Someone quotes "$2 billion in prediction market volume" in a headline, someone else quotes "$40 million in open interest" for the same platform in the same week, and both numbers sound big and neither one tells you what actually happened. That gap between the numbers people cite and what those numbers mean is the whole problem this article exists to fix.
If you are searching for prediction market volume, prediction market open interest, how big are prediction markets, or prediction market statistics, this is the direct explainer: what volume, open interest, and liquidity each actually measure, why comparing them across polymarket volume and kalshi volume is harder than it looks, and where to find the live, methodology-disclosed numbers instead of a recycled headline figure.
If you haven't read it yet, start one level back with How Prediction Market Probabilities Work — this page assumes you already know that a price is an implied probability, not a guarantee, and builds on top of that with the size and activity metrics.
TL;DR
- Volume is traded turnover over a period (a flow). Open interest is outstanding, unsettled exposure at a point in time (a stock). Liquidity is how much you can trade right now without moving the price (a depth measure). These are three different numbers, and headlines routinely swap them.
- There is no single, comparable "total prediction market volume" figure, because Polymarket, Kalshi, Manifold, and forecast platforms like Metaculus each report on a different basis — on-chain dollar value, fixed-notional contracts, play money, or nothing at all.
- A single day's own headline number can swing by an order of magnitude depending on a scope choice as simple as "open markets only" vs. "every market including closed ones" — methodology, not manipulation, drives most of the disagreement you see between sources.
- Volume on one market can signal conviction, new information, or a whale print — but it doesn't tell you the market is right, only that it's active. See Reading Whale Flow in Prediction Markets for that layer.
- CoinRithm's live stats page discloses its volume basis per venue rather than hiding it behind one blended number, and the data API exposes the same figures programmatically.
- CoinRithm is an aggregator and paper-trading sandbox — it computes and discloses these figures, it doesn't originate the underlying trades.
Volume vs. Open Interest vs. Liquidity
These three words get used almost interchangeably in casual coverage of prediction markets. They shouldn't be — they measure different things, and mixing them up is the single most common way a quoted number misleads.
Volume: turnover over a period
Volume is a flow. It's the dollar (or contract) value of trades that executed in a given window — the last 24 hours, the last 7 days, or since a market opened. A market can have enormous 24-hour volume and almost no open exposure left, because most of that volume was people getting in and out of the same position multiple times, or a market resolving and paying out. Volume answers "how much traded hands," not "how much is currently at stake."
Two volume numbers that look similar are not automatically comparable: 24h volume and lifetime volume (sometimes called total or cumulative volume) measure completely different things. Lifetime volume for a market that's been open for a year will dwarf its 24h number even on a quiet day — that's not a discrepancy, it's arithmetic.
Open interest: outstanding exposure right now
Open interest is a stock, not a flow: the total size of positions that are still open and have not yet been closed or settled, at a single point in time. In traditional futures markets this has a precise, universally reported meaning (outstanding contracts). In prediction markets it's less standardized — a CFTC-regulated exchange contract market and an on-chain order book don't compute "outstanding exposure" the same way, and a forecasting platform with no tradable contract at all has no open interest to report.
This is worth being precise about: CoinRithm's own aggregated numbers reflect open markets (how many markets are still live) and open liquidity (order-book depth on open markets) rather than a single blended "open interest" figure across all seven tracked venues, for the same reason volume can't be blended cleanly — see the next section. Treat any cross-venue "total open interest" claim you see elsewhere with the same skepticism you'd apply to a blended volume number.
Liquidity: how much you can trade without moving the price
Liquidity is neither a flow nor a stock of exposure — it's a measure of market depth: how much size sits in the order book close to the current price, i.e., how large a trade you could place right now before your own order started moving the price against you. A market can have decent lifetime volume and still be illiquid today if the people who were trading it have moved on. Liquidity is the number that matters most in the moment you're trying to actually place a trade; volume and open interest describe the market's history and current footprint, not its executability right now.
Confusing these three is exactly how a market gets described as "huge" because of a big lifetime-volume number, when the thing that actually matters to a trader looking at it today — can I get filled at a reasonable price — is liquidity, and could be thin.
Why Cross-Venue Volume Comparison Is Genuinely Hard
This is the part most coverage skips, and it's the core reason a single "total prediction market volume" number should make you ask "measured how?" before you repeat it.
Every venue reports volume on its own basis
Polymarket and Limitless are on-chain central limit order books, and their volume is traded on-chain value in USDC — a real dollar figure, verifiable on a block explorer. Kalshi is a CFTC-regulated exchange, and its volume is settlement notional: contracts multiplied by a fixed $1, which is the basis Kalshi itself reports, not an order-book dollar-traded figure. Manifold Markets runs on play money — its "volume" reflects mana, not real currency, so it's useful for market coverage and calibration culture but not for a dollar total. Smarkets is a GBP-denominated betting exchange, and its figure is lifetime matched volume — reported in pounds and converted to US dollars at an hourly exchange rate, so currency conversion is part of that number's basis. Metaculus and PredictIt don't publish a volume figure at all in a form an aggregator can pull, so an honest aggregator shows them as zero rather than estimating a number nobody actually reported.
