What this is
This listing points at two related things on Dune, the onchain analytics platform, and it is worth keeping them apart. The first is a community-built cross-venue board, Prediction Market Dashboard, published by the Dune author @datadashboards (renamed from @dunedata) — the board journalists, analysts and funds cite when they need a single cross-venue number. The second is Dune's own curated Prediction Markets dataset, a first-party data product that unifies Polymarket and Kalshi into one schema. The board is a dashboard built on top of data like this; the dataset is the underlying product.
Underneath both sits the platform itself: normalised blockchain datasets spanning over 130 chains, with dashboards, queries, APIs, warehouse syncs and an MCP surface for agents.
The board's headline numbers
As of 30 September 2026 the board showed $130.3B lifetime taker volume, $133.9B traded volume, $1.87B open interest, 2.32B transactions and 4.51M unique users across the venues it tracks, with Kalshi leading on transactions and Polymarket leading on users. The venue set is broad: Polymarket and Polymarket US, Kalshi, Limitless, Opinion, Predict, Myriad, Overtime, Novig, Gemini, ForecastEx and Hyperliquid HIP-4 outcome markets.
Those are the numbers that get quoted in articles, which is exactly why the next section matters.
Reading the board honestly
The board's author is unusually explicit about how its own figures can mislead, and that candour is the most useful thing on the page:
- Use taker volume for cross-venue comparisons. Traded and notional contract volume count contracts rather than dollars, so a venue with cheap contracts can look bigger than it is.
- No wash-volume filter is applied. Reported activity includes anything the venues print.
- Kalshi's feed is uploaded by Kalshi itself and never includes the current day, so Kalshi is structurally a day behind on this board.
- Hyperliquid is a manual refresh rather than an automated feed.
Put together, that means a single headline number is a reasonable orientation and a poor argument. The board is most useful when you look at the trend across venues and periods rather than one figure in isolation.
What Dune's curated dataset adds
Dune's first-party Prediction Markets collection does the normalisation work that a community board has to do by hand. It unifies Polymarket (onchain on Polygon, covering Combos parlays, from November 2022) and Kalshi (from June 2021) into one schema with a shared market-category taxonomy and a canonical Yes/No direction, and delivers trade-level data, hourly price history and resolution outcomes. The collection card quotes roughly 3+ years of history and about $170B in notional volume, with Hyperliquid HIP-4 outcome markets tagged alongside the two main venues.
Two properties matter more than the totals: every resolved market carries its final outcome, which makes the history usable for signal validation and calibration studies without further cleaning, and the vendor absorbs upstream schema changes so queries do not break when a venue's API shifts.
Where it fits in this directory
This is the measurement layer, not a venue and not a signal service. It does not price anything, take positions or make forecasts; it records what the prediction markets did. That makes it the natural cross-check on the venue pages in this directory — when a venue reports its own volume, the board and the curated dataset are where you go to see the number in context, alongside the venues it competes with.
Who it is for
- Journalists and analysts who need a citable cross-venue figure and, ideally, one caveat attached to it.
- Funds and research desks treating prediction markets as an alternative-data signal, which is a use case Dune markets directly to hedge funds.
- Builders who want trade-level prediction-market history in SQL rather than reconciling two venue APIs themselves.
- Anyone auditing a claim — a venue's own dashboard versus an independent index of the same activity.
What it is not
It is not a trading interface, a venue or a source of live prices: the community board can lag, Kalshi's feed excludes the current day, and the curated tables land within 24 hours of source. Free access is view-only, so even reading the SQL behind a public chart requires an account with credits. Nothing here is investment advice, and Bitnxt is not affiliated with Dune.





































