What is Hyperliquid
Hyperliquid is a new Layer 1 blockchain purpose-built to provide a comprehensive foundation for onchain finance. With a native, ultra-performant perpetuals and spot exchange, a general purpose EVM environment, and a number of net new primitives, Hyperliquid has rapidly emerged as a strong competitor to both centralized exchanges and general-purpose Layer 1s. Hyperliquid has undeniably been among the most successful new network launches of the past few years – read on to learn what all the HYPE is about.
Hyperliquid Architecture
Hyperliquid’s execution layer is split between two top-level native components: HyperCore and HyperEVM. HyperCore houses native primitives that support an ultra-high-performance exchange including onchain orderbooks, native oracles, staking, and vaults while HyperEVM provides a general-purpose smart contract platform where developers can build applications that interact with the native primitives on HyperCore.
Importantly, both HyperCore and HyperEVM exist as a shared state and are secured by Hyperliquid’s unified consensus protocol, HyperBFT. This allows the two components to specialize in different use cases while retaining the ability to seamlessly interoperate with each other.
This novel architecture can be understood as a hybrid between an app-specific blockchain and a general purpose chain, in many ways offering the best of both worlds to end users and developers. HyperCore is able to provide a best-in-class onchain trading experience largely due to its integration into the protocol itself, while HyperEVM provides general purpose functionality that is fully interoperable with HyperCore’s native components.
This design leads to a whole that is greater than the sum of its parts. Central limit orderbook exchanges are among the most throughput intensive and latency sensitive applications in all of crypto. HyperCore’s ability to handle 200K orders per second with a 70ms block time means it can handle an unprecedented amount of trading activity onchain. On pure general purpose chains, the premier exchange has to share blockspace with all other smart contract applications built on the network. On Hyperliquid, the flagship exchange has its own execution lane that handles core exchange functionality natively. In addition to the inherent exchange performance benefits this offers as compared to building on top of general purpose smart contracts, it also leaves a larger portion of HyperEVM’s throughput for all other apps that rely on smart contract capabilities.
In essence, Hyperliquid’s hybrid design can be viewed as a novel approach to scaling a general purpose platform for finance just as much as it can be viewed as a way to maximize performance for the core application. While HyperEVM throughput is set low initially, it will increase over time – the most important part of the initial rollout was not to affect the performance of HyperCore components. Depending on the measurement used, the platform as a whole may be considered more scalable than many of the fastest general purpose L1s even though HyperEVMs current total throughput capacity is lower than its general purpose competitors today.
HyperCore
The Hyperliquid Exchange
By initially building the Layer 1 infrastructure around the flagship orderbook exchange, the Hyperliquid Exchange enjoys a number of advantages over those built on general purpose L1s.
Source: The Hyperliquid Exchange UI
HyperCore offers a CEX-comparable trading experience that provides an excellent UI/UX and can handle the needs of professional traders who have historically been largely limited by latency and liquidity requirements to centralized venues. This is in part thanks to opinionated design choices informed by the Hyperliquid team’s trading background and made possible by direct integration with the broader protocol.
One example is the fact that cancellation orders take precedence over market orders during HyperCore block construction. This may seem like a small detail, but it’s extremely important since it gives makers a very slight advantage when they need to update their quotes, providing protection against toxic flow from other high frequency traders that may look to exploit a latency advantage to “pick off” out of date pricing. As a result of this protection, market makers are able to quote tighter and thicker liquidity around the assets they trade, which ultimately provides a significantly better trading experience for end-users on the platform. As an interesting aside, Loracle and Stalequant (two longtime Hyperliquid market makers) believe that this design choice means that when using volume as a proxy for retail usage, Hyperliquid actually has an inherent handicap. Because centralized exchanges do not have this built-in market maker protection, a material amount of volume traded on most CEXes is high frequency traders arbitraging each other. Since Hyperliquid prevents this with a speed bump at the block construction level, there is far less of this arbitrage activity on Hyperliquid, which results in “understated” volume metrics.
