Arbitrage Calculator

Enter the two opposite outcomes from any two venues. We tell you whether the prices cross, the exact stake split, and the P&L if both legs fill at the prices you entered. Either leg can be a sportsbook line or a Polymarket / Kalshi price. Our published MLB snapshot study found quoted crossings in 0.36–0.39% of its June 21–23, 2026 samples. Those sampled quotes do not establish that both legs could have filled.

Outcome A

Outcome B

Total Stake ($)
Total you want to put across both legs
Prediction-Market Fee per Share (¢)
Flat per-share assumption applied to legs entered in ¢. Enter the costs for your contract; 0 excludes them.
Positive Calculated Payoff
Equal payout if both legs fill as entered and settle as opposite outcomes of the same event.
No Arbitrage
No positive payoff at these inputs. At a sum of 100%, the calculated payoff is break-even before any omitted costs.

How to Split the Stake

Stake on Outcome A
Stake on Outcome B
Calculated Payout
Profit
ROI

The split equalizes payout under the entered assumptions. Rounding, stake limits, one-sided fills and different settlement rules can change the result. This calculator does not place orders.

Where cross-venue arbs come from

The calculator tests whether the two effective implied probabilities sum to less than 100%, then divides the stake to equalize the payouts. This requires two exhaustive, mutually exclusive outcomes with matching settlement rules. A sampled price difference does not verify executable size or simultaneous fills.

Check the contracts
  • Same event, participants, date and settlement definition
  • Mutually exclusive and exhaustive outcomes
  • Void, overtime and cancellation rules on both legs
Don't forget
  • Fees and the bid/ask eat thin arbs — include the ¢ fee
  • Prices move; one leg can vanish before you fill the other
  • Limits and resolution rules can differ across venues

Find the gaps automatically

ZenHodl's Edge Finder compares available venue prices against model probabilities. A model-price gap is a different measurement from a two-leg arbitrage calculation. Our per-sport CLV scorecard publishes the observed result—including negative values—without claiming that this one measure proves an edge.

Get ZenHodl Weekly

Dataset releases, research notes, and public results, including corrections.

Tuesday mornings. No spam.

Free tools for sharper betting and prediction-market trading.