Kalshi vs Polymarket: different counterparty, different settlement authority, and the same round trip costs $3.26 or $0.00 (August 2026)

The two venues are not competing products with different bonuses. They are different legal objects. Kalshi is a designated contract market on the Commodity Futures Trading Commission's public register, so your counterparty sits inside a federal registration you can look up. Polymarket's main property is close-only for United States users on both the frontend and the API, read from its own geoblock endpoint on 5 August 2026, and its separate US property named no operator and stated no regulatory status when we read it the same day. On cost the gap is wider than any promo: one 100-contract round trip, bought at $0.60 and sold at $0.66 with both legs taken, costs $3.26 in fees on Kalshi and anywhere from $0.00 to $3.25 on Polymarket depending on which category the market sits in and whether your orders rest. Fee schedule version 2026-08-04.a.

Every ranking comparison compares bonuses, and we measured how far that goes

On 19 August 2026 we pulled the pages ranking for this query and read thirteen of them. Every one carries a fees section. One of the largest, at roughly 6,500 words, carries a fees-and-costs H2 containing zero numerical fee figures. Eight of the thirteen were updated between 14 and 19 August, so this is not staleness. It is a subject-matter choice.

Three things were missing from all thirteen: a live price from both venues on the same contract; any reconciliation of Polymarket's two published fee schedules, which every page treats as one number; and a worked cost example with executed numbers, the closest being a hypothetical 100 contracts at 40 cents. The rest of this page is those three things.

Your counterparty is a different kind of thing on each venue

Custody is where we stop short of the answer you want. In community vocabulary Polymarket positions are described as fully collateralised and cash-settled, and Kalshi holds customer cash as a federally registered exchange. We have not verified either description against a primary custody document, we have not confirmed segregation arrangements at either venue, and NPR reported finding no sign of Polymarket at its stated Panama headquarters. Anyone stating confidently where your money sits on Polymarket should be asked which document says so. Our page on this is where your money sits, and it will not assert an answer either until the document exists.

The fee formulas, side by side, from the published schedules

Kalshi. Taker fee is roundup(0.07 × C × P × (1 − P)), where C is contracts and P is price in dollars, rounded up to the next whole cent. The published maker coefficient is 0.0175. That maker figure carries a live contradiction: Kalshi's newsroom states resting orders are fee-exempt while its help centre states maker fees are charged, both are Kalshi's own publications, and we have found no public reconciliation. Kalshi also acknowledges special-event fee tiers for elections, awards ceremonies and large sporting championships, and has never quantified them anywhere public.

Polymarket .com. Same formula shape, rate × C × P × (1 − P), with the rate set per category and no maker fee at all. Zero, not a discount:

Polymarket US. A separate documentation hub publishing a flat 0.06 taker coefficient and a 0.0125 maker rebate, stated effective 1 July 2026, read 19 August 2026. That is a different schedule from the one above, and the two are never reconciled by the platform or by anyone ranking for this query. Our fee model at version 2026-08-04.a carries the per-category .com rates only, so the .us figures above are read but not yet modelled.

This split is not trivia. It is the direct cause of the coefficient errors in this niche. Our notes record a competitor fee page asserting a single coefficient of 0.0625 for five months against documented per-category rates of 0.04 to 0.07, which is roughly where you land if you average two schedules you have not noticed are two.

The same trade, priced on both venues

Buy 100 contracts at $0.60, sell at $0.66. Gross move $6.00. Both legs taken.

Three things fall out of that table that no bonus comparison can reach.

At the same coefficient the venues differ only by the rounding rule. Kalshi taker and Polymarket crypto both charge 0.07. The exit leg computes to $1.5708 on both. Kalshi rounds it up to $1.58 and Polymarket does not, so the trade costs $3.26 against $3.2508. On this trade the rounding is worth $0.0092.

The rounding rule stops being trivial in the longshot corner of the board. One contract at $0.03 generates a raw Kalshi fee of $0.002037, which rounds up to a full cent. You pay 4.9 times the unrounded rate, and that penny is 33% of your $0.03 stake. Polymarket publishes per-category rates but we have found no published rounding rule for Polymarket at all, so we do not know whether the same order rounds there. That is on the unverified list below, and it is the single fee question most worth answering next.

The maker row is where the venues genuinely separate. Polymarket charges makers nothing in every category; Kalshi's published maker coefficient makes the same trade cost $0.82. Posting a resting order rather than crossing the spread saves more on either venue than any promotional credit either one advertises. What it costs instead is adverse selection.

Break-even by entry price is the number that decides which venue is cheaper

Break-even as a share of stake, for one leg, is rate × (1 − P). The contract count cancels the price, which is why this is a clean comparison and why the cheap end of the board is the expensive end in percentage terms.