The Vault Primitive
Hyperliquid’s Vault primitive allows end-users to contribute USDC to strategies managed by a third party (which can be discretionary or automated) and earn a share of the profits or losses. Initially the Vaults were built to enable end-users to provide liquidity directly to the market making strategies operating through the HLP vault to provide liquidity on the exchange. As outlined in an early blog post, the primary motivation behind doing so was to demonstrate to end-users that the core team was not building Hyperliquid for the purpose of profiting from market making strategies. After some consideration, the core team thought that users might find the feature useful and decided to make vault creation permissionless. While many vaults today are run by retail end users, vaults provide the technical foundation for a marketplace of end-user accessible professional trading strategies to develop on Hyperliquid over time.
Native Oracles
The Hyperliquid exchange relies on pricing data provided by each validator to compute funding rates for perpetual futures. Validators periodically publish spot oracle prices, taken as a weighted median of multiple exchange prices including major Centralized Exchanges and Hyperliquid itself. The final oracle price used is the stake-weighted median of each validator’s submitted oracle price. This approach eliminates risks posed by third party oracles by integrating oracle pricing with the validators securing the protocol itself.
HyperEVM
While HyperEVM is still in the very early stages of development at the time of writing, it has its own opinionated design choices intended to provide builders with a specialized environment for deploying financial applications that can interoperate with HyperCore’s native components.
Dual-Block Architecture
One such choice is HyperEVM’s dual-block architecture, which splits total HyperEVM throughput between small blocks that happen quickly and larger blocks that happen less frequently. The HyperEVM mempool is split into two independent mempools that source transactions for each of the two block types.
This design decouples block speed and block size, enabling incremental improvements to be made on each axis separately without the need to accept a forced tradeoff between the two. Users benefit from faster blocks with lower time to transaction confirmation, while builders benefit from larger blocks to more easily include large transactions like contract deployments.
Currently, fast blocks have a block time of 2 seconds with a 2M gas limit, while slow blocks have a 1 minute block time with a 30M gas limit. These are conservatively set initial configurations which are expected to be steadily increased over time.
Source: @Ren_gmi – note that in reality there are ~30 fast blocks between each slow block, so this image is not an exact representation of small vs large block output over time.
Precompiles
HyperEVM’s interoperability with HyperCore is achieved through precompiles, which are capable of reading from and writing to HyperCore state. This enables novel DeFi use cases that tap into both HyperCore’s perpetuals and spot exchanges and HyperEVMs general purpose smart contract layer.
For example, builders could create an Ethena-style stablecoin without any reliance on centralized exchanges as a hedging venue. Protocol Vaults can be tokenized, allowing vault deposits to be re-used as collateral and looped. Money Markets can offer higher LTV ratios than otherwise possible by hedging risk on Hypercore, perform liquidations directly into the spot orderbook, and send and receive assets directly to and from users’ perpetual or spot balances on HyperCore.
These are just a few of the things precompiles make possible, with many more use cases likely to emerge over time. In essence, precompiles effectively allow HyperEVM developers to build smart contracts on top of HyperCore’s central limit orderbook (h/t to the Ambit Finance / Hyperdrive team for this framing, we highly recommend reading their article on precompiles for even more in-depth coverage on the topic).
Ecosystem
Hyperliquid has a very strong pipeline of builders, many of whom are Hyperliquid-native and/or deploying their apps and products exclusively on Hyperliquid. We’ll highlight a few of these teams below, but note that this list is by no means exhaustive. For a complete list, we’d highly recommend taking a look at Hypurrcollective’s ecosystem page.
- Hyperdrive: HyperDrive is building a DeFi hub, offering a lending and borrowing solution, liquid staking, and the ability for users to tokenize HLP positions.
- Hyperlend: HyperLend is another Hyperliquid DeFi hub, supporting lend/borrow and tokenized HLP in addition to flash loans and liquid perpetual futures positions built with ERC-4626 vaults.
- Stakedhype: StakedHype allows users to deposit HYPE for stHYPE, a HYPE liquid staking token from the Thunderhead team. Thunderhead built the original Hyperliquid stats dashboard and has extensive experience offering liquid staking solutions in other ecosystems.
