Nearly 40% of Hyperliquid's daily active users now execute trades through third-party frontends rather than the protocol's native interface, with the top three builders — Based, Phantom and pvp.trade — collectively capturing more than $31 million in fees, per a 0xResearch analysis.

How do builder codes work?

Builder codes operate as a protocol-level parameter embedded in transaction payloads.

The codes let any interface append a builder address for on-chain fee capture, with surcharges capped at 100 basis points (1%) on spot trades and 10 basis points (0.1%) on perpetuals. The mechanism decouples execution from settlement, allowing frontends to monetize order flow without maintaining their own orderbook.

Third-party frontend share briefly exceeded 50% in late October before settling near the current 40% mark.

Does this reshape Hyperliquid's market structure?

The design pushes Hyperliquid away from a fully integrated exchange model. Centralized venues like Binance control the full stack across onboarding, routing, matching and custody.

Hyperliquid's architecture mirrors U.S. equities instead, where Robinhood and Charles Schwab own client relationships while wholesalers such as Citadel Securities and Virtu handle execution.

The protocol effectively runs two layers: a broker-style distribution tier where builders compete for order flow, and a central execution venue that concentrates liquidity and manages margining.

What risk does Solana's track record reveal?

Solana's memecoin terminals Photon and Axiom illustrate the danger. Photon grew first as the fastest memecoin sniper; Axiom responded with a broader product suite and aggressive rebate designs.

Together, terminals captured 77% of Solana's DEX revenue over the past year — $633 million versus $188 million for the underlying protocols, a 3.4x multiple. Pump.fun's decision to swap Raydium for an in-house AMM demonstrated how backends without user ownership can lose liquidity share.

Could Hyperliquid lose its role to aggregators?

Frontends have started resembling routers in the analysis, demanding better terms from backend venues. Cross-venue perp routing demands prime brokerage infrastructure, because a BTC-PERP on one venue is not fungible with one on another given differences in index composition, funding rates, liquidation engines and risk limits.

LiquidTrading, a Paradigm-backed terminal that raised $7.6 million in seed funding, has processed $5.6 billion in Hyperliquid volume while already routing trades to Ostium and Lighter through the same interface.

If a rival venue pairs builder rebates with a zero-fee model, wallets like Phantom and Rabby could route around Hyperliquid's 4.5 basis point charge and capture the entire fee stack.

What still defends Hyperliquid today?

More than 90% of Hyperliquid volume still originates from its native frontend, the analysis notes. HYPE staking lets builders stack fee discounts on top of revenue from referrals and builder surcharges.

Flow from third parties appears additive rather than cannibalistic, bringing new users through wallets and terminals rather than siphoning existing ones. The onchain orderbook itself remains a significant technical moat, and builder margins on perpetuals look healthy.

Still, the next phase is less a defense of existing volume and more a race for growth as competing venues compress fees and cross-venue clearing layers mature.