This is part of the 0xResearch newsletter. To read the full editions, Subscribe.
One of Hyperliquid’s core innovations is building codes. These tokens act as a protocol-level parameter in transaction payloads, allowing interfaces to append a generator address to capture automated charges on the chain. Originators can charge an additional fee of up to 100 basis points (1%) on spot and 10 basis points (0.1%) on criminality.
This separation of execution and settlement allows front-ends to monetize proprietary flow without the technical complexities of maintaining an order book or the capital inefficiency of providing liquidity. As shown below, third-party front-ends integrate Hyperliquid implementations and add their own variable fee levels on top, effectively creating a different pricing landscape for the same underlying implementation.
As such, building codes opened up the powerful flywheel of distribution. Nearly 40% of daily active users now trade through third-party front-ends rather than the native UI, a share that briefly topped 50% in late October. The top three builders, Based, Phantom, and pvp.trade, collectively took in more than $31 million in fees.
From a market structure perspective, this pushes Hyperliquid away from a fully integrated cryptocurrency exchange model and closer to a multi-layered brokerage of traditional stocks. On a centralized exchange like Binance, a single entity controls the entire stack across setup, routing, matching, and custody.
Hyperliquid’s design mimics the US stock market, where retail brokers (Robinhood, Schwab) own the relationship with the client and generate income from distribution, while they route orders to wholesale dealers (Citadel Securities, Virtu) who handle execution and settlement. In effect, the stack becomes two-level:
- A middleman-like distribution layer, where originators compete for order flow and differentiate between product and fee passing.
- A central execution location, where Hyperliquid centralizes liquidity and handles matching and margin.
Although this separation mechanism is new to cryptocurrency professionals, it has already been applied to Solana. Trading terminals like Photon and Axiom control user flow by focusing on the consumer layer. Photon grew primarily by being the fastest Solana memecoin sniper, while Axiom eventually challenged it with a broader product range and more aggressive points and discount designs. These terminals effectively served as building blocks: they sat on top of decentralized trading platforms, fixed their own fee spreads, and maintained accounting manually. Hyperliquid generation codes essentially turn this pattern into a primitive native protocol.
However, Solana’s example also highlights the risks. Trading terminals accounted for 77% of Solana’s DEX revenue over the past year, $633 million versus $188 million for DEXs, a 3.4x multiple that highlights that owning the front-end is often more valuable than owning the back-end. Specifically, is the front-end too valuable for Hyperliquid to give away?
The relationship between frontends and backends is rarely purely symbiotic. Frontends such as Jupiter aggregate different backends (Meteora, Raydium, Orca) and return the best path given the size, graph, and slippage constraints.
Source: Jupiter front end assembly example
This puts severe margin pressure on DEX backends. With no moat, it must be the cheapest way to win the flow. Since they do not own the user, the backends are also at risk of being replaced. We saw this when Pump.fun replaced Raydium as a liquidity backend with its own internal AMM, which significantly impacted Raydium’s share of volume.
Currently, Hyperliquid does not have this problem. By pioneering build codes on perps, the single build code environment has become powerful. However, if originators evolve from higher-HL UIs to true routers that can send flow to competing backends, they begin to resemble the intelligent order router in traditional finance. In this scenario, builders can:
- Overall cost optimization: Calculate spreads/slippage + patron/maker fees + additional builder fees – rebates + projected financing.
- Bargaining with venues: Ask for higher building quotas or discounts while threatening to direct flow elsewhere.
- Capturing the user relationship: While venues are forced to compete just to be the cheapest and best-executed wholesale liquidity provider.
Likewise, in traditional finance, wholesalers compete with broker-dealers for volume. Robinhood turns to Citadel Securities, Virtu, and Jane Street based on what offers the best order flow execution and payment.
While competing DEX platforms like Drift and Ostium have built-in generation tokens, none have emerged as a true contender yet. However, there is still a significant structural risk: If a place like Lighter ties builder rebates to a zero-fee model, it could theoretically allow wallets like Phantom and Rabby to bypass Hyperliquid’s 4.5 basis point fee. This would enable front-ends to capture the entire fee pool, effectively doubling their revenue per trade compared to Hyperliquid’s current model.
Liquid trade Serves as a leading indicator of this future. The Paradigm-backed terminal, which raised $7.6 million in its seed round, has facilitated $5.6 billion in trading volume on Hyperliquid. But more importantly, it also allows users to trade on Ostium and Lighter via the same interface. If major builders follow this path and begin actively routing flow based on place discounts rather than loyalty, Hyperliquid builder front-ends could evolve into a commoditized perp aggregator, directly threatening the protocol’s ability to capture value.
However, there is still a fundamental difference. Spot is easy to accumulate because each swap is atomic and the asset is fungible across places. One transaction equals one fill, and the router can seamlessly split a trade across multiple buckets. However, with perpetrators, attitudes are fixed and place-specific. The BTC-PERP position at Place A cannot be exchanged with the BTC-PERP position at Place B due to differences in index composition, funding rates, liquidation drivers and risk limits.
To usefully route criminals across locations, the market needs one of two difficult solutions:
- User segmentation: Users have to hold collateral in multiple places, which is capital inefficient and leads to poor user experience.
- Main brokerage layers: The router must act as a clearing layer, solving the difficult problems of credit extension, cross-margin, and clearing coordination.
While non-fungibility provides a short-term defence, the harsh reality is that front-ends are rational economic actors; They will migrate if a competitor offers higher profit margins. However, data suggests that this threat is currently contained. Despite the large number of users on third-party backends, the vast majority of the volume, over 90%, still originates from Hyperliquid’s native frontend.
Furthermore, the HYPE token adds a retention layer. Builders can hold HYPE to access fee discounts, allowing them to stack revenue streams: referrals, builder fees, and volume-based discounts. However, the cost of switching to increasingly better graphics may not be worth it for existing front-ends. Finally, the influx of builders appears to be additive rather than cannibalistic. These are new users entering the ecosystem via wallets and terminals, not users switching interfaces.
Therefore, while generation tokens provide an efficient scaling vector, expecting Hyperliquid to maintain complete dominance over its distribution layer is unrealistic. As the sector matures, Hyperliquid will have a harder time defending its lead against aggregators and lower-fee competitors. However, building a high-performance order book on blockchain is still a massive technical moat, and with front-end margins remaining healthy, incentives for builders to switch are low. However, in a rapidly expanding market, this is not a battle to retain scale, but a more competitive race for growth where Hyperliquid remains the heavyweight to beat.
Get news in your inbox. Explore Blockworks newsletters:
