Practical answers about cross-venue prediction market arbitrage. Numbers below are from the live universe of arbs we're tracking right now.
The same event listed at two different prices on two different venues. You sell on the expensive side and buy on the cheap side. Your profit is locked in the moment both legs fill — it doesn't matter who wins the election, game, or whatever the market is about.
Example: Polymarket says "Marco Rubio wins" is selling at $0.34, Kalshi says the same outcome is selling at $0.30. You sell on Polymarket (collect $0.34), buy on Kalshi (pay $0.30), pocket $0.04 per share locked.
Each Telegram alert gives you two clickable URLs and the exact prices to use. Click the first link, place the buy order on one venue. Click the second link, place the sell order on the other. Total time: about 60 seconds. Your profit is locked the moment both legs fill.
In practice, arbs persist for hours — often a day or more — not seconds, so you have time to place both legs.
You're not racing HFT bots. These are structural cross-venue mispricings that persist because Kalshi (US retail) and Polymarket (crypto-native) have different user bases and nobody is automatically bridging them.
It depends on which arbs are live, how much depth they have, and which venues you can trade — and honestly, it's usually modest. Cross-venue arbs are structural: they exist because Kalshi and Polymarket have different user bases, so they're real and fillable, but they're not lottery tickets.
A real example measured on the board in September 2026 — a snapshot, not an illustration, and not necessarily still live today: Chelsea to win the Premier League, quoted on both venues. Walking the actual order books, $52.84 of locked-in profit on $3,471 of committed capital — 1.5% net of fees. It resolves with the season, about nine months out, so that is roughly 2.0% annualized, and your capital is tied up for the whole of it. The top of the book shows a wider 2.2%, but the edge thins as you fill, so the walked figure is the one that matters.
That is the honest shape of this market today: edges are real and small, and most of what is available resolves months away rather than days. Short-dated spreads pay far better per day when they appear, which is what the alerts are for.
We don't publish a bankroll–to–ROI table, because your realized return depends on what's live when you're funded, your venue access, and fill depth — not on a fixed formula. Starter is $99/year; you can cancel anytime and keep access to the end of the term.
Yes — just like any prediction market position. When you take an arb, capital stays in your venue accounts until that market resolves.
Most of what's live is long-dated, and you should know that before you fund rather than after. Measured across the board as it stands: the median arb locks capital for about 26 months, and weighting each arb by the capital it actually absorbs gives roughly 22 months. The shortest are around three months, a few are open-ended with no fixed resolution date, and nothing on the board today resolves inside 30 days. Think of it like a CD: the capital is committed until the market resolves and your payout is fixed at the moment you take the arb — not a savings account you can pull from anytime.
Kalshi charges ~7% of net winnings. Polymarket International charges effectively zero (its taker fee is currently waived), but Polymarket US charges about 6% on the same shape — so which venue an arb routes through changes what you keep. Polygon gas is ~$0.50 per transaction.
The dashboard deducts the correct per-venue fee from every edge it shows, so what you see is already net of fees.
For cross-venue arbs, yes — you need funded accounts on both Polymarket and Kalshi to execute the two legs.
For NegRisk arbs (same-event multi-outcome trades inside Polymarket), you only need a Polymarket wallet.
Kalshi requires US residency + completed KYC. Polymarket is blocked in some jurisdictions including Ontario and parts of Europe. Verify access at both venues before subscribing.
Polymarket and Kalshi for cross-venue mispricings, plus Polymarket NegRisk events for same-event multi-outcome arbs. More venues are planned but not promised.
Telegram alerts cover cross-venue arbs only. NegRisk baskets are scanned and shown on the dashboard, but they are not pushed to Telegram — we hold them below the delivery confidence bar until the basket-pricing model has a track record behind it. Check the dashboard for those.
The scanner runs continuously. New arbs that clear the confidence gate go out on the next delivery cycle. Each arb alerts exactly once. Once an arb has been sent it will not be sent again, even if its edge widens later — one notification per opportunity, no reminders. Watch the dashboard if you want to track an arb you chose not to take.
No. We detect and deliver. You make all trading decisions, fund your own venue accounts, and place your own orders. We give you the signal; you decide what to do with it.
Three real ones:
1. Capital lockup. You can't pull money out before each market resolves. Anywhere from a few weeks to 2+ years per arb.
2. Execution slippage. Orderbooks tick. By the time you click through to place your order, the price may have moved 1-3 cents from what the alert showed. We display orderbook depth so you can size accordingly, but small slippage is normal.
3. Thin-inventory days exist. Some days surface 10+ new arbs, others surface 1 or 2. We can't manufacture inventory we don't see.
Speed bots chase micro-arbs on the same venue. Tiny edges that close in milliseconds, requiring colocated servers and custom infrastructure.
Converge finds cross-venue structural mispricings where the same outcome is priced differently on Polymarket vs Kalshi. These edges are real but usually modest, and they persist for hours to days because they're driven by different user bases and venue plumbing friction, not latency. You're not racing bots — you're seeing what most retail traders never look for.
Yes. Cancel through Stripe at any time. No contracts, no commitments, no questions.