Kalshi vs Polymarket US vs Robinhood: one of these three is not an exchange

Two exchanges and a broker standing in front of them

The card grid above puts three products side by side, which makes them look like three of a kind. They are not. Kalshi and Polymarket US are CFTC-designated contract markets. Robinhood is not an exchange at all: Robinhood Derivatives, LLC is a registered futures commission merchant, and the event contracts in the app are routed to KalshiEX, ForecastEx or Rothera Exchange and Clearing. No designated contract market is registered under the Robinhood name.

Rothera is the one that confuses the picture, because Robinhood part-owns it — a joint venture with Susquehanna took 90% of the former MIAXdx on 20 January 2026, and Robinhood describes the exchange as independently managed. Part-owning an exchange is not being one. Every contract in the app still settles under somebody else's rulebook.

Two things follow, and they are the whole reason this page exists.

First, "Robinhood versus Kalshi" is mostly Kalshi versus itself. When your order goes to KalshiEX you are holding a KalshiEX contract, resolved by Kalshi's markets team under Kalshi's rules. The only variable is what you paid to reach it.

Second, you do not choose the destination, and the destination sets half the bill. Robinhood's fee schedule quotes the KalshiEX exchange fee as $0.01 per contract per side and, for ForecastEx and Rothera, points at each exchange's own schedule instead of naming a number. So for two of its three routes the second layer is unquoted, and the routing decision that determines it is not yours. ForecastEx has its own card and its own schedule; you will not know in advance whether it applies.

Polymarket US is not one of the three destinations. The amended designation order of 24 November 2025 vacated the clause that barred futures commission merchants from intermediating on that exchange, so a brokered route there is now possible in principle — but Robinhood's own disclosure names KalshiEX, ForecastEx and Rothera and nothing else. Today, Polymarket US is a place you go directly or not at all.

What the broker layer costs, priced across the curve

Both layers are published, so this is arithmetic rather than opinion. On a KalshiEX-routed fill, per contract:

• Kalshi, direct. 0.07 × contracts × price × (1 − price), rounded up to the cent. A curve that peaks at the midpoint and collapses towards both ends.
• Robinhood. A commission of k × price × (1 − price) with k at 10%, or 5% with the $5-a-month Gold subscription, capped at $0.01 per contract — plus the exchange's own $0.01 per contract, per side, which is flat.

At 50 cents the difference nearly disappears. Kalshi's curve is at its own maximum of 1.75 cents; Robinhood's commission has been flattened to a cent by the cap, and the exchange cent sits on top, for 2.00 cents. A quarter of a cent per contract — about 14% more. Gold changes nothing here: 10% and 5% of expected earnings at the midpoint are 2.5 cents and 1.25 cents, and both are above the ceiling.

The gap is widest where the cap starts to bind, around 11 and 89 cents. There Kalshi charges about 0.7 cents and Robinhood charges the full 2.00 — roughly three times as much, and 1.3 cents a contract in absolute terms, the largest spread anywhere on the curve.

Outside that band the commission falls away but the exchange cent does not. At 10 cents the same fill is 1.90% of the payout through Robinhood against 0.63% direct. At 1 cent it is about 1.10% against 0.07% — fifteen times. This is the part that surprises people: Robinhood's cost has a floor of one cent per contract per side that nothing reduces, while below about 17 cents Kalshi's entire taker fee is less than that floor on its own.

Gold narrows the middle of that picture without closing it. At 10 cents it halves the commission from 0.90 to 0.45 cents, taking the gap against Kalshi from 1.27 cents to 0.82. The flat exchange cent is untouched at every price.

Price a round trip and it doubles on both sides: 4.00 cents a contract through Robinhood at the midpoint against 3.50 direct, because the exchange fee applies to opening and closing trades and so does Kalshi's taker fee. What a trade actually costs works through the four incompatible units these numbers arrive in.

What the extra cent buys, stated fairly

It buys three real things, and they are worth something to a particular reader.

