Why Kalshi and Polymarket Prices Differ on the Same Question

Open Kalshi and polymarket.com side by side, find what looks like the same question, such as what the Federal Reserve will do at its December 2026 meeting, and the two prices may not match. When they differ, the reason is rarely that one venue is right and the other wrong. More often you are looking at two different contracts, with different fees, different order books and different rules about who may trade.

That matters for anyone reading prediction market prices as probabilities. A price of 54 cents on one venue and 56 on the other, on what looks like the same question, is not two cents of free money. It is a prompt to check whether the contracts, or the venues, are as alike as they look. This guide walks through why gaps appear and how to tell a real disagreement from a mechanical one, using each venue's own published terms. Facts checked on 9 November 2026.

The two contracts are rarely identical

Start with the words on the page. Both venues list the Fed's December 2026 decision, and both split it into five outcomes: a cut of more than 25 basis points, a 25 basis point cut, no change, a 25 basis point rise and a rise of more than 25. The questions look the same. The terms are not.

polymarket.com's "Fed Decision in December?" market defines the rate as "the upper bound of the target federal funds range" and resolves to "the amount of basis points the upper bound of the target federal funds rate is changed by versus the level it was prior to the Federal Reserve's December 2026 meeting." If the Fed moves by an amount that is not one of the listed options, "the change will be rounded up to the nearest 25 and will resolve to the relevant bracket." Its resolution source is "the FOMC's statement after its meeting scheduled for December 8-9, 2026", and it "may resolve as soon as the FOMC's statement for their December meeting with relevant data is issued." If no statement is released by the end date of the next scheduled meeting, it resolves to "No change".

Kalshi's "Fed decision in Dec 2026?" market is worded around the action itself. Its 25 basis point rise market reads: "If the Federal Reserve does a Hike of 25bps on December 09, 2026, then the market resolves to Yes." Its rules add that "The outcomes are mutually exclusive: at most one can resolve to Yes", and that if the scheduled FOMC meeting "is canceled and does not occur on its scheduled date", the "Fed maintains rate" outcome "resolves to Yes and all other outcomes resolve to No." Kalshi's contract terms for the series name the Federal Reserve System as the Source Agency and set an expiration time of 2:05 PM ET.

In an ordinary decision, a 25 basis point move announced on 9 December, the two land in the same place. The differences sit in the unusual cases: an unusual move size, which polymarket.com rounds up to the nearest 25 basis points; a meeting that is cancelled or produces no statement, which both treat as no change but describe in different words; and timing. Traders who read those cases can price them, and that can open a small gap between two contracts that look identical. For the basics of how a price maps to a probability, our guide to how prediction market odds work is the place to start.

Fees: what each venue charges

The quoted price is not the price you pay, and the two venues charge in different ways.

Kalshi's fees page says it "makes money by charging a transaction fee on the expected earnings on the contract", and that some markets carry different fees from others. Its published fee schedule, last updated and effective on 7 July 2026, gives the general trading fee as round up(0.07 x C x P x (1-P)), where C is the number of contracts and P is the price in dollars, with a multiplier that varies by market. Some markets also carry maker fees, charged on resting orders when they are eventually matched, at round up(0.0175 x C x P x (1-P)). Kalshi lists its Fed decision series as one with maker fees.

polymarket.com runs a central limit order book, and its fees page says it "charges a small taker fee on certain markets". The formula is fee = C x feeRate x p x (1 - p), where C is the number of shares and p is the price. Only takers pay: "Makers are never charged fees." The rate depends on the market's category: 0.05 for economics, 0.04 for politics, 0.07 for crypto, and nothing for geopolitical and world events markets, which "are fee-free". The December Fed market is listed with economics fees.

Both formulas are largest at a price of 50 cents and shrink towards 0 and $1. On 100 contracts bought at 50 cents as a taker, Kalshi's general formula gives about $1.75 and polymarket.com's economics rate gives $1.25.

The practical point: a two-cent gap between the venues has to be wider than the cost of trading on both before it is anything more than a gap on screen. On small trades it often will not be.

Spreads and depth are the second filter

A quoted price can also mislead. polymarket.com's own documentation explains that "The displayed price is the midpoint of the bid-ask spread", and that if the spread is wider than $0.10, the last traded price is shown instead. It adds that you "won't necessarily trade" at the displayed price: you pay the ask when buying and receive the bid when selling.

