Open the Kalshi market endpoint for any live contract and you get back two numbers that look like they measure the same thing and do not. One field is called volume_fp. Another is called open_interest_fp. Both are counts of contracts, both move throughout the day, and both get casually described as "activity".
They measure different things. Volume counts contracts traded. Open interest counts contracts still outstanding. A market can have enormous volume and almost no open interest, or the reverse, and each shape tells you something specific about who is actually holding a position when the question resolves.
Deposit $20, get $50 to trade
Welcome bonus on the world's largest prediction market.
18+ · New users only · Availability and terms vary by region · Trading involves risk of loss · BeGambleAware.org
The one-line answer, and then the honest version
Volume is a flow. Open interest is a stock. If that distinction is enough for you, you can stop reading, but it is also the kind of sentence that sounds clearer than it is. The practical difference only lands once you watch a single contract trade back and forth between two people. Each time they transact, volume rises by one. Open interest does not move at all, because the same contract keeps changing hands. Nothing new has been created, nothing has been settled, and the total exposure in the market is unchanged.
Now imagine a fresh buyer steps in and a fresh seller takes the other side. A new contract has been created. Volume rises by one. Open interest rises by one. Both have moved together, and they have moved for the same reason: new money has committed to a position that did not exist a second ago.
That is the core mechanic. Volume captures turnover. Open interest captures commitment.
How this shows up on Kalshi
Kalshi's own API documentation makes the split explicit. The Get Market endpoint returns volume_fp, volume_24h_fp and open_interest_fp as three separate fields on the same market object, alongside price fields such as yes_bid_dollars, yes_ask_dollars and last_price_dollars. Kalshi describes volume_fp as the market volume "in contracts", and open_interest_fp as "the number of contracts bought on this market disconsidering netting".
The _fp suffix marks a fixed-point count: a string with two decimals, such as "10.00", because Kalshi supports fractional contracts. It is not a dollar amount, which is one of the easiest things to get wrong when quoting a Kalshi figure. The field notional_value_dollars is not a dollar volume either: Kalshi defines it as the total value of a single contract at settlement, in dollars. To turn a contract count into money you need a price, and which price you use changes the answer.
The documentation also notes that the market response is cached for a short time, so prices, sizes and status can lag the exchange by a few seconds; Kalshi points to its WebSocket API for current state. For a definition article this is a footnote. For a trader trying to read an intraday squeeze it matters.
And there is no example figure here on purpose. Kalshi's documentation page lists the fields and their shape, not the live value for any specific market, so no honest snapshot number can be lifted off it.
How this shows up on Polymarket
Polymarket's international exchange, which US traders can't use, does not list an open interest field on the market object in its Gamma markets reference. Open interest sits in a different place: Polymarket's Data API has an open interest endpoint, GET /v2/oi, which returns what it calls "priced gross open interest" for one or more markets, or for the whole exchange if no market is named. Note the word "priced". Polymarket's figure is the outcome tokens held, multiplied by their price, with no netting. That makes it a money value, not a count of contracts like Kalshi's, so the two venues' numbers cannot be compared directly.
The mechanic underneath is set out in Polymarket's documentation. Trading is collateralised in pUSD (Polymarket USD), a token backed by USDC that replaced bridged USDC in Polymarket's April 2026 exchange upgrade. Shares in event outcomes are priced between 0.00 and 1.00. Splitting creates a complete set: every 1 pUSD produces 1 YES token and 1 NO token. Merging reverses it: every 1 YES token and 1 NO token returns 1 pUSD. Polymarket's help centre puts the trading side directly: "Shares are created when the sum of what each side is willing to pay is equal to $1.00".
So on Polymarket the stock-versus-flow split still applies, but the vocabulary is different. Volume is still turnover of outcome tokens. Open interest is the value of outcome tokens still outstanding. Same idea, different plumbing.
Why one rises and falls without the other
A worked sequence makes this obvious. Say a market opens with zero volume and zero open interest. Trader A buys a YES contract from Trader B, who is opening a fresh NO position. Volume goes to 1. Open interest goes to 1. A new contract exists.
Now Trader A decides to get out and sells the YES to Trader C, who is opening a new YES position. Volume goes to 2. Open interest stays at 1. Trader A has exited; Trader C has entered; the total number of live contracts has not changed, because one side of the existing contract simply changed hands.
Third move: Trader B wants out of the NO position, and Trader C is happy to close the YES against it. Volume goes to 3. Open interest goes to 0. Both sides have exited, the contract has been retired, and nobody is left holding exposure to the outcome.
Three trades, three volume ticks, open interest moved 0 → 1 → 1 → 0. The volume figure never tells you which of these shapes is happening. Open interest does.
What each number actually tells you
Volume is the right lens for liquidity. If 50,000 contracts changed hands yesterday, you can estimate how tight the bid-ask is likely to be, how large a position you can enter without moving the price, and how often someone is willing to trade against you. Our guide on what liquidity means in prediction markets walks through why thin volume kills otherwise clean trades, and what volume actually means on a prediction market covers the volume figure itself.
