Prediction markets and surebets in Genesis API: endpoints, filters and fields
In short
- Six endpoints: prediction arbitrage and platform status from the Start plan; analytics, smart money, large trades and the trader leaderboard on Pro and Max.
- Arbitrage bundles come priced through the order books with fees: trade size, profit in $ and %, APR, days to settlement and risk codes.
- Sport mode returns surebets across 13 platforms, pre-match only. When a leg is a bet at odds, profit_usd is null because the stake limit of that bet is unknown.
Genesis API serves prediction market data through six GET endpoints: arbitrage bundles across Polymarket, Kalshi, Limitless, Predict.fun and Opinion, surebets across 13 platforms, platform status, analytics, smart money, large trades and the trader leaderboard. It is the data behind the Prediction arbitrage, Analytics and Smart Money pages, as JSON. Build a bot, a dashboard or alerts on it; trades and bets are placed by you on the platforms.
Endpoints and plans
| Endpoint | Returns | API plans |
|---|---|---|
/predictions/arbitrage | Arbitrage bundles and surebets | Start, Pro, Max; llm and sport modes — Pro, Max |
/predictions/platforms | Platform status | Start, Pro, Max |
/predictions/analytics | KPIs, price moves, money flows | Pro, Max |
/predictions/smart-money | Large trades, whales, insider patterns | Pro, Max |
/predictions/trades | Polymarket trade tape | Pro, Max |
/predictions/traders | Top 300 Polymarket traders by PnL | Pro, Max |
On Trial all six return 403 plan_restricted; on Start, so do requests with mode=llm and mode=sport. The response format is covered in the guide to responses and errors, the quota in the guide to limits.
Prediction arbitrage: modes
The mode parameter picks which bundles /predictions/arbitrage returns.
mode | What you get | kind | Plans |
|---|---|---|---|
exact | The same event on two platforms, the default | cross | Start, Pro, Max |
dutch | A set of outcomes of one event on one platform | dutch | Start, Pro, Max |
llm | Event pairs verified by AI | cross | Pro, Max |
sport | Surebets on matches | sport | Pro, Max |
Profit is already computed through the platforms' order books with fees. A request for exact bundles with at least 1% profit that settle within 30 days:
curl -G "https://api.genesisarbitrage.com/v1/predictions/arbitrage" \
-H "X-API-Key: gns_live_XXXXXXXXXXXXXXXX" \
-d mode=exact -d min_profit_pct=1 -d max_days=30
Filters
| Parameter | Type | What it does |
|---|---|---|
mode | string | exact (default), dutch, llm or sport |
category | string | Category, as in the category field |
platform | string | At least one leg on this platform; several — comma-separated |
min_profit_pct | number | Minimum profit, % |
min_profit_usd | number | Minimum profit, $; rows with profit_usd = null are skipped |
min_apr | number | Minimum APR, % |
max_days | number | At most this many days to settlement |
sport, league | string | Sport and league, for mode=sport |
Filters are optional and combine. Write the platform as in legs[].platform: polymarket, kalshi, fonbet. An unknown mode returns 400 bad_request.
A bundle and its legs
An illustrative example: a mode=exact row from the documentation, shortened.
{
"id": "x:exact:kalshi:KXFRENCHPRES-27-JCAS~polymarket:0x8c1f…",
"kind": "cross",
"match": "exact",
"title": "Will Jean Castex win the 2027 French presidential…",
"category": "Elections",
"ends_at": 1843308000,
"days_left": 562.87,
"size_usd": 107175.08,
"fees_usd": 28.08,
"profit_usd": 153.69,
"profit_pct": 0.14,
"apr_pct": 0.1,
"shares": 107356.85,
"risks": ["rules_differ", "end_mismatch"],
"legs": [
{
"platform": "polymarket",
"side": "yes",
"outcome": "Jean Castex",
"price_avg": 0.0013,
"cost_usd": 140.3,
"fee_usd": 5.6,
"url": "https://polymarket.com/event/next-french-presidential-…"
},
{
"platform": "kalshi",
"side": "no",
"outcome": "No",
"price_avg": 0.997,
"cost_usd": 107034.78,
"fee_usd": 22.48,
"url": "https://kalshi.com/events/KXFRENCHPRES-27"
}
]
}
| Field | Meaning |
|---|---|
id, first_seen, updated_at | Stable id, when the bundle appeared and was updated |
kind, match | cross, dutch or sport; match is exact, llm or sport |
title, category | Event or market and its category |
ends_at, days_left, starts_at | Settlement time, days to settlement, match start |
size_usd, fees_usd, shares | Trade size and fees in $, number of shares |
profit_usd, profit_pct, apr_pct | Profit in $ and %, APR; apr_pct is null under one day |
edge_cents, risks | Price edge in cents per share, risk codes |
legs[].platform, market, outcome | The leg's platform, market and outcome |
legs[].side | yes or no — what to buy; bet — a bet at odds |
legs[].price_best, price_avg, price_worst | Best, average and worst price walking the book |
legs[].shares, cost_usd, fee_usd, url | The leg's shares, cost, fee and market link |
A leg price is a fraction of a dollar: a share returns $1 if its outcome wins, so 0.997 is 99.7¢. If both platforms settle the same way, one leg wins whatever happens, and shares is the amount you receive. rules_differ means the platforms' settlement rules may differ; end_mismatch means their settlement dates differ.