So the real comparison is between what you can buy at and what you can sell at on each venue, not between two headline numbers. If one venue's best bid is 54 and its best ask 58, its midpoint is 56, but nobody can trade at 56 there. A difference between two venues that is narrower than their spreads is not a disagreement at all. Our breakdown of what liquidity actually means on a prediction market explains why some books stay thin.

Who can trade on each venue

The venues also differ on who may open a position, and that is published.

Kalshi describes itself as "subject to U.S. regulatory oversight by the CFTC". polymarket.com is not available in the United States or the United Kingdom: its geoblocking page lists both among the regulatory-restricted jurisdictions where users "can close existing positions but cannot open new ones". Polymarket US is a separate venue with its own markets and rules.

We cannot see who trades on either venue, so this guide does not describe their customers. What can be said is that the two prices are set by order books open to different sets of people, and that some of those people cannot trade on the other venue. That is one reason a gap can last longer than it would between two listings of the same share.

Resolution timing

Look again at the Fed contracts. polymarket.com's market "may resolve as soon as the FOMC's statement for their December meeting with relevant data is issued", with the statement as its source. Kalshi's terms point to the Federal Reserve System as the Source Agency, with an expiration time of 2:05 PM ET. On an ordinary meeting day both settle within hours of the decision. Where the two can part is the unusual case: a cancelled meeting, a statement that comes late, or a move size neither venue lists. Prices close to a deadline can reflect those paths as well as the decision itself.

How to read a gap in practice

You can work through it in five steps.

First, open each market on its own venue and read the rules, not the headline: the resolution source, the expiration, and what happens in the unusual cases. Second, check the best bid and best ask on each side, not the displayed price. Third, work out the fees you would pay on each venue, using each venue's own formula. Fourth, check who can trade on each: someone in the US or the UK cannot open a new position on polymarket.com. Fifth, ask whether what is left of the gap is wider than the two spreads plus fees. If it is not, it is noise. If it is, the next question is whether the differences in the contract terms explain it.

For venue-specific help, our guides on how to read Kalshi odds and how to read Polymarket odds walk through turning a price into a probability on each venue.

What iPredicta does with this

iPredicta is a UK-based prediction markets discovery platform. We track markets across ten venues and show their prices side by side wherever a question trades on more than one, so readers can see a gap and look into it. Our Kalshi vs Polymarket page compares the two venues, and our comparison page carries the live board. Treat a visible gap as a prompt to read the contracts.

Frequently asked questions

Why do Kalshi and polymarket.com show different prices for the same question?

Because they are rarely pricing exactly the same contract. On the Fed's December 2026 decision, for example, polymarket.com rounds an unusual move up to the nearest 25 basis points and resolves from the FOMC statement, while Kalshi's contract names the Federal Reserve System as its source and sets out its own rule for a cancelled meeting. On top of that, each venue charges its own fees, the displayed price is not the price you can trade at, and the two venues are open to different people: polymarket.com does not let anyone in the US or the UK open a position.

Can I just buy the cheaper side on one venue and sell on the other?

Only if you can trade on both, and many people cannot: polymarket.com lists the United States and the United Kingdom as places where users "cannot open new ones", meaning new positions. Even where both are open to you, you pay the ask to buy and receive the bid to sell, each venue charges its own fee, and the two contracts may settle differently in unusual cases. Treat a gap as information about the market. Our guide on spotting mispriced prediction markets covers where real mispricing tends to appear.

Why do fees matter so much to the gap?

Because anyone trying to close a gap pays a fee on each venue. Kalshi's general trading fee is round up(0.07 x C x P x (1-P)) under its fee schedule effective 7 July 2026, with maker fees on some markets. polymarket.com charges takers C x feeRate x p x (1 - p) at a rate set by category, 0.05 for economics markets, and never charges makers. On 100 contracts at 50 cents, that is about $1.75 on Kalshi's general rate and $1.25 on polymarket.com's economics rate. A gap has to be wider than the combined cost before it is worth anything.

Is one venue more accurate than the other?

There is no general answer. Accuracy can only be measured market by market, against how each one resolved, and the two venues list different markets under different terms. Where both list a question, comparing their prices, spreads and terms tells you more than either alone. Our piece on why prediction markets are accurate, and where that breaks down covers the mechanics behind this.