- Valantis: Valantis is building the stHYPE AMM in collaboration with StakedHype. The stHYPE AMM integrates directly with the stHYPE withdrawal mechanism, ensuring that the peg between stHYPE and HYPE is maintained even in high stress periods. AMM reserves are also lent out for extra rewards on Hyperlend, improving capital efficiency and enhancing rewards for LPs.
- Kinetiq: Kinetiq is also building an LST, kHYPE. kHYPE will use StakeHub (built in collaboration with Anthias Labs) to score validators based on their performance and automatically allocate stake across validators accordingly.
- Felix Protocol: Felix is an authorized Fork of Liquity V2 with specific adaptations for Hyperliquid. Users can mint feUSD, Felix’s native CDP stablecoin, against select assets to access liquidity without selling their core holdings. Felix takes security extremely seriously, as evidenced by a number of in-depth articles (1, 2) on their approach. Like Kinetiq, Felix is also building in close collaboration with Anthias Labs.
- HyperFlash: HyperFlash is building FlashHYPE, a HYPE LST that will receive additional rewards from a planned MEV client.
- Silhouette: Silhouette is building a private trading layer on top of Hyperliquid that allows traders to tap into native onchain liquidity privately.
- Unit: Unit is the asset tokenization layer on Hyperliquid, which today enables both spot BTC and spot ETH to be traded on Hypercore. The Unit team seems to be evaluating enabling equity trading on Hyperliquid as well.
Product Innovation
In addition to its novel protocol architecture, Hyperliquid also meaningfully differentiates with a number of integrated products and features that aim to provide users and builders with everything they need to use Hyperliquid as a platform for onchain finance.
Hyperliquidity Provider (HLP)
HyperCore’s aforementioned Vault primitive was initially designed to enable the HLP (Hyperliquidity Provider) vault. HLP has been one of the most successful new crypto products of the past few years in its own right – users can deposit USDC to act as an LP for a variety of market making strategies run by the Hyperliquid core team on the exchange.
By allowing end users to permissionlessly deposit and earn a pro-rata share of profits or losses, HLP has democratized access to sophisticated market making strategies and the ability to perform liquidations, both of which typically provide extremely strong risk-adjusted rewards and were previously inaccessible to the broader retail market. On top of the base strategy rewards, the vast majority of platform fees were redirected to HLP prior to HYPE TGE with zero platform fees taken on HLP performance.
Builder Codes
Builder codes enable third party developers to build on top of HyperCore, taking a fee on fills they send on behalf of users. This allows trading front ends, onramps, social trading apps, and more to tap into Hyperliquid as the backend for their application. Builder codes remove the need for upcoming exchanges to build their own infrastructure and bootstrap liquidity from market makers to compete with incumbents.
Conceptually, builder codes work similarly to broker programs on centralized exchanges, but with the critical difference that the implementation is transparent, open, and permissionless. The vision of builder codes is to enable Hyperliquid to become the AWS of liquidity infrastructure – just as AWS abstracted away the complexity of managing cloud infrastructure for companies who required it, Hyperliquid aims to abstract away the need for trading applications to build and maintain complex exchange backends. This allows builders to focus on their product’s unique differentiation and outsource the backend to Hyperliquid.
Regional CEXes, wallets, and more can offer perpetual futures trading without having to worry about day one liquidity, listings, and building an entire exchange backend. Builder codes allow anyone to tap into Hyperliquid’s active, liquid orderbook, enabling front ends to focus on their core strengths and/or differentiating features. A few notable examples of apps leveraging builder codes today include pvp.trade, Okto Wallet, Dexari, and Nest.
Token Launch Support
In addition to providing an exchange that can compete with centralized offerings, Hyperliquid also provides an alternative solution for new projects looking to launch their tokens.
HIP-1: Native Token Standard
Hyperliquid’s HIP-1 token standard allows projects to permissionlessly create capped-supply spot assets. The pricing to launch a new token is transparent and based on a dutch auction model, where deployment gas decreases linearly from an initial price to 10,000 USDC. If the prior auction fails to complete, the initial gas price is 10,000 USDC. Otherwise, the initial price is set to 2 times the last price paid to mint a token. In the future, Ticker auctions will be denominated in HYPE. You can find the most recent tickers sold (and the price deployers paid for them) on Hypurrscan.