In the example 107,356.85 − 107,175.08 − 28.08 = $153.69 of profit, or 153.69 / 107,203.16 = 0.14%. The money is locked for 562.87 days until settlement, so the annual rate comes to 0.14 × 365 / 562.87 ≈ 0.1%. The max_days filter screens such bundles out.
sport mode: surebets
A sport row is one market of one match: 2 or 3 legs on different platforms that together cover all its outcomes. There are 13 platforms: Polymarket, Kalshi, Predict.fun and 10 betting platforms — Fonbet, Betcity, Olimpbet, Marathon, 5051X, Melbet, Winline, BC.Game, EGB and Dafa Sports. Pre-match only: once a match starts, the surebet leaves the list.
An illustrative example: a mode=sport row from the documentation with recalculated prices.
{
"id": "s:8812:RT:TOT:2.5:",
"kind": "sport",
"title": "Arsenal — Chelsea",
"sport": "soccer",
"league": "Premier League",
"teams": ["Arsenal", "Chelsea"],
"market_type": "TOT",
"period": "RT",
"line": 2.5,
"starts_at": 1794700800,
"size_usd": 1000,
"profit_usd": null,
"profit_pct": 2.39,
"legs": [
{
"platform": "polymarket",
"side": "yes",
"outcome": "Over 2.5",
"covers": ["O"],
"price_avg": 0.505,
"odds": null,
"stake_unit": 0.505,
"currency": "USD"
},
{
"platform": "fonbet",
"side": "bet",
"outcome": "Under 2.5",
"covers": ["U"],
"odds": 2.12,
"stake_unit": 0.471698,
"currency": "RUB"
}
]
}
| Field | Meaning |
|---|---|
sport, league, teams, flags | Sport, league, teams, match tags |
market_type | WIN — result, HCP — handicap, TOT — total, BTTS — both teams to score, CS — correct score |
period, line, score | Settlement period, handicap or total line, score for CS |
legs[].covers | Outcomes the leg covers: O — over, U — under |
legs[].odds | Decimal odds; null for a YES or NO purchase |
legs[].stake_unit | Stake share: how much to put on the leg to get $1 if it wins, fees included |
legs[].currency | The platform's currency |
For a bet at odds, stake_unit = 1 / odds: 1 / 2.12 = 0.471698. The stake units across legs add up to 0.505 + 0.471698 = 0.976698, below 1, hence a profit of 1 / 0.976698 − 1 = 2.39%. To split your own amount, use leg stake = amount × stake_unit / Σ stake_unit.
size_usd is the amount of purchases on prediction platforms that their order books can absorb. $1,000 at 0.505 buys 1,000 / 0.505 ≈ 1,980.2 YES shares of "Over 2.5". For "Under 2.5" to return the same, the Fonbet stake is 1,980.2 × 0.471698 ≈ $934.06.
Fonbet takes bets in rubles (currency = RUB): convert the amount at your own rate, since the exchange is not part of profit_pct. profit_usd is null here because the betting platform does not publish its stake limit.
Platform status
/predictions/platforms shows how fresh the arbitrage data is. It takes no parameters; data holds two lists:
{
"platforms": [
{
"platform": "kalshi",
"mode": "rest",
"quotes_at": 1794675858,
"full_scan_at": 1794675012,
"markets": 26187,
"hot_markets": 188
}
],
"bookmakers": [
{
"bookmaker": "fonbet",
"title": "Fonbet",
"ok": true,
"events": 1840,
"matched_events": 212,
"updated_at": 1794675855,
"currency": "RUB"
}
]
}
platforms[].mode— the prediction platform's feed mode:ws,rest, ordownwhen the feed is not working.quotes_atandfull_scan_at— the time of the latest quotes and of the latest full market scan;marketsandhot_markets— markets in total and hot ones.bookmakers[]— betting platforms:ok— whether collection works,eventsandmatched_events— events in total and matched with other platforms,currency— the betting currency.
