Find arbitrage and +EV prices with Odds-API.io

What you need

Code status: The Python in this tutorial was run against the example responses published in the Odds-API.io docs (checked 3 November 2026), not against a live API key. Live responses can carry extra fields; the code only reads documented ones.

1. Fetch arbitrage opportunities

bookmakers is required; minProfit is a percentage; includeEventDetails=true adds teams. Margins are calculated on netOdds where exchange or prediction-market fees apply ^[1]. The snippet recomputes each margin from the legs as a sanity check.

# reuses get() from the Python tutorial
arbs = get("/arbitrage-bets", bookmakers="Bet365,SingBet,Betano",
           minProfit=1, includeEventDetails="true", limit=50)

for a in arbs:
    # re-check the margin ourselves from the legs (netOdds where fees apply)
    s = sum(1 / float(l.get("netOdds") or l["odds"]) for l in a["legs"])
    mine = (1 - s) / s * 100
    ev = a.get("event", {})
    print(f'{ev.get("home")} v {ev.get("away")} | {a["market"]["name"]} {a["market"].get("hdp", "")}'
          f' | API {a["profitMargin"]:.2f}% vs recomputed {mine:.2f}%')

On the docs' sample (Betfair Exchange over 2.102 net, Bet365 under 2.05) the recomputed margin is 3.78% against the API's 3.79% — a rounding difference.

2. Scale the stake split

optimalStakes are for totalStake (1,000 by default); scale every leg by the same factor and stay under each leg's maxStake ^[1]. Cross-check with the Arbitrage calculator.

def scale_stakes(arb, bankroll):
    """Scale optimalStakes to a bankroll, keeping every leg under its maxStake."""
    factor = bankroll / arb["totalStake"]
    for st in arb["optimalStakes"]:
        leg = next(l for l in arb["legs"]
                   if l["bookmaker"] == st["bookmaker"] and l["side"] == st["side"])
        if leg.get("maxStake"):
            factor = min(factor, leg["maxStake"] / st["stake"])
    return [{**st, "stake": round(st["stake"] * factor, 2),
             "potentialReturn": round(st["potentialReturn"] * factor, 2)}
            for st in arb["optimalStakes"]]

3. Fetch value bets (+EV)

/value-bets needs one bookmaker; fair odds come from a sharp-book panel and EV = fair probability × odds − 1 ^[2]. Because expectedValue is a fraction, you can back out the fair probability. Use the Expected value (+EV) calculator and Kelly criterion calculator on any row.

vbs = get("/value-bets", bookmaker="Bet365", sport="football",
          includeEventDetails="true")

for vb in sorted(vbs, key=lambda v: v["expectedValue"], reverse=True)[:10]:
    side = vb["betSide"]
    odds = float(vb["bookmakerOdds"][side])
    fair_p = (vb["expectedValue"] + 1) / odds      # from EV = p x odds - 1
    ev = vb.get("event", {})
    print(f'{ev.get("home")} v {ev.get("away")} | {vb["market"]["name"]} {side} @ {odds}'
          f' | EV {vb["expectedValue"]:.2%} | implied fair p {fair_p:.1%}')

4. Know the limits

Arbitrage results are prematch only and removed as soon as a price moves ^[1]; bookmakers may limit accounts that take arbs ^[4]. Value is an estimate against a sharp consensus; track Closing line value to see whether it holds up ^[5].

Related glossary terms and tools

• Arbitrage
• Sure bet
• Expected value
• Kelly criterion
• Closing line value
• Arbitrage calculator
• Expected value (+EV) calculator
• Kelly criterion calculator
• Live odds in Python
• No-code odds alerts with n8n or Make
• Odds data in Claude and ChatGPT via MCP
• Prediction-market prices vs bookmaker odds
• Odds-API vs SportsGameOdds

Frequently asked questions