Afx Reaches $1.3 Billion In Perpetual Volume Through A Self-Funded Model

What Happened

AFX is one such competitor, recording more than $1.3 billion in cumulative perpetual volume within a little over a month of its mainnet launch, according to Head of Growth Ken C. Its ALP liquidity vault attracted more than $21 million over the same period, providing an early indication of demand across its crypto and traditional asset markets.

Market Context

Perpetual futures have become one of crypto’s busiest on-chain trading categories, with the top 10 perp DEXs recording $6.7 trillion in volume during 2025. This growth has intensified competition among exchanges seeking faster execution, deeper liquidity, and a greater share of derivatives activity.

In an interview with BeInCrypto, Ken discussed AFX’s sovereign blockchain, its plans for AI-powered trading, and a self-funded token model centred on community participation.

A Sovereign L1 Built Around Perpetual Trading

“The order book, matching, the settlement, the margin and the risk systems are all designed around trading from day one,” Ken said.

A sovereign L1 gives AFX control over execution, settlement and risk management across its perpetual markets. Ken linked this choice to the demands of leveraged trading, where congestion can alter an order at the moment a trader enters or exits.

AFX reports finalization of around 100 milliseconds, which Ken compared with a human blink lasting between 100 and 400 milliseconds. Even a small delay can affect entry prices, exits, liquidation risk and automated strategy performance.

Preparing AFX for AI Agent Trading

The exchange plans an AI agent trading competition during the second season of its points programme. Ken also compared upcoming event-led tools with “Bloomberg Terminal plus event-led AI agent trading.”

Users can review which assets an agent traded, when it entered or exited, which prices it received and how it behaved during volatile periods. These records also show whether an agent followed its assigned strategy and remained within its risk limits.

“If you are trusting an AI agent with capital, you need more than just a screenshot. You need a verifiable track record,” he said.

Liquidity Supports Markets Beyond Crypto

“Listing more markets is easy. Making them tradable is the hard part,” Ken said.

AFX offers perpetual exposure across crypto and several traditional asset categories, including stocks and commodities. The ALP vault provides base and backstop liquidity, and held more than $21 million at the time of the interview.

AFX pairs vault liquidity with market-specific open interest caps and a multi-stage liquidation engine. These controls support trading during sharp price movements and reduce exposure to manipulation.

Why It Matters

“Your chain could be congested because someone is minting NFTs,” he said, describing how unrelated activity on a shared network can affect a Bitcoin trade.

“Slow execution can negatively impact your trade performance,” he said. Consistent finalization gives human traders and algorithmic systems a more predictable environment during volatile periods.

AFX is developing native APIs and SDKs for automated strategies, while scoped wallets and symbol-specific subaccounts define access. Risk limits, kill switches and test environments give users control over agent activity.

“The first users are active traders, but over time we expect more professional traders and AI agents to come on board,” Ken said.

Details

“We want to be known as the platform built for traders and tuned for AI agents,” Ken said.

“Predictable execution is everything,” Ken said. AFX uses a dedicated fair-ordering mempool to reduce front-running and sandwich attacks, helping automated systems receive consistent treatment when they submit orders under comparable conditions.

Giving Agents Verifiable Performance Histories

“Every order, fill and liquidation can be verified,” Ken said.

On-chain histories allow strategy developers to demonstrate performance through recorded execution, while users can compare agents using the same underlying data.