Kalshi Arbitrage Scanner
Two scanners in one: political markets and sports game lines, both Kalshi vs Polymarket. Every row carries Kalshi's real per-contract fee and what crossing actually costs — so a divergence that cannot be traded reads that way.
PRO EDGE
Same event, different prices on two venues — and what it actually costs to cross
What is this?
"Arbitrage" = buying low on one venue and selling high on another. The scanner compares Kalshi and Polymarket on the same event — those two venues only.
Example: Polymarket prices an event at 52¢. Kalshi prices the other side at 44¢ (implying 56¢ for a win). Buying YES on one and NO on the other pays exactly $1.00 whichever way it lands. Whether that is profitable is a fee question, not a gap question: Kalshi charges 0.07 × P × (1−P) per contract — 1.75¢ at 50¢, about 3.5pp round trip on two legs — so a 2-3pp divergence is underwater before slippage. Every row shows the known cost of crossing and whether the divergence clears it.
Real-World Example
→ The Gap Appears
Breaking news drops at 2pm. Polymarket updates fast — a market moves from 40¢ to 65¢. Kalshi hasn't updated yet — still at 43¢. The same event, a 22-point divergence, and far clear of the ~2-3pp it costs to cross.
That size is rare. Most days the two venues agree within 0-2pp, which does NOT clear the cost line — the scanner labels those untradeable rather than dressing them up as edge.
✅Action: Read the cost line before the signal. A divergence under it cannot be traded profitably, whatever the label says.
Bottom line: When two venues price the same event differently, one of them is wrong — but only the divergences that clear the real cost of crossing are worth acting on.
Full guide →Quick Answer
the Arb Scanner flags live price divergence between Kalshi and Polymarket on the same political or sports market. It flags ARB only when the provable riskless pair nets at least 2¢ per contract after both venues' real costs, and WATCH when the divergence clears the known Kalshi-leg hurdle by at least 2pp without a provable pair — the raw gap alone is never the signal. Kalshi's real fee is about 1.75¢ per contract at 50¢ (~3.5% of stake, ~3.5pp round trip on two legs), so each row shows the known cost of crossing and whether the divergence clears it. Confirm resolution terms and book depth before taking a position.
If you're asking whether Kalshi vs Polymarket arbitrage is real: yes, but rarely free money. The scanner flags ARB only when the riskless pair still nets ≥2¢ per contract after both venues' real fees and spreads — fees and thin books are priced in, not ignored. What it cannot price is resolution criteria: two contracts on the same event can settle on different terms, and that check is still yours before taking a position.
Caveat: a flagged divergence on screen is not the same as an executable trade. Always verify resolution terms and book depth on both platforms before sizing the position.
What Is the Cross-Platform Arb Scanner?
The Arb Scanner is a live arbitrage detection tool that scans Kalshi and Polymarket simultaneously for the same event priced differently. When Kalshi says 55% and Polymarket says 42%, one of those prices is wrong. The scanner finds that gap automatically and tells you if it's large enough to trade.
Unlike manual comparison, the scanner normalizes prices across both platforms, matches markets describing the same event by title-token coverage, and scores every matched pair on the raw price gap — the difference in implied probability between the two platforms — with a symmetric KL-divergence reading shown alongside as a secondary signal.
How It Works
Run the scanner and it fetches live data from both Kalshi and Polymarket. Matched pairs whose divergence clears real venue costs by 2pp (WATCH), or whose riskless pair provably nets 2¢+ per contract (ARB FLAGGED), are surfaced immediately with direct trade links to both platforms. The scanner does not save state — every run reflects live prices.
When to Use This Tool
Use it when you want to check for executable arb opportunities across Kalshi and Polymarket before placing a trade. Check the scanner at market open, after major news events, and before any contract with high volume on one platform. Pair with the EV Calculator and Kelly Criterion to size any flagged position. New to the two exchanges? See our Kalshi vs Polymarket comparison for legality, fees, and which side to trade. Pro subscribers get direct trade links and real-time refresh.
