Over the last 30 days, Hyperliquid processed $234.3 billion in notional volume across 904 markets, filling 144.1 million trades from 242.8K unique trader addresses. Those are big numbers, but Hyperliquid prints big numbers every month. What made this month different is the shape of the volume, not the size of it. Pull the raw trade data and the tape splits cleanly into three lanes: core crypto perps, builder-deployed HIP-3 markets tracking real-world assets, and a brand new class of HIP-4 outcome markets. For the first time, more than a quarter of everything traded on Hyperliquid was not a crypto perp.
Every stat in this post comes straight from onchain trade data, pulled with Quicknode SQL Explorer, which indexes over 500 billion rows of Hyperliquid history across 37 tables and makes all of it queryable in plain SQL. No indexers, no pipelines, no waiting on a dashboard to add the metric you want. Here is what the month looked like.
Hyperliquid became a three-lane market
Core crypto perps are still the depth engine of the venue. They did $170.6 billion over the window, 72.8% of all volume, and on the biggest days in early June the core tape alone cleared $13 billion. But the second lane is no longer a rounding error. HIP-3 builder-deployed perps did $63.8 billion across 158 markets, which works out to 27.2% of everything traded on Hyperliquid. These are markets like crude oil, the S&P 500, silver, and single stocks, listed permissionlessly by builders and settled on the same margin system as everything else.
The third lane is HIP-4. Outcome markets did $37.1 million across 287 outcome legs, which is small in dollar terms but represents an entirely new product category trading inside the same account and the same execution venue. A year ago this chart would have been one solid color. Today it is a stack.
Run this yourself: the segment split above is a single GROUP BY over the hyperliquid_trades table. SQL Explorer ships with 40+ pre-built Hyperliquid queries covering trades, fills, funding, liquidations, and positions, so you can reproduce every chart in this post without writing a line of infrastructure.
A few huge markets led the HIP-3 tape
The honest version of the HIP-3 story is that volume was heavily concentrated. Not every RWA market worked, and most of the 158 builder markets are still small. What matters is that a handful became serious enough to sit beside crypto perps on the leaderboard. CL, the crude oil perp, was the single largest HIP-3 market of the month at $12.1 billion. Index exposure followed close behind, with XYZ100 at $7.6 billion and SP500 at $7.5 billion trading around the clock, including hours when the underlying markets are closed. Commodities filled out the top five, with BRENTOIL at $6.1 billion and SILVER at $4.9 billion.
Together, those top five markets did roughly $38.2 billion, about 60% of all builder perp volume. Behind them sits a long tail of single-stock perps that would have sounded absurd a year ago: MU did $3.8 billion, SNDK $2.3 billion, NVDA $2.2 billion, with GOLD, INTC, MRVL, and CRCL all printing nine to ten figures. Indices, energy, and metals are doing the heavy lifting, and equities are coming up fast underneath.
SPCX turned pre-IPO speculation into a 24/7 perp tape
The cleanest single-market story of the month is SPCX, a perp offering exposure to one of the most demanded pre-IPO names in the world. From its first observed trade on May 17 at 22:33 UTC, SPCX did $397.0 million over 789.2K fills in under three weeks. The market opened with a launch print of $179.90, ran to a period high of $230.00, and slid to a latest price of $164.04, down 8.8% from launch and 28.7% from the high. Along the way the market produced 12.1K liquidations totaling $15.4 million, which is what real leverage meeting a real two-sided market looks like. This is price discovery for a private company, happening in public, onchain, around the clock.
SPCX was not alone. The window saw a wave of new HIP-3 listings find immediate volume from a standing start: BB did $182.3 million, ARM $173.3 million, PURRDAT $160.0 million, and QNT $92.0 million, all within days of launching.
Catch the next SPCX from its first fill: Quicknode’s Hyperliquid infrastructure streams HyperCore trades, order books, and liquidations in real time over gRPC and WebSocket, and SQL Explorer lets you backfill the full history the moment a market launches. The Hyperliquid docs cover both layers, HyperCore and HyperEVM, through a single endpoint.
HIP-4 outcome markets traded in bursts
HIP-4 is the newest lane and the smallest, so the right framing is activity rather than dominance. Outcome markets did $37.1 million across 621.2K trades and 287 outcome legs in the window, trading in bursts around event flow rather than in a steady stream. The more telling signal is breadth. Active outcome legs held steady at 17 through mid-May, then nearly doubled to around 31 by early June as new event markets came online. Event risk is starting to get pulled into the same venue as everything else, and the surface area is expanding faster than the volume.
What it means
Hyperliquid’s moat is widening horizontally. Core crypto perps still supply the depth, but HIP-3 lets builders list equities, commodities, indices, and pre-IPO exposure, while HIP-4 starts pulling event risk into the same execution venue. The result is one account, one margin system, and a much broader set of things to trade. And because all of it settles onchain, every number in this post is reproducible by anyone with a SQL query.
Build on the same data
If you trade these markets on MEXC today, Hyperliquid’s onchain tape gives you a second lens on price discovery, liquidation flow, and volume concentration. You can see exactly where leverage is building up, which RWA and pre-IPO names are pulling real volume versus just headlines, and when an outcome market starts heating up, all from the raw fills rather than a smoothed dashboard. It is the kind of read that usually lives behind a desk’s data team, available to anyone willing to write a query.
Everything in this post took one afternoon and zero infrastructure. SQL Explorer gives you direct SQL access to 500B+ indexed rows of Hyperliquid trades, orders, fills, funding, liquidations, and staking data, in the browser or through a REST API your systems can consume directly. It is live in beta on all paid Quicknode plans.
If you want to go deeper: start with the SQL Explorer docs, clone the open-source SQL Explorer Cookbook to ship a dashboard in minutes, or browse the Hyperliquid developer guides for tutorials on portfolio trackers, oracle prices, and more. Create a free Quicknode account, and you can be querying the same tape this post was written from in under five minutes.
Methodology and data notes
All figures were queried with Quicknode SQL Explorer on the hyperliquid-core-mainnet cluster, primarily from the hyperliquid_trades and hyperliquid_liquidations_hourly tables. The window runs from 2026-05-06 15:17 UTC to 2026-06-05 15:17 UTC, so the first and last calendar days are partial. HIP-3 markets are classified by builder coin prefixes (xyz:, cash:, tradexyz:, vntl:, km:, hyna:, flx:, fix:), and HIP-4 outcome legs are coins beginning with #. Unique traders are deduplicated addresses appearing as buyer or seller; per-market participant figures count buyer and seller sides separately.