Read the rows before the columns. On Kalshi, moving your entry from $0.90 to $0.30 raises break-even from 0.70% to 4.90%, a 4.20 point swing. Switching from Kalshi to Polymarket's 0.04 band at $0.90 moves it from 0.70% to 0.40%, a 0.30 point swing. Entry price moves the number fourteen times further than venue choice does at that price. Every comparison that answers "which is cheaper" with a single number has thrown away the variable that matters most. Both calculators are live: Kalshi fee calculator and Polymarket fee calculator. The full Kalshi derivation, including the rebate accumulator, is in Kalshi fee math.

Our own log says the cross-venue trade is negative before the market moves

This is the part an affiliate cannot publish, because publishing it breaks the marketing.

Our paper trader (Collector B, 500 markets, 8 tags) opened five positions on 28 July 2026 with $500 deployed. Running all five through the verified per-category fee arithmetic:

The two-leg column adds a Kalshi leg at 0.07 to the Polymarket position. It is negative at entry: minus $3.48 on $500, before the market moves at all. That is the honest answer to "can you run the same position on both venues", and it is one observation on five positions, so read it as five worked examples and not as a rate. n=5 supports no rate.

The last row is worse than a bad outcome. At an entry price of $0.2965 in a 0.04 category, break-even is 0.04 × 0.7035 = 2.81% of stake. It was entered on a claimed edge of 1.20%. It was arithmetically unwinnable at the moment of entry. The cause is a flat percentage threshold in our own scanner that ignores the price-dependence of the fee, and the same 1.20% claimed edge on the Hyperliquid position at $0.8693 was genuinely profitable because break-even there is 0.91%. Same edge number, opposite economics, decided entirely by entry price. That threshold fix has not shipped, and the scanner still publishes a 1.0% minimum edge against real fee-adjusted break-evens of 3.63%, 2.56% and 2.04% depending on rate.

One more finding that belongs on a versus page and appears on none of them: buying YES and NO on the same Polymarket market for a combined price under $1.00 does not occur. Gamma exposes bestBid and bestAsk only, the NO side is the arithmetic complement, so yes_ask + no_ask = 1 + spread, always at or above $1.00 by construction. Two independent systems reached that conclusion from different directions. Our scanner finds momentum, not arbitrage, and we say so rather than let the word do work it has not earned.

Who writes the settlement rules, and what happens when you disagree

On Kalshi the answer is documented and uncomfortable: Kalshi writes the rules for the contract and determines the basis for settlement, with no independent intermediary. That is the Massachusetts Attorney General's characterisation of the arrangement, not ours. We have found zero documented appeal paths outside Kalshi itself. What does exist, because Kalshi is CFTC-registered, is an escalation ladder above the exchange: exchange support, then the BBB, then CFTC reparations, then a state attorney general, then the NFA.

On Polymarket we are not going to describe the resolution mechanism, because we have not verified it against a primary document and describing an oracle design from memory is exactly the failure mode this page exists to document. What we can say is that no equivalent US escalation ladder has been verified by us, and polymarket.us disclosed no regulatory status at all when read on 5 August 2026.

On responsiveness there is one hard number, and it belongs to Kalshi: its Better Business Bureau profile records 214 complaints, of which 174 are closed without a response from the business, which is 81%. Name that metric precisely. It counts complaints the business did not answer through the BBB channel. It is not a fraud rate, not a satisfaction score, and it has no denominator, because Kalshi does not publish an active-trader count anywhere we have found. We do not hold a comparable figure for Polymarket, so no comparison is available on this line and we are not going to imply one.

The dominant one-star theme on both Trustpilot and the BBB is withdrawals and KYC holds. If you want the carveout-hunting version of this, it is how to read a market's rulebook and how event contracts settle.

US access is where these pages get the question wrong

Three different questions get answered as one, and separating them resolves every contradiction on this SERP:

1. Is it lawful for me to trade? State law.
2. Is the platform geofenced for my state? Platform policy.
3. Am I promo-eligible? Marketing terms.

Kalshi. Federally registered as a DCM, which is not a state-law ruling. At least six state-level actions landed on prediction market contracts during 2026. Our notes record Massachusetts contracts enjoined, Nevada under injunction, a $5 million fine connected to Ohio (issuing body and order number unverified), cease-and-desists in Connecticut, Minnesota's SF 3432 effective 1 August 2026, and Illinois SB 3019. We have not pulled those dockets and you should treat the list as a pointer, not a citation. Kalshi's referral programme is separately US only, pays non-withdrawable trading credits, and those credits expire 7 days from issuance.

Polymarket .com. The geoblock endpoint, read 5 August 2026, lists the United States close-only on frontend and API, alongside the United Kingdom, four named Canadian provinces, France, Germany, Italy, Poland, Singapore, Taiwan and Thailand among others, with Iran, Syria, Cuba and North Korea blocked completely. That page carries no date and no version number, which is why we diff it daily rather than quote it once.

Polymarket US. A sign-up page and two links when we read it on 5 August 2026. No operator named, no CFTC or DCM status stated, no state availability list, no statement on whether US users may open positions.

