Is Kalshi cheaper than sportsbooks?
How often a prediction-market price beat the best US-regulated sportsbook on the same bet, measured on our stored prices.
Usually not, once its trading fee is paid. From Aug 12 to Oct 28, 2026, we compared 4,759 Kalshi prices across 1,327 games with the best price among the US-regulated sportsbooks we capture for that sport, on the same side, market and line. Each comparison uses the sportsbooks' last capture before the start and Kalshi's order book as we stored it at that capture or up to 3 hours earlier. Kalshi's ask plus its taker fee was the better price 14.1% of the time (669), the same price 5.7% (271), and the sportsbook was better on the other 3,819.
Before the fee, Kalshi's ask was the better price 85.5% of the time, so the fee is what usually decides it.
The sportsbook side is not the whole US market in every sport. For MLB and NBA our price history carries only DraftKings and FanDuel (2,786 of the 4,759 prices), and Kalshi was better 16.2% of the time there; against the wider set of books in NFL, NHL, college football and soccer it was better 11.0%.
Sportsbooks compared, by sport. MLB and NBA: DraftKings and FanDuel only. NFL: BetMGM, BetRivers, Caesars, DraftKings, FanDuel, Fanatics and theScore Bet. NHL, college football and soccer: Bally Bet, BetMGM, BetRivers, Caesars, DraftKings, FanDuel, Fanatics, Hard Rock Bet, betPARX and theScore Bet.
By sport and market. Each price is one side of one market in one game.
| Sport | Market | Prices | Kalshi better, after fee | Same price | Kalshi better, before fee |
|---|---|---|---|---|---|
| MLB | Moneyline | 807 | 6.7% | 5.9% | 79.7% |
| MLB | Spread | 1,140 | 16.1% | 7.5% | 94.1% |
| MLB | Total | 602 | 25.2% | 12.0% | 95.3% |
| NFL | Moneyline | 65 | 4.6% | 4.6% | 83.1% |
| NFL | Spread | 88 | 13.6% | 12.5% | 85.2% |
| NFL | Total | 122 | 20.5% | 5.7% | 88.5% |
| NBA | Moneyline | 97 | 16.5% | 3.1% | 91.8% |
| NBA | Spread | 96 | 34.4% | 4.2% | 92.7% |
| NBA | Total | 44 | 29.5% | 9.1% | 90.9% |
| NHL | Moneyline | 206 | 5.8% | 1.9% | 84.5% |
| NHL | Spread | 398 | 11.3% | 1.8% | 80.9% |
| NHL | Total | 410 | 15.6% | 2.0% | 80.7% |
| College football | Moneyline | 244 | 12.3% | 2.9% | 73.0% |
| Soccer | Moneyline | 440 | 6.1% | 1.6% | 72.0% |
| All | All markets | 4,759 | 14.1% | 5.7% | 85.5% |
For comparison only: Polymarket’s international exchange is not available to US bettors and is not part of Best Price. Moneylines, after Polymarket’s published sports taker fee (5% of P(1-P), in effect since Jul 10, 2026, before every price here), by sport.
| Sport | Market | Prices | Polymarket better, after fee | Same price | Polymarket better, before fee |
|---|---|---|---|---|---|
| MLB | Moneyline | 2 | too few to rate | too few to rate | too few to rate |
| NFL | Moneyline | 58 | 34.5% | 5.2% | 77.6% |
| NBA | Moneyline | 96 | 46.9% | 4.2% | 85.4% |
| NHL | Moneyline | 198 | 28.8% | 7.6% | 85.9% |
| College football | Moneyline | 276 | 41.7% | 2.9% | 73.9% |
| Soccer | Moneyline | 446 | 34.5% | 2.2% | 74.4% |
| All | Moneyline | 1,076 | 36.4% | 3.7% | 77.6% |
One comparison is one side of one market in one game: a team to win, a team at a spread, or the over or under at a total. For each game we take the last sportsbook capture before the scheduled start in which at least 2 US-regulated books priced the game. Offshore and international books never count. A side needs at least 2 of those books pricing it, and the best of them is the sportsbook price. Which books that is depends on the sport, because our price history does not carry every book in every sport:
- MLB and NBA: DraftKings and FanDuel only.
- NFL: BetMGM, BetRivers, Caesars, DraftKings, FanDuel, Fanatics and theScore Bet.
- NHL, college football and soccer: Bally Bet, BetMGM, BetRivers, Caesars, DraftKings, FanDuel, Fanatics, Hard Rock Bet, betPARX and theScore Bet.
Spreads and totals are compared only where the exchange lists a contract at exactly the books’ line. The median capture was 27 minutes before the start.
The exchange price is the ask, what a buyer pays, from the order book as we stored it at the sportsbook capture or before it: the newest stored book at or before that moment, and only if it was captured no more than 3 hours earlier. The two prices can therefore be up to 3 hours apart; we store an exchange book only when it changes, so an older book is the book that was still standing. It counts only when the book is two-sided with a spread of 5c or less. No midpoint and no price derived from the other side’s bid is ever used. On a Kalshi moneyline we price each team only on its own “wins” contract, never on the “no” side of the other team’s contract, which can be the cheaper way into the same result; that makes the Kalshi share a conservative one. Kalshi is compared after its taker fee of 7% of P(1-P) per contract, rounded up to the cent, the same price Troxler’s Best Price uses. On a large order Kalshi rounds the fee once for the whole order rather than per contract; at that per-contract rate Kalshi was the better price 24.4% of the time.
Ties go to the sportsbook. Both prices are compared as American odds, the exchange price rounded to a whole number. When they are equal the exchange is not counted as better, because a bettor gains nothing by switching and Best Price shows the book in that case. Counted for the exchange instead, ties would put Kalshi at 19.8% and Polymarket at 40.1%.
Polymarket here is the international exchange, not Polymarket US, and it is not available to US bettors. It is shown for comparison only and is not part of Troxler’s Best Price, which includes only prices a US bettor can take. It is compared after the taker fee Polymarket publishes for sports markets, 5% of P(1-P) per share (Polymarket fee schedule, read Oct 29, 2026). That rate has applied since Jul 10, 2026 (Polymarket changelog: raised from 3% that day), before the first price compared here, so it applies to every Polymarket comparison on this page. Polymarket is measured on moneylines only.
Sample. 1,890 games had a qualifying sportsbook capture. A share below 30 prices is printed as “too few to rate”. Sides whose stored exchange book was more than 3 hours older than the capture were left out. Putting those 1,936 Kalshi sides back in, scored the same way, gives 13.2% (885 of 6,695) in place of 14.1%. For Polymarket, 835 sides back in give 37.0% (708 of 1,911) in place of 36.4%. Measured on Oct 29, 2026 from prices captured Aug 12 to Oct 28, 2026.
Powered by Kalshi Research