Open interest is the right lens for exposure. If open interest is 10,000 contracts against a binary outcome, then 10,000 contracts are waiting for resolution, each due to pay a fixed amount to whichever side is right. On Kalshi that amount is the contract's notional_value_dollars; on Polymarket each winning token redeems for 1 pUSD. That tells you how much the market has actually committed to the question, as distinct from how often it has shuffled the same chips around the table.
The two can diverge sharply. A market with huge volume and tiny open interest is being day-traded; positions open and close inside a session, and few people hold exposure for long. A market with large open interest and modest volume is being held, not traded; conviction is in place but the book is quiet. Reading a market by only one of these numbers gives you half the picture.
Where open interest surprises you
Two scenarios are worth knowing. The first is the end of a market's life. Open interest does not simply fade away; it is settled. On Polymarket, once a market resolves, redeeming turns outcome tokens into pUSD: each winning token returns 1 pUSD and each losing token returns nothing. On Kalshi, each contract settles at its stated value. Until then, open interest falls only as far as holders choose to close their positions early, so a high figure on the eve of resolution is a measure of how much is about to be paid out.
The second is a stale market with real exposure. If a contract has barely traded in a week but open interest is still substantial, that is not a dead market. That is a market where people have taken positions and are sitting on them. Price can move sharply if a single large holder decides to exit, because the thin recent volume means there may be no cushion of counterparties ready to absorb them. This is the shape where the implied probability from reading prediction market odds can mislead you, because the last traded price is old.
Trade 400+ coins
The premier crypto platform, trusted by 100M+ users.
18+ · Eligibility and terms apply · Crypto is volatile; capital at risk
The reading trap
The most common mistake with these two numbers is treating them as interchangeable headlines of "activity". A press release that cites "$X in volume" is making a flow claim. A chart that shows "open interest at an all-time high" is making a stock claim. Both can be true at the same time about the same market and they are not the same story.
And treat the units with care. Kalshi's volume_fp is a contract count, not a dollar figure; quoting it in dollars without converting through a price is a straightforward misread. Polymarket's open interest figure is already priced, in money, so setting it beside a Kalshi contract count compares two different things. Polymarket's shares are priced between 0.00 and 1.00 in pUSD, so a volume figure in shares and a volume figure in pUSD notional are different numbers too.
Facts checked on 21 November 2026, against Kalshi's Get Market API documentation, Polymarket's Gamma markets reference, its Data API open interest reference, its pUSD and position-management pages, and its help-centre page "What is Polymarket".
18+. For information only, not financial advice.
Frequently asked questions
What is open interest in a prediction market?
Open interest is the total of a prediction market's positions that are currently outstanding, meaning positions that have been opened and not yet closed or settled. On Polymarket each complete set is a YES token and a NO token created from 1 pUSD of collateral. When a buyer and a seller both open new positions against each other, open interest rises. When both sides close out, it falls. When one holder sells to a new buyer, open interest does not move at all, because the position has only changed hands. It is a measure of standing exposure, not of trading activity, and it tells you how much is actually committed to the outcome.
How is open interest different from volume?
Volume is a flow; open interest is a stock. Volume counts every contract that trades over a given period and keeps ticking up each time the same contract changes hands. Open interest counts only the contracts still live in the market, so it moves up when a new position is created and down when an existing position is closed. Kalshi's API makes the distinction explicit, returning volume_fp, volume_24h_fp and open_interest_fp as separate fields on the same market, each a count of contracts. A market can post large volume while open interest stays flat, which means traders are churning the same positions rather than committing fresh capital to the question.
Does Polymarket show open interest?
Yes, though not on the market object in its Gamma markets reference. Polymarket's Data API has a separate open interest endpoint, GET /v2/oi, which returns "priced gross open interest" for a market, or for the whole exchange. It is a money value (tokens held multiplied by their price, with no netting) rather than a count of contracts, so it is not directly comparable with Kalshi's open_interest_fp. Underneath, the mechanic is the same: 1 pUSD splits into 1 YES token and 1 NO token, and a YES and NO pair merges back into 1 pUSD.
Can a market have high volume but low open interest?
Yes, and this is one of the most useful signals in reading a prediction market. High volume with low open interest means contracts are being traded actively but positions are not being held; traders are opening and closing inside the same session, or the same contracts are rotating between a small set of counterparties. It is the signature of a day-traded market rather than a conviction-held one. The reverse shape, low recent volume with substantial open interest, means holders are sitting on positions and not trading them, which can leave the price stale and vulnerable to sharp moves if a single large holder decides to exit.
What happens to open interest when a market resolves?
It is settled and paid out. On Polymarket, redeeming after resolution converts outcome tokens into pUSD: each winning token returns 1 pUSD and each losing token returns nothing. On Kalshi, each contract settles at its stated value, which the API gives as notional_value_dollars. Before resolution, open interest falls only when holders close positions early, by trading out or, on Polymarket, by merging a YES and NO pair back into pUSD. So a large open interest figure close to resolution is not a warning sign; it is the amount waiting to be paid.
