“We’re not here to replace anyone. We’re here to empower them.” That was how Jeff, co-founder and CEO of Hyperliquid Labs, summed up Hyperliquid’s long-term goal in an interview on Empire, the Blockworks podcast.
The episode, hosted by BlockWorks research analyst Shaun Da Devens, aired on Oct. 7, 2026. Jeff used the conversation to lay out how Hyperliquid connects HIP-3, HIP-3 Star, HIP-4, portfolio margin, native lending, priority fees and its on-chain order book into a single roadmap.
From perpetuals to a base layer for all finance
Shaun opened by noting how much Hyperliquid has shipped this year: HIP-4 for outcomes-style payoff structures, portfolio margin, native lending built on top of it, priority fees, and HIP-3 Star running on testnet. Asked how those pieces fit together, Jeff said the answer is simple: Hyperliquid is trying to build the tooling that can eventually host all financial activity.
He said a vision by itself means little if builders cannot actually deploy what they want. If they cannot build on one system, they will build somewhere else. For that reason, Hyperliquid is not trying to stack feature after feature. The core idea, he said, is to refine a small number of primitives that combine cleanly with each other. Portfolio margin was his example because it lets lending and trading primitives, whether perpetuals or spot, lock together naturally.
Shaun said Hyperliquid often seems to ship something before the broader market understands why it matters. HIP-3, in his telling, was a case like that. During testnet, crypto-native users already saw that it could use price oracles to list any asset permissionlessly, but it took products such as Tradex.xyz and Shoku to show what strong execution on top of it looked like. Jeff said it is hard to predict how long that recognition takes, but he sees the same pattern across Hyperliquid’s primitives: each one may contain a larger story that has not yet been written. HIP-4, he said, is one example because options-like products can eventually be built and expanded on top of it.
Shaun also pointed to HIP-3’s growth in real-world asset open interest, saying it had gone from $300 million at the start of the year to $3.9 billion by the time of the interview, a 13x increase. He added that the market’s comparisons for Hyperliquid had shifted from AVO and DYDX toward CME and Nasdaq. Jeff said no analogy is perfect, largely because there is nothing else in the market doing exactly the same thing. To him, that is a good sign. He said Hyperliquid looks more like public infrastructure such as the internet or Linux than a product built for a crowded field of near-identical rivals. In that framing, the goal is a neutral system, not ten copies fighting one another.
HIP-3 Star adds optional controls rather than taking anything away
On HIP-3 Star, Shaun said the model allows deployers to list their own markets as HIP-3 does, while adding permission controls such as whitelists. He noted that Payward, Kraken’s parent company, has already said publicly that it wants to use that route to open Hyperliquid to U.S. users once it receives regulatory approval.
Jeff described HIP-3 Star as a distilled response to builder feedback. In his words, it takes an already solid base and adds optional switches. He said transparency, self-custody and redundancy are positive traits under any regulatory structure. HIP-3 Star, he argued, is not subtractive. It is additive because it leaves more choice with the institutions and individuals that want to build on top of Hyperliquid.
He repeated another theme from the interview as well: the split between decentralized finance and traditional finance does not really exist in his view. There is only finance. Hyperliquid started as a very high-performance layer-1 chain centered on perpetuals, but has since grown into spot, outcomes and native lending. Jeff said the roadmap still has work left on both performance and primitives. Performance will always be a challenge, he said, but that is a “high-class problem.” Every order-of-magnitude jump in usage unlocks something meaningful while introducing fresh engineering constraints. And the list of primitives, in his view, should never really end for a protocol that wants to support a global financial system.
Priority fees and the attempt to internalize the cost of latency races
Shaun called priority fees one of Hyperliquid’s more underappreciated recent updates. He broke the mechanism into two parts: read priority fees, which let traders stream unconfirmed order flow from the mempool, and write priority fees, which let an order attach HYPE to move ahead in line. So far, he said, about 180,000 HYPE has been burned through the design.