HIP-2: Hyperliquidity
Where HIP-1 provides a token standard for spot assets issued on Hyperliquid, HIP-2 (also known as Hyperliquidity) provides a simple mechanism for projects to seed newly-issued assets with liquidity. Inspired by Uniswap, Hyperliquidity enables deployers to define and fund a passive liquidity provision strategy secured by Hyperliquid’s underlying consensus to bootstrap liquidity for newly-issued assets.
By simply defining several parameters including the start price of an LP range, the number of orders in the range, the size of each order in the range, and the number of levels that begin as bids instead of asks, token deployers can launch native spot assets with a passive liquidity provision strategy that guarantees a 0.3% spread every 3 seconds.
Notably, Hyperliquidity is just one automated participant in a general purpose orderbook. This means that unlike on AMM platforms, active liquidity providers can quote alongside Hyperliquidity, allowing markets to adapt as assets mature and demand for liquidity grows.
These two features provide upstart projects with everything they need to launch a token and seed it with initial liquidity, all without the need to make any special deals with exchanges or third party market makers. Hyperliquid’s consensus itself will provide liquidity around a new token pair, and the only cost to the end user is the price paid to acquire their ticker of choice and any bids they’d like to seed for their token at launch.
Usage Metrics
With ~$1.1 trillion in perpetual futures volume settled, 400K unique addresses, and tens of thousands of active addresses each day, Hyperliquid has undeniably won over a very significant portion of real users who trade onchain (and many who previously did not). It has steadily become the most dominant onchain perpetual futures protocol by a significant margin, with over 6 times as much volume as its nearest competitor at the time of writing. Of particular note is the fact that usage grew considerably after Hyperliquid’s TGE, which stands in stark contrast with most protocols that distribute pre-TGE incentives, yet another indisputable indication of genuine end-user traction.
This sticky user base speaks to the fact that Hyperliquid’s product is not only best-in-class for DeFi, but a real competitor with other top tier trading venues, namely centralized exchanges. Pictured below are HypeFlows’ dashboards tracking Hyperliquid perpetual volume and Open Interest relative to centralized exchanges, both in terms of aggregate market share and compared to leading individual centralized exchanges.
Tokenomics
The HYPE token plays a number of important roles on Hyperliquid, including acting as the native staking asset for validators and delegators. HYPE holders can delegate their stake to validators of their choice to earn a ~2.3% SRR (Staking Rewards Rate) at the time of writing. HYPE is also the native gas token on HyperEVM, and thus is required to pay HyperEVM fees.
It’s also recently been announced that HYPE stakers will be eligible for trading fee discounts starting on or after April 30th based on the amount of HYPE they stake. In the future, HYPE will also be used for governance and to pay for ticker auctions – governance functionality has yet to be implemented and ticker auctions are still paid in USDC.
Since launch, it’s been very clear that HYPE is designed with value accrual to users and holders at top of mind. 100% of platform fees are directed to users and token holders, with some going to HLP depositors and the rest going to the Hyperliquid Assistance Fund, which today uses its portion of trading fees to buy HYPE on the open market. Since HYPE TGE the assistance fund has accumulated ~19.5M HYPE, worth nearly a quarter billion dollars at the time of writing.
The Hyperliquid Team
Hyperliquid is built by a very small but extremely talented team that comes from a professional trading background. Formerly known as Chameleon Trading, the now-Hyperliquid team initially started with proprietary market making strategies before expanding into DeFi trading. After using many of the prominent DeFi platforms at the time, they found many flaws with existing platforms and decided to build their own exchange to address these issues and provide an onchain trading experience that met the needs of professional traders. After FTX collapsed, the Hyperliquid team shifted all of their focus to building Hyperliquid. As Jeff Yan, Hyperliquid co-founder, describes in his interview on the When Shift Happens podcast, the Hyperliquid team viewed this as a critical juncture in crypto history after which the world was finally ready for DeFi.