Go Pro · 26% off your first year
Every edge the engines find, the moment they find it
Pro opens every engine on the site — the cross-market arb scanner, the Mispricing Scanner, Thee Oracle, and the silver, gold, oil and bitcoin edge models that price commodity contracts in real time.
$111 for the first year, then $150/yr. Code applies automatically. Ends Dec 31. Monthly is $14.99/mo.
Every signal graded in public, losers left on the board. See the record →
Related Tools
How the scanner works
The scanner fetches live markets from Kalshi and Polymarket every time you run it and normalizes all prices to implied probability. The two lanes then pair differently. POLITICS runs a title-match algorithm — scoring the share of the shorter title's content words that appear in the longer one, with year/month guards so markets about different dates never pair — and any pair clearing 60% token coverage is scored. SPORTS does no title matching at all: it pairs on team, exact Eastern-time game date and opponent, so a sports pair is either correct or absent. Games already underway are dropped, because Kalshi leaves game markets open through live play and an in-play price is one venue lagging the other rather than a real divergence.
Signals are scored net of cost, not on the raw gap. ARB means the provable riskless pair nets at least 2¢ per contract after both venues' real costs, and requires both books quoting two sides.WATCH means the divergence clears the known Kalshi-leg hurdle by at least 2pp but the riskless pair is unprovable or nets under 2¢. Kalshi's real fee is 0.07 · P · (1−P) per contract, which is 1.75¢ on a 50¢ contract: about 3.5% of stake, and roughly 3.5pp round trip on two legs. Each row therefore carries the known cost of crossing on the Kalshi leg — half its quoted spread plus that fee — and says whether the divergence clears it. Polymarket prices are the international (Gamma) book; US traders route to Polymarket US, a separate exchange with its own order book. The scanner also shows a symmetric KL-divergence reading as a secondary measure of how structurally inconsistent the two prices are (a 10pp gap at 50/40 is a different beast than 10pp at 90/80).
What to do when you see ARB FLAGGED
First: verify that both markets resolve on the same event, on the same date, with compatible terms. Kalshi and Polymarket sometimes phrase questions differently even when they're describing the same event. Second: check bid-ask spreads on both sides — a flagged arb with wide spreads may not be executable at the displayed midpoint. Third: check liquidity. If one side has thin order books, your fill price will differ from the displayed probability.
If all three checks pass, the trade is straightforward: buy YES on the cheaper platform and hedge on the more expensive one, or leg into both YES positions if the combined cost is under $1.00.
Frequently Asked Questions
What is the best Kalshi arbitrage scanner?
PredictionMarketsPicks' Arb Scanner is built for the one edge most scanners miss: cross-platform gaps. Instead of only surfacing within-Kalshi mispricings, it lines up the same event on Kalshi and Polymarket, normalizes both to implied probability, and scores the gap net of what crossing actually costs. ARB means the provable riskless pair nets at least 2¢ per contract after both venues' real fees and spreads, and it requires both books quoting two sides — so it is never claimed off an invented spread. WATCH means the divergence clears the known Kalshi-leg hurdle by at least 2pp but the riskless pair is unprovable or nets under 2¢. It covers macro, politics, and five sports, shows a KL-divergence secondary read, and flags candidates to verify rather than promising free money: the engine can price fees, spreads and depth, but it cannot read resolution criteria, so confirming both contracts settle on identical terms is still yours. Polymarket prices are the international (Gamma) book; US traders route to Polymarket US, a separate exchange with its own order book.
What is a Kalshi arbitrage scanner?
A Kalshi arbitrage scanner compares live Kalshi prices against another platform — here, Polymarket — on the same event, and scores it net of cost rather than on the raw gap. A midpoint-to-midpoint divergence is not a profit: crossing costs land on both legs, so a wide-looking gap can still be underwater. Each row therefore shows the net figure per contract pair after both venues’ real books and real fees, and that number is frequently negative — which is the useful part. ARB is reserved for a pair that nets at least 2¢ per contract after those costs with both books quoting two sides; WATCH is a directional divergence clearing the known Kalshi-leg hurdle by at least 2pp where the riskless pair is unprovable or nets less. A Polymarket leg under $500 of 24-hour volume never signals at all. A flagged gap is a candidate to investigate, not free money: confirm both contracts resolve on identical terms and that the books are deep enough to fill before taking a position.