One account and one app. If you already hold a Robinhood brokerage account there is no second identity check to pass — the derivatives account is an approval layered on the one you have, granted on trading experience, investment profile and state of residence. Compare that with starting cold. Kalshi wants a driver's licence or passport photographed live through its app; an upload from the camera roll or a picture of a screen is rejected, name and date of birth must match the document exactly, and no PO boxes. Polymarket US requires the identity check to be complete before you can deposit at all, with manual review quoted at three to five business days when the automatic pass fails.

One tax story. Event contracts sit beside the rest of the account, and Robinhood covers reports and statements for them as an ordinary topic. Kalshi issues documents at year end and itemises every transaction for download. Polymarket US's card sets tax_reporting: false, and nothing on it describes a year-end document at all.

Access to three exchanges' listings through one screen, which is genuinely the widest set of questions of the three, precisely because it is an aggregate.

Against those, the things the cent does not buy: the app is mobile-only for trading and is not on Robinhood Legend; there is no API, no WebSocket and no SDK; orders are limit-only, either immediate-or-cancel or good-til-date expiring at 3 AM ET the next calendar day, so nothing you place survives a week; there is no order book on screen; and Robinhood publishes no list of the states it serves. Maryland and Nevada are named — and Robinhood's own two help pages disagree about whether Maryland's block covers sports event contracts or all of them — with a warning that further state restrictions apply and a pointer to the app. The block also follows your current physical location, not only your residence.

The honest summary is that Robinhood loses on every axis except convenience, and convenience is a real axis. It is just one that costs a cent a contract, per side, forever.

Kalshi and Polymarket US are two exchanges, not one brand with two doors

This is the second thing the grid hides. The two look interchangeable and agree on almost nothing except the taker fee.

The taker fee is the non-difference. Coefficients of 0.07 and 0.0695 produce 1.75 cents and 1.74 cents on a 50-cent contract. Anyone choosing between these two on the headline rate is choosing on a hundredth of a cent.

The maker side is where they separate. Kalshi charges makers on 160 of the 11,078 series it published on 19 September 2026 — 1.4% — and nothing on the rest. Polymarket US pays the maker, at a −0.0125 coefficient credited at the point of the fill: 0.31 cents a contract at the midpoint. Rest both legs of a round trip and Kalshi charges you nothing while Polymarket US hands you 0.62 cents a contract. That is the single largest structural difference between them.

It comes with an operational cost that belongs in the same paragraph: every Thursday from 02:00 to 04:00 ET Polymarket US cancels every open order on the exchange, answers 503 on the API, and reopens with empty books. A resting-order strategy there is rebuilt weekly. Kalshi's card records no equivalent window.

Volume is rebated on one and not the other. Polymarket US pays takers back weekly off the previous calendar month's notional — 10% from $250,000, 25% from $1 million, 50% from $10 million — and will place you in a tier immediately on verifiable proof of trailing-30-day volume elsewhere. That takes an effective taker rate of 1.74% down to 1.57% and eventually 0.87%. Kalshi publishes no volume rebate on event contracts at all — its own volume tiering applies to perpetual futures, which are a different product priced in basis points.

One exchange varies the rate by series and the other does not. Kalshi's coefficient carries a per-series multiplier, and its schedule lists 155 non-standard series: some add maker fees, and some — several crypto and political series among them — sit at a multiplier of zero and carry no fee on either side. Worth checking before you assume 1.75%, in both directions. Polymarket US has one coefficient for the whole exchange, no per-category table, and the same rate on sports as on politics, which is the more predictable arrangement and never the cheaper one.

Both resolve as the exchange, and on both the decision is final. The difference is whether you can read the source first. Kalshi exposes a machine-readable settlement_sources array — a name and a URL — on every event through its public API, so the document your money depends on is a field you can pull before you take the position. Polymarket US names primary, secondary and tertiary sources inside each contract's terms, and its rulebook lets the company modify contract specifications after you hold the position, including the settlement date, the payout condition, the interpretation of resolution criteria and the designation of alternative source agencies. Rule 10.4 lets named officers determine the final outcome at sole discretion and reverse an obvious error; Rule 10.5 settles contracts about a person who dies or is incapacitated at last traded prices, final and not subject to review. Neither exchange offers an appeal — but one publishes the source as data and the other as a clause you can be moved off. Who decides the outcome covers what that distinction is worth.