Put a Polymarket dollar figure, a Kalshi notional figure, and a Manifold mana figure into one sum, and you get a number that is technically additive but not meaningfully comparable to any single one of its inputs. That's why CoinRithm's stats page explicitly separates monetary totals — volume, 24h volume, and liquidity — to real-money venues only, and reports play-money and forecast platforms for market counts and probabilities, not dollar figures. That's not a limitation to apologize for; it's the only honest way to combine seven fundamentally different reporting bases.
The scope you choose changes the number by an order of magnitude
Even within one platform's own data, a single methodology choice can move the headline figure dramatically. Take "24-hour volume": do you count only markets that are currently open, or every market including ones that closed and settled recently? Settlement and backfill activity on a just-closed market can register as heavy trading well after the actual event stopped being interesting — so a "24h volume" figure that includes closed-market settlement noise can be many times larger than the same figure restricted to markets still open for trading. Neither number is fabricated. They're answering different questions: "how much settlement activity happened in accounting terms" versus "how much is genuinely trading right now." CoinRithm's own stats page deliberately reports the open-markets-only figure as its headline, specifically because it's the honestly-current number rather than the technically-larger one.
There's no shared identifier across venues
The same real-world event — an election, a Fed decision, a championship game — exists as separately worded contracts on multiple platforms with different slugs and no common ID to join them on. So even "total volume on this specific question across every venue" requires an entity-matching step before you can sum anything, and that matching is itself imperfect. CoinRithm's Compare view exists to do that matching visibly, so you can see which sources are being compared rather than trusting a black-box sum.
Put together: "how big are prediction markets" doesn't have one clean answer. It has an answer that depends on which venues you include, which basis you sum them on, what time window you pick, and whether you count closed-market settlement noise. Any figure quoted without disclosing those choices is one journalist's or one tool's editorial decision dressed up as an objective fact.
What Volume Signals for a Single Market — and What It Doesn't
Zoom in from the market-wide picture to one specific market, and volume is genuinely useful — within limits.
What a volume spike can signal:
- Conviction. A sudden jump in volume alongside a price move suggests traders are putting real size behind a view, not just nudging the price with small orders.
- New information arriving. Volume often spikes exactly when news breaks — a poll, a data release, a ruling — because that's when disagreement about the right price is highest and people trade to express a new view.
- A large single print (a whale trade). One big fill can move both the price and the volume number on its own. See Reading Whale Flow in Prediction Markets for how to read a single large trade specifically — a whale print is not automatically a signal that the whale is right, and big prints are just as often hedges or exits as directional bets.
What volume does not tell you:
- Whether the market is correctly priced. High volume means the market is actively trading, not that traders agree with each other or that the current price reflects the true probability well.
- Which direction caused the volume. A market can trade heavily because two large positions are fighting each other, not because everyone suddenly agrees.
- Long-term conviction from a short-term number. A 24h volume spike can be one large trader rebalancing a single position, not a crowd forming a new consensus.
The useful discipline is treating volume as a prompt to look closer — read the market, check whether the move is broad-based or one print, and check what's happening on other venues for the same question — rather than as a verdict on its own.
How to Read Venue-Level Volume
When you compare venues instead of individual markets, a different distortion shows up: event mix, not just genuine attention, drives most of the difference in total volume between two platforms.
A platform running a single high-profile market — a presidential election, a championship final — can post enormous volume on that one question alone during peak weeks, dwarfing a quieter competitor that has broader but calmer coverage. That's a real signal about that specific event, but it says little about the platform's overall size or activity in a normal week.
The opposite distortion runs the other way: a venue that lists many small, recurring daily markets (for example, a fresh "will this crypto asset close up or down today" market every 24 hours) accumulates market count and cumulative volume quickly through sheer repetition, even if each individual market is thinly traded. A high market count or a large lifetime-volume figure driven mostly by many small recurring markets is a different kind of "big" than one that comes from a small number of markets trading with heavy, sustained size.
Neither pattern is wrong — they're just different shapes of activity, and a single ranking number collapses that shape into one comparison that hides it. Reading venue-level volume well means checking the category and market-count breakdown alongside the dollar total, not just the total by itself — CoinRithm's stats page publishes both a per-venue share table and a per-category breakdown for exactly this reason.
Where to Get the Real Numbers
Rather than repeating a specific dollar figure here — one that would be stale within days — the useful thing this article can give you is where to find the number live, with its basis disclosed.
