Hyperliquid Was a Preview

Hyperliquid spent three years building the ground floor before it could focus on the exchange. The next team won't have to.

Hyperliquid, a single app running on its own blockchain, has an FDV of about $67 billion. Every Layer 2 token CoinGecko tracks, more than a hundred of them, adds up to about $25 billion. One app that owns its chain is worth more than two and a half times an entire category of networks built to host apps like it. (CoinGecko: Hyperliquid, CoinGecko: Layer 2)

Monolith supporters tend to write Hyperliquid off as a one-off: a very successful app that makes a lot of money, so of course it is valuable. That is fair. Hyperliquid is a story about an app that made it. But look closely at how it made it, and a large share of the work its team did in those first years had nothing to do with the app people use. It went into building a blockchain underneath it. That part of the story is the one that has changed.

Three Years on the Ground Floor

Hyperliquid's team began working on onchain trading in the summer of 2022. They wanted a fully onchain order book, the same kind of exchange professional traders use, with every order, cancel, and liquidation recorded on chain. No existing blockchain could handle that load, so they set out to build their own. (ASXN)

The first version went live as a closed alpha in early 2023, running on Tendermint, an off-the-shelf consensus engine. It worked, but it became the bottleneck, topping out at around 20,000 orders per second. So the team wrote a consensus protocol of their own from scratch, HyperBFT, and moved the whole network onto it in May 2024. Alongside it they built the order book and clearinghouse into the chain itself, built price oracles into consensus, and built a bridge secured by their own validators so users could move money in. For most of its life, deposits still ran through a bridge on Arbitrum, an Ethereum Layer 2, and it took until February 2025 to ship HyperEVM, a general-purpose smart contract layer, and until late 2025 to support native USDC directly. (Bitget, ASXN, Yahoo Finance)

All of that was done by a remarkably small team, reported at around five engineers and ten to eleven people in total, funded out of the founders' own trading profits, with their trading business put into maintenance mode to keep the lights on. Jeff Yan has described the early days simply: "it was just an idea and a small team grinding day and night." (ASXN, TechFlow)

That is roughly three years of a small, exceptional team working around the clock, much of it on infrastructure no trader ever sees: consensus, networking, validators, bridges, and the plumbing to make it all fast enough. They did it because there was no other way to get what they wanted.

What All That Work Bought

What they wanted was control. Because Hyperliquid owns its chain, its order book runs as part of the network itself rather than as one contract competing for blockspace. Its fees stay with the protocol instead of leaving as gas to someone else's validators, which is why it can direct nearly all of them to buying back its own token. And it ships changes on its own schedule. That control is what turned a great exchange into one that keeps the value it creates. As of November 10, 2026, Hyperliquid held about 56% of onchain perpetuals open interest and 44% of daily volume, and in the last 30 days it collected about $61 million in fees. (Castle Labs, DefiLlama)

The lesson most people take from Hyperliquid is that you need an extraordinary team to pull this off. The more useful lesson is that an extraordinary team was needed mostly because the ground floor did not exist yet. Remove the need to build it, and the bar drops to where it should be: building a great app.

The Ground Floor Now Exists

Canopy is that ground floor, and it is worth being specific about how it replaces each piece Hyperliquid had to build.

Start with the chain itself. Hyperliquid's first problem was that no existing network could host its order book, so it had to build a new blockchain before it could build an exchange. On Canopy, every app is already its own blockchain. A team starts from Canopy's open-source template, which ships a complete node: block production, transaction handling, peer-to-peer networking, and storage. The team writes only the app's own logic, the rules for what transactions do and how state changes, as a plugin in Go, TypeScript, Python, Kotlin, or C#. Roughly 200 lines are enough to define a working app. Everything else a chain needs is already there.

Consensus is the next piece, and it is where Hyperliquid spent some of its hardest engineering. The team launched on Tendermint, outgrew it, and then wrote HyperBFT from scratch, a consensus protocol based on HotStuff. Canopy apps come with NestBFT, which is also built on the HotStuff family of protocols, and adds stake-weighted leader election that resists manipulation and denial-of-service attacks, signature aggregation that keeps messages small, immediate finality, and built-in protection against attempts to rewrite old history. A Canopy team inherits that work rather than repeating it. That does not mean every app will match Hyperliquid's throughput, which depends on the app and its validators, but no team has to design and maintain a consensus protocol of its own.

Security is where the gap is widest. A new chain is only as safe as the stake behind it, which has traditionally left teams with two bad options: raise a large amount of money to buy a decentralized validator set before having users, or start with a few trusted operators and decentralize later. Hyperliquid took the second path, and its concentrated validator set drew criticism as a result. On Canopy, an app is secured from its first block by a committee of validators who already secure the Canopy network. They join an app's committee by adding it to their existing stake in a single transaction, with no new keys and no new collateral, and their entire bond can be slashed for misbehaving in that app. Apps can reward that committee in their own token from the start, and apps that attract enough support also share in Canopy's protocol rewards. If an app's validators go offline, other Canopy validators can step in through ordinary transactions on a network that keeps running. When an app is ready, it can graduate to its own validator set with its full history intact.

Moving money in and out was another long project for Hyperliquid, which relied on a bridge on Arbitrum for most of its history. Canopy includes token swaps that let users exchange assets with external chains such as Bitcoin and Ethereum without those chains changing anything. A committee of Canopy validators watches both sides of a trade, reaches agreement on what happened, and releases escrowed funds only once the other side has paid.

Finally, there is the question of paying for all of it. Hyperliquid funded years of development from its founders' trading profits. On Canopy, a team can launch through Canopy Terminal, where its token trades on a bonding curve from day one and the app moves onto its own network with real liquidity and a base of holders once it reaches the graduation threshold. Upgrades after launch work the same way they did for Hyperliquid, on the team's own schedule, by shipping a new version of the app that activates at a set block height, without waiting on anyone else.

The result is that the part of Hyperliquid's journey that took a small team years can now be skipped entirely.

What's Left Is the App

None of this guarantees a hit. Plenty of apps that owned their chains never found their users, and dYdX is the obvious reminder that a great product still matters most. But that is exactly the point. When the technology was the barrier, only teams with rare expertise and deep pockets could even try. With that barrier gone, the next Hyperliquid can come from whoever builds the best app.

Where CNPY Fits

In this model CNPY is not the currency of any particular app. Each app has its own token and captures its own value. CNPY's role is the collateral that each app's committee is built from while Canopy secures it. More apps, protecting more value, need more of that collateral to be credibly secure, and the people who bond it also decide which apps receive protocol issuance.

The Next One

Hyperliquid showed what an app can become when it owns its chain. It took three years of building the ground floor to get there once. If you have an idea for the next one, you can start on the part that matters. The Canopy build guide is where to begin.

Nothing in this essay is financial advice.