Jeff said he would not describe that model as bribery. He compared it instead to the way Ethereum evolved by making priority fees part of the protocol and routing them into burn mechanics. “Bribery,” he said, is a label from an older system, and in his view that system is toxic. The point here is to make markets clear more efficiently. If a trader brings real alpha, that edge should not be overwhelmed by somebody else’s advantage in microwave towers or transoceanic fiber.
That is why he sees priority fees as a way to internalize what would otherwise be zero-sum or even negative-sum externalities. Rather than forcing market access to depend on expensive infrastructure, the model is meant to let anyone trade on more equal footing with very little hardware or connectivity overhead.
Jeff defends public order books and says open data is the fairer model
The interview also turned to the current debate around privacy and transparency. Shaun said some market participants question whether blockchains should be fully open, and cited research comparing hidden versus publicly broadcast TWAP execution on Hyperliquid. The result, he said, was counterintuitive: public TWAP performed better because market makers had time to absorb that flow. He asked whether Hyperliquid should build a private layer on top.
Jeff said his view may be controversial. He went back to the pre-electronic era, when trading was shaped by who could shout louder and physically position themselves better in a pit. Electronic markets changed that by letting everyone see the full picture. In his argument, when everyone can see the order book, the order book itself becomes more liquid. He called that a clear win for openness.
He pushed the point further. An order book, he said, should not be run by a privileged centralized intermediary, and a globally neutral financial system should not leave that much power in one hand. Hyperliquid has kept its order book on-chain from day one, to the point where users can even see which address is behind each resting order. Jeff contrasted that with parts of traditional finance, mentioning Brazil’s market as an example where broker information is disclosed. If the data already exists, he said, then the real question is simply who gets access and who holds privilege. On that basis, making the data public is the fair outcome.
If he had to start over, Jeff says he would likely build options
Shaun then asked what Jeff would build if he switched roles and could no longer work on base-layer infrastructure. Jeff said he would look for products that do not yet exist at scale but have large upside. Options would probably be at the top of that list.
His reasoning was straightforward. Perpetuals satisfy a lot of demand, but some users want to express more complex market views. HIP-4, he said, creates an obvious path for building options protocols because it inherits the spot order book and portfolio margin framework, while also allowing delta hedging through perpetuals venues. In his view, convex payout products have long reached escape velocity in traditional finance, but they still have not truly taken off in crypto.
Shaun noted that HIP-4 is often framed from the outside as a way to bring prediction-market style products onto Hyperliquid, though it clearly goes beyond that. Jeff agreed and said the real value lies in composability inside the same ecosystem. Perpetuals, spot and outcomes together unlock more than any one of them can alone.
Spot, options and asset issuance are the next obvious pieces
Shaun referred back to Jeff’s earlier comparison between building Hyperliquid and playing Go: you cannot calculate every branch of the game, so you move ahead one, two or three steps at a time with conviction and instinct. Asked what the next obvious move is, Jeff answered plainly: options. To him, completing the trio of spot, perpetuals and options is the clearest next step.
At the same time, he said spot is a major unlock but harder to bootstrap because simply holding an asset is not itself economic activity. Tokenizing real-world assets is also more difficult than tokenizing only their price exposure through perpetuals. The ceiling for portfolio margin, in his words, depends on the quality of assets supported on-chain, which means a strong asset issuance layer is also necessary.
Jeff said that a year ago, HIP-3 barely existed in the conversation and now dominates it. In the same way, spot and options could become the focus a year from now.
Success, in Jeff’s view, is new things being built and actually used
Asked how he measures success as an infrastructure founder, Jeff said he leans more on qualitative standards than on raw metrics. What matters most to him is whether something genuinely new, something that did not previously exist, is being built and then used by people because they prefer it.
That, he said, is the purpose of infrastructure: to be used and to enable construction. He closed with a message to the community about staying humble and keeping the end user in mind rather than getting trapped in the idea of defeating opponents. Three years ago, he said, Hyperliquid was only an idea and a small group of people working around the clock. Growth, in his view, should not turn into arrogance.