CoinRithm's live stats page is built to be the citable reference: a headline 24h volume figure computed from open markets only, alongside lifetime volume, open and total market counts, open liquidity, and per-venue and per-category share tables — plus a methodology section that states, in plain terms, which venues' figures are on-chain dollar value, which are fixed-notional contracts, which are play money shown for coverage only, and which don't publish volume at all. As of mid-2026, total tracked 24-hour volume across real-money venues runs on the order of hundreds of millions of dollars per day — check the live page for the current figure and its exact basis, since it updates continuously and any number printed in an article is a snapshot, not a standing fact.
For programmatic access to the same figures — pulling them into a script, a dashboard, or an AI agent's context rather than reading them off a page — the prediction market data API exposes an overview endpoint with the same totalVolume, volume24h, totalLiquidity, per-venue (bySource), and per-category (byCategory) fields the stats page renders, all keyless and free. The Prediction Market Data API article covers that full surface, including the events and whales endpoints, in more depth.
Citing These Numbers
If you're a journalist, researcher, or an AI assistant answering a question about prediction market size, the live stats page is built to be cited directly rather than paraphrased from a stale screenshot. It states its own methodology on the page, so a citation can point to a specific, disclosed basis instead of an unsourced headline number. Attribution is welcome and appreciated whenever you use these figures:
Data by CoinRithm — https://www.coinrithm.com
That applies whether you're quoting the 24h figure in an article, building a chart from the overview endpoint, or having an AI assistant answer "how big are prediction markets right now" — cite the live page, not a number copied from an earlier article (including this one).
How CoinRithm Fits In
CoinRithm aggregates volume, open-market counts, and liquidity across seven prediction-market and forecasting venues — Polymarket, Kalshi, Metaculus, PredictIt, Limitless, Manifold Markets, and Smarkets — and publishes both the numbers and the methodology behind them, rather than a single blended figure with no basis attached. The live stats page is the citable reference for the current picture; the data API exposes the same figures programmatically; and the Compare view lets you see the same underlying question matched across venues so you can judge divergence yourself instead of trusting a single number.
To be direct about what CoinRithm is and isn't: it computes and discloses these figures as an aggregator, it doesn't originate the underlying trades, and its paper trading simulator lets you practice acting on volume and liquidity signals with mock funds — never real money.
Frequently Asked Questions
What's the difference between prediction market volume and open interest?
Volume is turnover — how much traded over a period, like the last 24 hours. Open interest is a snapshot of outstanding, unsettled exposure at one moment. A market can have heavy 24h volume with little open exposure left if most of that volume was people entering and exiting the same position, or the market settling.
Why do different sites report different total prediction market volume figures?
Because each venue reports volume on its own basis — Polymarket and Limitless in on-chain dollar value, Kalshi in fixed-notional contracts, Manifold in play money — and because scope choices like "open markets only" vs. "every market including closed ones" or "24h" vs. "lifetime" change the number significantly. A different total usually reflects a different, undisclosed methodology, not necessarily an error.
How big are prediction markets right now?
As of mid-2026, tracked 24-hour volume across real-money prediction market venues runs on the order of hundreds of millions of dollars per day, but the exact figure changes continuously and depends on which venues and which time window you're counting. Check CoinRithm's live stats page for the current number with its basis disclosed, rather than relying on a figure printed in any single article.
Does high volume mean a prediction market is priced correctly?
No. High volume means the market is actively trading — traders disagree enough, or new information arrived recently enough, to generate real turnover. It does not mean the resulting price is correct. A heavily traded market can still be mispriced; volume tells you the market is being actively contested, not who's right.
Is prediction market liquidity the same thing as volume?
No. Liquidity measures order-book depth right now — how much you could trade before moving the price against yourself. Volume measures how much already traded over a past period. A market can have large historical volume and thin liquidity today if the traders who drove that volume have since moved on.
Where can I find real-time prediction market volume data instead of a static article figure?
CoinRithm's stats page updates continuously and discloses its methodology on the page. For programmatic access, the data API's overview endpoint returns the same figures as JSON, free and keyless.
Does CoinRithm publish a single "open interest" number across all venues?
No, and deliberately so — like volume, open interest isn't reported on a comparable basis across a CLOB, a regulated exchange, and forecast platforms with no tradable contract at all. CoinRithm instead reports open market counts and open liquidity per venue, which is the accurate way to represent "current footprint" without inventing a blended figure the underlying venues don't actually agree on.
Continue reading: Prediction Markets vs. Sports Betting — how event-contract trading actually differs from wagering, including how each side handles size, pricing, and payout.
Disclaimer: This article is for educational and informational purposes only and is not financial, legal, or investment advice. CoinRithm aggregates prediction market data for research and development purposes and does not execute real-money trades; execution happens on the underlying platform. Any figures referenced above reflect the live stats page's methodology at time of writing and change continuously — always check the live stats page for the current numbers.