The practical US answer as of August 2026: Kalshi has a federal registration you can verify and a state-law question you must check yourself, while Polymarket splits into a property you cannot open positions on from the US and a property that tells you nothing about itself. The deeper trust write-ups are is Kalshi legit and is Polymarket safe, with the full write-ups at Kalshi review and Polymarket review.

What we could not verify

This section is the point of the page.

• Whether Polymarket rounds its fees, and in which direction. Kalshi documents a roundup to the whole cent. We found no equivalent published rule for Polymarket, which means the sub-cent comparisons above are exact for Kalshi and unrounded for Polymarket.
• How the two Polymarket schedules relate. We read a flat 0.06 taker and 0.0125 maker rebate on the .us hub and per-category 0.04 to 0.07 with a zero maker fee on the .com hub. Nothing we read explains which applies to whom, or whether one supersedes the other.
• Custody and fund segregation at either venue. Not verified. Not asserted.
• Polymarket's resolution mechanism and any appeal path. Not verified against a primary document, so not described.
• The competitor coefficient claim. Our notes record a fee page asserting 0.0625 for five months. That page serves HTTP 403 to fetchers, so we have never read it ourselves and the claim sits in our notes without a page read we can stamp.
• Docket numbers and case captions for every state action listed. The $5 million Ohio figure in particular is recorded without an issuing body or order number.
• The read date on the 174-of-214 BBB figure. Complaint counts move and our notes do not preserve the exact date it was read.
• Kalshi's maker fee. Newsroom and help centre disagree. Unresolved, and it changes the maker row in the table above.
• Kalshi's special-event fee tiers. Acknowledged to exist for elections, awards ceremonies and large championships. Never quantified publicly.
• Liquidity depth on either venue. We collect prices, not book depth at size, so we make no claim about which venue fills a large order better.

Nobody pays us for this page, and our own log is down $28.30

PredictionEdge currently takes no referral revenue, and whether it ever will is undecided. There is no Kalshi code anywhere on this site; Kalshi's published programme is an ordinary refer-a-friend scheme by its own framing, with no application form and no rate card. A Polymarket code exists on our scanner page and it has earned nothing to date.

Our paper log is public and unflattering: 10 trades, 6 closed, net −$28.30 on $500 deployed as of 4 August 2026, including one position that the arithmetic said could not win before it was opened. Our own scanner threshold is wrong and named as wrong above. If a comparison page cannot tell you what it gets paid and cannot show you a losing trade, apply that to what it tells you about fees. The vocabulary used here is defined in the glossary.

Questions people type before choosing a venue

Is Kalshi or Polymarket cheaper? Depends on category and order type, not on the brand. The same 100-contract round trip at $0.60 to $0.66 costs $3.26 on Kalshi as a taker, $1.8576 in a 0.04 Polymarket category, and $0.00 in a geopolitical market or on any Polymarket maker order. Kalshi's 0.07 is flat across all markets.

Can I use Polymarket in the US? Not on polymarket.com. Its own geoblock endpoint, read 5 August 2026, lists the United States as close-only on both the frontend and the API, meaning you can exit positions but not open them. A separate property, polymarket.us, exists, but when read the same day it named no operator and stated no regulatory status.

Does Polymarket charge maker fees? No. Makers are charged zero on polymarket.com in every category, which is a zero rather than a discount. Kalshi publishes a maker coefficient of 0.0175, which on that same 100-contract round trip works out to $0.82. Kalshi's own newsroom and help centre contradict each other on whether resting orders are charged at all.

Can you arbitrage Kalshi and Polymarket? Not on our numbers. Our five paper positions, run through both fee schedules, show the two-leg cross-venue book at minus $3.48 on $500 deployed before the market moves. Separately, buying YES and NO on one Polymarket market under $1.00 does not occur: Gamma prices sum to exactly $1.00 by construction.

Which is better for sports, Kalshi or Polymarket? On fees alone, Polymarket. Sports sits in Polymarket's 0.05 band against Kalshi's flat 0.07, so at a $0.50 entry the break-even move is 2.50% against 3.50%. We have not measured book depth at size on either venue, so we make no claim about which one actually fills a large sports order better.

Who decides how a market settles on Kalshi versus Polymarket? On Kalshi, Kalshi does. The Massachusetts Attorney General's complaint describes it as writing the contract rules and determining the basis for settlement with no independent intermediary, and we have found zero appeal paths outside Kalshi. On Polymarket we have not verified the resolution mechanism against a primary document, so we do not describe it.

Are cheap contracts a safer way to compare the two venues? No, the opposite. Break-even as a share of stake is rate × (1 − P), so a $0.90 contract needs a 0.70% move on Kalshi while a $0.03 contract needs 6.79%. Rounding compounds it: one Kalshi contract at $0.03 pays a full cent, which is 4.9 times the unrounded fee and 33% of stake.

What happens to my open positions if my state bans prediction markets? Nobody has published the mechanics, and we have not verified them either. What is documented is the timing: at least six state-level actions landed during 2026, and Minnesota's SF 3432 took effect on 1 August 2026. Ask the venue in writing before a known effective date rather than after it.