key for real-world applications.

• Phase 2 Catalysts: Full testnet, advanced validation protocols, and mainnet in 2025.

Therefore, REI positions itself as a “smart agent” builder. If you’re looking for sophisticated, adaptive A.I. on-chain, keep an eye on REI.

Vapor ($VAPOR) (MC: ~$154m)

A native A.I. Agent Launchpad on Hyperliquid, built with Ai16z’s Eliza framework.

• Deflationary: You burn $VAPOR to launch an agent.

• Liquidity Pairings: Agents launched on Hyper EVM pair with $VAPOR, locking supply.

• Already testnet live; once mainnet picks up steam, supply dynamics could push price higher.

Mode (MC: ~$114m)

An ETH Layer 2 that focuses on A.I. Agents x DeFi. They’re betting that by 2026, over 80% of on-chain transactions will be initiated by A.I. agents.

• #3 chain in the OP Superchain with $500M+ TVL.

• Hosted an AI Agent Hackathon with ~667 registrations, building Eliza plugins in less than a week.

• They’re integrating BitcoinOS, bridging BTC into the A.I. mix.

Therefore, if the A.I. x DeFi narrative matures, Mode stands to benefit from massive user growth.

GAME (@game_virtuals) (MC: ~$390m)

The earliest gaming infrastructure project for Virtuals, described as “Shopify for Virtuals.”

• Developed by the Virtuals Protocol team.

• Fees in $GAME keep the ecosystem running.

• Already powering 200+ projects, with 150k daily requests.

It’s overshadowed by Virtuals, but has shown 400% growth in the last 30 days. It’s possible that $GAME remains undervalued simply because the team is too busy building to hype their project.

Soul Graph ($GRPH) (MC: ~$36m)

Focuses on the “soft” skills of an A.I. agent—personality, persistent memory, real-time conversation.

• 2,000+ “souls” created and 15k+ voice minutes served in their playground.

• Integrated with Ai16z’s Eliza plugins.

• 85% of supply goes to the community; dev vesting is locked for a year.

Character.ai recently sold for $2.7B, but that product is basically just chatbot entertainment. Soul Graph’s approach is a real-time personality layer for truly human-like A.I. agents.

Other Notables

• Bully (@dolos_diary) – Social-media-based agent frameworks.

• TopHat (@tophat_one) – A Solana A.I. agent launchpad.

• Swarm Node (@Swarmnode) – Think AWS Lambda for A.I. frameworks.

• Cookie Fun (@cookiedotfun) – Analytics aggregator for the A.I. agent sector.

Which to Choose?

1. Get Core Exposure: If you want it simple, grab Virtuals + Ai16z. Each is so integral to this sector that missing out on either seems risky.

2. Consider Beta Plays: If you believe in alternative programming languages or new chain expansions, look at Arc (Rust) or Vapor (Hyperliquid).

3. Allocate for Risk: A barbell strategy—go large on Virtuals/Ai16z, sprinkle the rest on smaller frameworks.

4. A Little of Everything: Virtuals, Ai16z, a handful of these mid-caps, plus some < $10m microcaps.

Criticisms of A.I. Agent Sector

• “Too many frameworks and launchpads!” – That’s normal in crypto’s early days. Eventually, only the best or most adaptive will survive.

• “Most will go to zero!” – Yes, but those that don’t can 100x. Risk is the name of the game.

• “A.I. sector has topped.” – The best time to enter is often during fear and skepticism.

• “All A.I. agents suck.” – Early iPhones didn’t even have copy/paste. Innovation can be messy. That’s exactly why it can be lucrative.

Final Thoughts

I wasn’t certain we’d get a brand-new, game-changing primitive this cycle—like DeFi was in 2020 or NFTs in 2021. But we got it: the rise of on-chain A.I. agents. It’s energizing both devs and retail, and it could become the hallmark narrative of this bull run.

Therefore, if you missed out on DeFi Summer or Alt L1 mania, here’s your chance. This sector is young, developing quickly, and still lightly capitalized compared to major L1s. Now is the time to decide if you’ll act or remain on the sidelines. Once the broader market rebounds, money and attention will flood in—and you’ll wish you had a piece of this new A.I. frontier.

Don’t overthink it. Don’t “be cute.” The Universe is giving you another shot. Will you take it?