Quick answer: is Kalshi vs Polymarket arbitrage actually exploitable?
Rarely, and the fee math is why. Kalshi charges 0.07 x C x P x (1-P) per contract, which is 1.75 cents on a 50-cent contract — about 3.5% of stake, and roughly 3.5 percentage points round trip across two legs. So a 2-3pp divergence is underwater before you account for anything else. On top of that, the headline gap is a midpoint-to-midpoint comparison and nobody trades at the midpoint: you cross the spread on entry. Every row therefore shows what crossing actually costs on the Kalshi leg (half its real spread plus its real fee) and whether the divergence clears it. A divergence below that number cannot be tradeable. Above it, still verify resolution terms and order-book depth before acting.
What is cross-platform arbitrage in prediction markets?
Cross-platform arbitrage occurs when the same event is priced differently on Kalshi and Polymarket. If Kalshi prices an event at 55¢ and Polymarket prices it at 42¢, one of those prices is wrong. Buying the cheaper side and hedging on the more expensive side — or buying YES on both if their combined cost is under $1.00 — captures the gap as profit regardless of outcome.
How does the Arb Scanner detect price gaps between Kalshi and Polymarket?
The two lanes match differently, because the problems are different. POLITICS pairs on title-token coverage: at least 60% of the shorter title's content words must overlap, with year and month guards so markets about different dates never pair. SPORTS does not use title matching at all — it pairs deterministically on team, exact Eastern-time game date, and opponent, which is why a sports pair is either right or absent rather than fuzzily scored. Both lanes then normalize prices to implied probability and compare the two venues' midpoints. Games already underway are dropped: Kalshi leaves game markets open through live play, so an in-play price is one venue lagging the other, not a divergence. A symmetric KL-divergence reading is shown alongside as a secondary read on how structurally inconsistent the two prices are.
When does the Arb Scanner flag ARB vs WATCH?
The signals are cost-aware, not raw price gaps. ARB means the riskless pair — buy the cheaper venue's YES, buy the other venue's NO — provably nets at least 2 cents per contract after both venues' real fees and spreads; it is only ever claimed when both venues expose a two-sided order book, and each row shows how much size the edge held for. WATCH means the fair-value divergence clears the known cost of crossing on the Kalshi leg — half its real quoted spread plus its real per-contract fee — by at least 2 percentage points, but the riskless pair cannot be proven (one venue quoted a single price) or nets under 2 cents. ALIGNED is everything else, including any divergence that fails to clear known costs. Rows on markets with negligible 24-hour volume never signal — a price gap against a dead book is stale pricing, not disagreement. When a venue quotes one price with no book, the scanner reports the divergence and the known cost rather than inventing the missing side, which would flatter every number.
Does the Arb Scanner cover sports like the NBA and World Cup?
Yes — the scanner covers 5 sports: NBA, NHL, MLB, NFL, and the 2026 World Cup, matching Kalshi game markets against Polymarket on the same matchup. Sports gaps are scored on the same implied-probability gap as macro markets, and trade routing sends you to Kalshi first. Season gating means a sport only appears when its games are live, so the active list shifts through the year — the World Cup slate, for example, opens in mid-June.
Do I need a paid plan to use the Arb Scanner?
The Arb Scanner is a Pro tool at $14.99/mo, which unlocks direct trade links, the Sports tab, and real-time refresh across Kalshi and Polymarket. The free tier still lets you check edge on any single contract through the EV Calculator, but live cross-platform gap detection and one-click trade routing are Pro-only. Annual billing runs $150/yr, which works out to two months free versus paying monthly.
How fresh are the prices the Arb Scanner shows?
The scanner saves no state — every run pulls live prices, and the sports gap feed refreshes on its own 15-minute cycle. Each page load reflects the current Kalshi and Polymarket order books, not a cached snapshot. Because cross-platform gaps can close in seconds once order flow reacts, re-check both books at the moment of execution rather than trusting a price you loaded minutes earlier.
Get The 7 Oracles' daily edge — subscribe free
No spam. Unsubscribe anytime.