The Hyperliquid team’s background in trading, understanding of the needs of professional market makers, and deep familiarity with blockchain design were all likely instrumental in their ability to build an exchange as advanced as Hyperliquid. In addition to enabling them to build a best-in-class product, this skillset also enabled them to build Hyperliquid without the need to raise any money by funding development of the exchange through continued market making operations.
Hyperliquid’s “Immaculate Conception”
Any discussion of Hyperliquid would be incomplete without mention of Hyperliquid’s initial token distribution. Since the Hyperliquid team self-funded development, they were able to distribute their token however they saw fit. At TGE, 31% of the total supply (and virtually 100% of the initial circulating supply) of HYPE was distributed to early users, with a further 38.888% held in reserve for future emissions and community rewards. 23.8% was set aside for current and future core contributors, with 0.3% allocated to community grants and 0.012% allocated to HIP-2 liquidity.
The Hyperliquid team is extremely committed to providing an even playing field. They do not pay for any integrations, nor do they provide special incentives for market makers. Market makers who traded on Hyperliquid prior to TGE were eligible for the genesis distribution, but they were subject to the same criteria as retail end users. Figment, as well as all other institutional validators, are competing on a level playing field with community validators for foundation support to enter the active set when it becomes permissionless.
This commitment to fairness has further set Hyperliquid apart from any current competition. As Jeff Yan described in a podcast with Blockworks’ 0xResearch, the Hyperliquid team’s vision is to build a blockchain to house all of finance. In Jeff’s view, the level of neutrality required of a platform that will achieve that goal is incompatible with having insiders at the base layer.
He draws a comparison to Bitcoin, suggesting that if Satoshi had taken any outside investment before launching Bitcoin it likely could not have been as successful as it is today. To that end, Hyperliquid’s approach to investment (or more accurately, lack thereof) has been built around replicating Bitcoin’s “immaculate conception”. Even for a highly-skilled and well-capitalized team like Hyperliquid, self-funding is an inherently more challenging path to take, but it is one that has won them an unprecedented amount of trust from their core userbase.
Recent Incidents and Responses
While Hyperliquid has been incredibly successful since inception, this level of success has inevitably made it a high-profile target for bad actors in the ecosystem. Over the past month there have been two noteworthy incidents that were seemingly attempts to attack the HLP vault in particular.
ETH: Margin Withdrawal
In the first incident, a trader opened a very large long position on ETH and was liquidated. HLP backstopped the liquidation, but due to the size of the position and the corresponding price impact, HLP ultimately took a $4M USD loss on the position. It’s not 100% clear if this user actually intended to attack HLP or they were just a user trading with very high leverage and large size.
The Hyperliquid team responded quickly to prevent the issue from re-occurring by adjusting requirements for users who wish to transfer margin out of open positions on the exchange. This mitigates the ability for a user to attack in this manner, since in order to be profitable in the attack they would need to move the mark price significantly from their entry. We recommend that the interested reader also refer to Jeff’s recent tweet outlining the rationale for this adjustment.
JELLY: Illiquid Self-Trade
Shortly after this initial incident, a more serious attack on HLP took place on a much less liquid token. In this case, a trader opened a $4M Hyperliquid short position on JELLY after having accumulated a large JELLY position on-chain during the prior week (seemingly in an attempt to corner the supply and orchestrate a short squeeze).
Shortly after opening the short position on JELLY, the trader purchased a significant amount of spot JELLY on-chain. The price impact of the spot purchase caused the traders short position on Hyperliquid to be liquidated, with HLP performing a backstop liquidation and taking over the position. Because the Liquidator component of HLP shared collateral with other component vaults in the strategy, auto-deleveraging failed to trigger when it otherwise should have.
This left HLP holding a very large short position on a very illiquid asset, which put the vault at significant risk. The Hyperliquid validator set quickly convened for a vote which ended in a unanimous decision to delist JELLY and settle the contract at the price at which the manipulation started (based on activity on both Hyperliquid itself and other venues) rather than the price at the time of delisting. The Hyperliquid Foundation subsequently re-imbursed any traders with open long positions as though the contracts had settled at the price at the time of delisting, which ensured that real users were made whole.
