Kalshi and Polymarket together processed over $44 billion in trading volume in 2025, and by April 2026 their combined monthly handle had reached $24 billion, topping the average monthly figure run through legal US sportsbooks, according to a Pew Research Center analysis of data from The Block. Kalshi alone pulled in $263.5 million in fee revenue last year; annualized, that number now sits above $1.5 billion.
For any founder watching those figures, the natural next question is whether something similar can be built from scratch. The short answer is yes. The harder question, one most people skip until it's too late, is what that actually requires legally and operationally.
The mechanics behind the model
A prediction market runs as an exchange for binary contracts. A user buys a YES contract at 60 cents if they think an event has a 60% chance of occurring. If they're right, the contract pays $1. If wrong, it expires at zero. Another user takes the opposite side, believing the probability is lower. The platform matches them and collects a fee from both parties when the trade settles.
Crucially, the platform itself holds no position. It doesn't set odds, doesn't keep a margin and has no exposure to the outcome. That's the structural argument operators use to distinguish themselves from sportsbooks, where the house sets the line, absorbs the risk and pays winners from losers' stakes.
Kalshi's fee structure charges both makers and takers on a probability-weighted per-contract basis. Fees peak at 50/50 odds and taper off as contracts move toward certainty in either direction. At high volume, those fractions add up fast, which is exactly what the 2025 revenue figures reflect.
Whether the exchange model survives regulatory scrutiny depends entirely on jurisdiction. In the US, the Commodity Futures Trading Commission regulates event contracts under the Commodity Exchange Act, which is why Kalshi spent years in a legal fight with the CFTC before winning the right to list political event contracts in 2024. Polymarket, operating offshore and restricting US users, sits under a different framework entirely.
Any founder looking to replicate the model needs to resolve those questions before writing a line of code, not after. The compliance path, licensing jurisdiction and fee structure are inseparable from the product itself.
This article is for informational purposes only and does not constitute financial or legal advice.