Bundles with a platform in down or with ok = false are best skipped: their prices may have moved.
Analytics, smart money, trades and traders
Four endpoints on Pro and Max mirror the screens from the Analytics and Smart Money guides.
| Endpoint | Parameters | What's inside |
|---|---|---|
/predictions/analytics | none | KPIs, price moves over 1h and 24h, money flows, markets settling soon, new markets |
/predictions/smart-money | none | KPIs, money flows, insiders, concentration, smart_vs_price |
/predictions/trades | period, wallet, category, side, min_usd, insider | Polymarket trades, newest first |
/predictions/traders | period: 30d (default), 7d, all | Top 300 Polymarket traders by PnL |
The insider label describes the behaviour pattern of a wallet and proves nothing: a new address of an experienced trader can trade the same way.
Trade tape parameters; period sets the threshold:
| Parameter | Type | What it does |
|---|---|---|
period | string | live (default) — trades from $1,000, from $100 for a smart wallet; today and 7d — takers from $5,000 |
wallet | string | Trades of one wallet |
category | string | Market category |
side | string | buy or sell |
min_usd | number | Minimum trade amount, $ |
insider | boolean | true — insider-labelled trades only |
curl -G "https://api.genesisarbitrage.com/v1/predictions/trades" \
-H "X-API-Key: gns_live_XXXXXXXXXXXXXXXX" \
-d period=today -d side=buy -d min_usd=10000
{
"ts": 1794675812,
"wallet": "0x3a1c…f09e",
"tags": ["whale", "smart"],
"title": "Will Bitcoin reach $150k in 2026?",
"outcome": "No",
"side": "buy",
"price": 0.82,
"shares": 30487.8,
"usd": 25000,
"role": "taker",
"insider": false,
"bid_now": 0.81,
"ask_now": 0.83
}
usd = price × shares: 0.82 × 30,487.8 ≈ 25,000. bid_now and ask_now are the outcome's current prices: following this trade now costs 0.83, 1¢ more than the trader paid.
/predictions/traders gives PnL for all time, 30 and 7 days (pnl_usd), roi_pct, won out of resolved and the pnl_series sparkline. Put an address from the leaderboard into the trade tape's wallet.
How to use it
Check your plan and the platforms
/account/usageshows your plan,/predictions/platforms— how fresh the quotes are.Request bundles with filters
min_profit_pctandmax_dayscut the noise and long-dated bundles.Check risks and rules
With
rules_differ, read the settlement rules of both platforms.Poll no faster than your plan's freshness
Use an interval of at least
meta.freshness_secand send theETagback inIf-None-Match: a304response costs no quota.Trade yourself
Open the market from
legs[].urland check the price in its order book.
The same exact request in Python:
import os
from datetime import datetime, timezone
import requests
API = "https://api.genesisarbitrage.com/v1"
HEADERS = {"X-API-Key": os.environ["GENESIS_API_KEY"]}
params = {"mode": "exact", "min_profit_pct": 1, "max_days": 30}
r = requests.get(f"{API}/predictions/arbitrage", headers=HEADERS,
params=params, timeout=10)
if r.status_code == 403:
raise SystemExit("Endpoint or mode is not part of your API plan")
r.raise_for_status()
body = r.json()
ts = datetime.fromtimestamp(body["meta"]["updated_at"], timezone.utc)
print(f"Bundles: {body['meta']['count']}, data as of {ts:%H:%M:%S} UTC")
for b in body["data"]:
print(f"{b['profit_pct']:.2f}% · ${b['profit_usd']:.2f} · "
f"{b['days_left']:.0f} d · {b['title'][:40]}")
for leg in b["legs"]:
print(f" {leg['platform']}: {leg['side']} "
f"{leg['outcome']} at {leg['price_avg']}")
if b["risks"]:
print(" risks:", ", ".join(b["risks"]))
On the site, the same bundles are covered in the Prediction arbitrage and Sport mode guides; the platforms' official APIs — in the articles on the Polymarket and Kalshi APIs and on betting odds via API.
Questions and answers
Can I buy contracts or place bets through Genesis API?
No. Genesis API only reads data and places no orders: you trade and bet yourself on the platforms.
Why is profit_usd null for a surebet?
At least one leg is a bet at odds, and the platform does not publish its stake limit. The profit percentage is known; work out each leg's amount from stake_unit.
Does the API include live surebets?
No, pre-match only. Why live is riskier is covered in the article on pre-match vs live.
Which plan do I need for surebets?
Pro or Max: on Start, mode=sport returns 403 plan_restricted. Prices are on the Genesis API Pricing page.