Their listings cannot be lined up, and this page will not pretend otherwise. Kalshi's card reports a composition — the category field aggregated over 200 live open events on 19 September 2026, dominated by economics at 57 and sports at 55. Polymarket US's card reports totals on the same day: 2,611 open sports events and 1,167 open politics events. A ratio built from those two figures would be invented. What both cards do support is narrower and more useful: Polymarket US's own getting-started page still said sports only, with politics "coming soon", while its API was returning a thousand open politics markets. Read the API or the app, not the welcome page.

Only one of the three is reachable from outside the United States. Kalshi has served international traders since March 2026, with 55 restricted jurisdictions as of the exchange notice effective 22 June 2026 — the UK, Canada, Australia, New Zealand, France, Italy, Ireland, India and China among them, with Germany, the Netherlands, Spain, Japan and South Korea not on it. The restriction is a warranty you sign rather than a documented block, and the funding rail is the narrower half: international users have debit card, wire from $1,000, or crypto, and no ACH, PayPal or Venmo. Polymarket US and Robinhood are US-residents-only and say so.

Getting money out differs more than getting it in. Polymarket US returns withdrawals to the original funding source only, first-in-first-out against the deposits that funded them, after a three-to-four-day clearing wait on cards and ACH; promotional credit never leaves at all. Kalshi makes ACH free both ways, charges up to 2% on card deposits, and documents wire withdrawals as not currently supported below $500,000.

What the fee field carries, now that it carries both layers

Worth saying plainly, because this page was written when the field said something else. economics.fees_taker is normalised across the whole site as the effective taker cost at 50 cents, as a percentage of the $1.00 a winning contract pays. Kalshi reads 1.75 and Polymarket US reads 1.74, and in both cases that is the entire charge.

Robinhood now reads 2, which is also the entire charge on the route this catalogue can price: a cent of commission, where the cap binds, plus a cent of KalshiEX exchange fee. It used to read 1

• the commission alone - and a reader sorting this category by taker fee would have seen the most expensive of these three ranked first as the cheapest.

One caveat the single number still cannot carry. The 2 is the KalshiEX route. Robinhood's own fee schedule quotes no number for ForecastEx or Rothera and points at each exchange's schedule instead, so on those two routes the second layer is unquoted - and the routing decision is not yours.

Which to take, and when to switch

Take Kalshi unless you have a specific reason not to. It undercuts the brokered route at every price — by a quarter of a cent at the midpoint, and by roughly three times below 17 cents, where its entire taker fee is less than the flat exchange cent Robinhood passes through. It is the only one that publishes its settlement source as a field you can read before committing, the only one whose API key can be issued without owning an iPhone — see Kalshi API — and the only one of the three open outside the United States. The two things to accept are that there is no appeal from a resolution, and that kalshi.com refuses non-browser clients, so the July 2026 fee schedule behind those numbers had to be read through a proxy on 20 September 2026 rather than opened directly.

Switch to Polymarket US the moment your resting orders are a strategy rather than a convenience. That is the one axis on which it wins outright: 0.31 cents a contract received where Kalshi charges nothing, plus a taker rebate ladder that starts at $250,000 a month and reaches half the fee at $10 million. Budget for rebuilding the book every Thursday morning, for withdrawals that go back only the way they came, and for a rulebook that permits contract specifications to change under a position you already hold.

Take Robinhood only under one condition: you buy near the midpoint, a handful of times a year, and the alternative is not opening an account at all. At 50 cents the entire premium is a quarter of a cent a contract, and that is a fair price for one app, one statement and no second document check. Leave the moment any of four things becomes true — you buy away from the middle, you want to rest an order, you want programmatic access, or you want to know which exchange is holding your position. The third and fourth are not available at any price there, and the first two get expensive quickly.

If none of these is reachable where you are standing, the offshore venues are a different set of trade-offs entirely — see Polymarket vs Limitless vs Myriad and why a venue is unavailable where you are.