This was a somewhat controversial decision, with some arguing that JELLY contracts should have been settled at the price at the time of delisting. However, one may also argue that this approach wouldn’t provide a sufficient disincentive for any future attempts to attack HLP, which consists largely of end-user deposits. By making users whole, the platform settled all non-attackers at favorable prices while simultaneously demonstrating the validator set’s will and ability to coordinate to prevent and/or punish attempts at market manipulation on the platform.
In the aftermath of this incident, the Hyperliquid team has made a number of further updates to HLP, HyperCore itself, and validator governance capabilities to protect against similar threats in the future. HLP’s liquidator vault is now limited to a small percentage of HLP’s total account value and has more sophisticated logic around taking backstop liquidations. ADL will also now be automatically triggered if the liquidator loses more than a certain threshold, and open interest caps have also been redefined to be dynamic relative to an asset’s market cap. It’s worth noting that the Hyperliquid team does not expect these adjustments to ADL trigger conditions to result in ADL being triggered in normal market conditions.
In addition to these changes, validators are now able to vote onchain to delist assets that fall beneath certain thresholds. Stalequant, who was previously mentioned for his insights into Hyperliquid volume metrics, has put together a website that recommends adjustments to leverage and delisting for different assets on Hyperliquid based on a number of criteria. These include but are not limited to market cap, spot volume, futures volume, and Hyperliquid liquidity (e.g., HLP share of global volume and open interest, slippage, etc). After initially voting to delist MYRO, the validator set took action again based on Stalequant’s recommendations, voting to delist MYRO, RLV, HPOS, ORBS, BNT, NTRN, and ILV on April 4th.
Hyperliquid Outlook
As a platform that flawlessly handled one of the largest TGEs in crypto history, is the single largest holder of USDC on Arbitrum, and consistently settles billions of dollars of trading volume a day, Hyperliquid’s track record for security and robustness when factoring in usage, load, and attractiveness as a target for bad actors is still extremely impressive despite recent attack attempts.
Hyperliquid is a net new design, blending the best of app-specific and general purpose blockchains to provide a highly-optimized platform for finance. By building the Layer 1 architecture around an ultra-performant orderbook exchange, Hyperliquid is able to offer a best-in-class onchain trading experience that rivals the performance of centralized exchanges. The vast majority of high value activity onchain today is centered around trading. Centralized exchanges, in particular those that support perpetual futures, are among the most profitable businesses in crypto today.
By providing a CEX-comparable perpetual futures and spot trading venue, Hyperliquid is able to support the most valuable use case of crypto today fully onchain and internalize 100% of the value it generates. Today, this value accrues towards HYPE holders and HLP depositors via HYPE buybacks and vault accruals, respectively. The addition of general purpose smart contract capabilities secured by the same consensus underlying Hypercore provides a foundation for smart contract based development that is interoperable with the orderbook exchange itself. Finally, with support for creating native tokens and passively providing liquidity, Hyperliquid is able to enable the full lifecycle of DeFi protocols – builders can deploy apps to HyperEVM, launch their tokens, and trade them with deep liquidity and seamless UX on HyperCore.
While Hyperliquid may not be optimal today for non-finance-related high-throughput application classes (e.g., fully onchain gaming), this is clearly an intentional tradeoff made as part of an opinionated design – Hyperliquid is built to be a maximally performant and efficient layer 1 to facilitate onchain finance, and thus far has been built to accept the tradeoffs required to optimally meet the needs of its target user and developer base.
Figment’s Involvement
Figment has operated a validator on Hyperliquid testnet since December of 2024, and is actively pursuing foundation support to enter the mainnet active set once it becomes permissionless. If you’re interested in staking HYPE, integrating Hyperliquid into an existing application, or are building in the ecosystem, feel free to reach out to our team – Figment’s protocol team has extensive knowledge of the protocol and is eager to contribute to the development of the ecosystem.
