Perpetuals Trading
Every trade, every funding payment and every resting order on perpetual venues, with each market classified by what it tracks and who launched it. Currently covering Hyperliquid HyperCore.
Tables in this collection
Frequently asked questions
What is Dune's Perpetuals Trading dataset?
The venue publishes its own transaction record. The work Dune does is turning it into something you can answer a question with. Every market is classified by what it tracks, who launched it and on what terms, and carries a written contract spec that tells two identically named markets apart. Volume is counted once per match rather than once per side, so the figure matches what the venue reports. Positions are rolled up per account per day, so profit and loss after fees and funding is one scan instead of a walk through billions of executions. The order book is reconstructed every minute, so spread and the cost of trading size come off a single row. Hyperliquid is the first venue covered.
What is a perpetual?
A perpetual is a derivative that tracks the spot price of an underlying asset, a token, a stock, a commodity, a currency or an index, and has no expiry date. You do not own the asset. You take a long or short position on its price, and you can keep that position open as long as you meet the margin requirement. Because the contract never settles, its price is kept close to spot by a funding rate: periodic payments exchanged between longs and shorts. When the perpetual trades above spot, longs pay shorts. When it trades below, shorts pay longs. That incentive is what anchors the contract to the asset it tracks. Perpetuals are the main trading instrument in crypto. The same structure is increasingly used to take positions on shares, metals, currencies and indices.
Which venues are covered?
Hyperliquid today, covering both the markets Hyperliquid runs itself and every market launched on its engine by an outside operator.
What are operator-deployed markets?
Hyperliquid lets outside teams launch their own markets on its trading engine, set the leverage terms, supply the reference price and collect a share of the fees. In the venue's own documentation these are called HIP-3 markets. They are where most of the equity, commodity, index and currency exposure on the venue lives, and every one of them carries its operator's identity in the market reference table.
How far back does the history go?
Executions from July 2025. Funding, positions and account balance movements from late September 2025. Order book depth from late June 2026.
How fresh is the data?
The curated tables refresh hourly. Market reference data updates as soon as a new market is registered, typically within the hour. The order book view is live over its source. A low-latency execution feed is coming soon.
Why does the execution table show about twice the volume the venue reports?
Because it records both sides of every match, which is what you want when you are reconstructing a single account's activity. For volume, use the hourly or daily market activity tables, which count each match once. That is the figure quoted everywhere on this page.
Can I measure what it costs to get in and out of a position?
Yes. The order book table holds the resting buy and sell orders at each price level, at one-minute resolution, best price first. Spread, the imbalance between the two sides, and the price you would have paid for a given size all come off a single row without a join. Depth is captured to one hundred price levels on each side.
Can I see who holds which position?
Yes. Positions are snapshotted every hour per account and market, with the direction, the size, the value and the funding paid or received. Rolled up per market it reproduces the venue's open position figure, so you can see how concentrated a market's exposure is and who is carrying it.
Can I rank traders by what they actually made?
Yes. There is a daily rollup for every account covering what it traded, what it earned or lost after fees and funding, and what it moved in or out of its balance. It is built so per-account questions do not require scanning the full execution history.
Does it include spot trading or prediction markets?
No. This dataset is perpetuals only. Spot markets and prediction markets on the same venue are indexed separately. Talk to data solutions to get access to that data.
Can I separate real-world asset markets from crypto markets?
Yes, and it is one query. Every market carries a category and an asset type, so shares, commodities, currencies, indices and private-company markets filter apart from crypto without maintaining your own list of tickers.
How do I access it, SQL, API or my own data warehouse?
Query it directly in Dune, pull it via API, ask through Dune MCP inside your own AI tools, or access it in Snowflake, BigQuery or Databricks via Datashare.
Is this data free, or does it require an Enterprise plan?
The perpetuals dataset ships as an Enterprise add-on.