I spent part of this week digging into Polymarket’s raw performance data to do something most people never bother to do with crypto claims: I went straight to the source and pulled the numbers myself instead of trusting a headline. What I found genuinely surprised me.
Among prediction markets with transparent, on-chain, officially verifiable revenue, Polymarket has quietly become one of the highest-earning applications in all of crypto in 2026, and the gap between it and its closest verifiable competitors isn’t close. It’s not a modest lead. It’s closer to a different league entirely.
I want to be precise about the scope of this comparison from the start, because it matters. This article is not a ranking of every prediction market on earth. Large platforms like Kalshi operate off-chain, as private companies, and don’t publish audited revenue through any public, checkable source the way on-chain protocols do. So this is specifically a look at platforms whose fees and revenue are tracked transparently on-chain, verifiable by anyone, at any time, through official data. Within that specific group, the numbers tell a genuinely lopsided story, and I want to walk through exactly how lopsided, how the next two biggest verifiable earners compare, and where you can go to check every figure yourself.
Why This Comparison Has To Be Scoped This Way
Before getting into the numbers, I think it’s worth explaining why “on-chain, officially verifiable revenue” is the right lens here rather than a limitation I’m working around. A platform like Kalshi processes enormous trading volume and is likely generating substantial real-world revenue. But because its trades settle off-chain through a centralized company structure, there’s no public smart contract data for anyone, including analytics platforms like DefiLlama, to independently verify. Any number attached to Kalshi’s revenue right now is an outside estimate, not something the company has officially disclosed or that can be checked against a live, public data source.
Polymarket, Predict Fun, and PancakeSwap Prediction are different. Every trade, every fee, and every dollar of revenue they generate moves through public smart contracts, which means the numbers aren’t self-reported or estimated, they’re independently observable by anyone who wants to look. That’s the specific category I’m comparing here, and I think it’s the fairest, most honest way to talk about revenue leadership in this space without mixing verified data with outside guesswork.
The Number That Stopped Me
According to on-chain data tracked by DefiLlama, the industry-standard analytics platform that most of crypto relies on for exactly this kind of verification, Polymarket has generated roughly $117.85 million in cumulative protocol revenue so far in 2026. This isn’t trading volume, and it isn’t gross fees paid by users. It’s actual protocol revenue, the portion of fees Polymarket keeps after paying out maker rebates, holding rewards, and referral incentives, the same kind of “net revenue” metric you’d use to evaluate a real business.
What makes that number land even harder for me is the starting point. Polymarket generated essentially zero revenue in 2025, deliberately operating without trading fees while it scaled global liquidity and user growth. The moment it turned fees on, revenue went from nothing to well over a hundred million dollars inside a single year. Over just the past 30 days alone, the platform pulled in $16.29 million in protocol revenue on $76.42 million in gross fees, and DefiLlama’s own trailing-year annualized estimate now puts Polymarket on pace for roughly $163.6 million in revenue and $377.5 million in fees going forward.
The Second-Place Verifiable Earner Isn’t Particularly Close
The next biggest earner among on-chain, officially trackable prediction markets is Predict Fun, a BNB Chain-based prediction market. Predict Fun has generated approximately $6.95 million in cumulative protocol revenue so far this year, according to its official DefiLlama tracking page at https://defillama.com/protocol/predict-fun. That’s a real, legitimate number for a platform of its size, built off roughly $21.45 million in cumulative fees and a trailing 30-day revenue figure of just over $1 million.
But laid next to Polymarket’s $117.85 million, the scale difference is stark. Within this specific, verifiable category, Polymarket has out-earned Predict Fun by a factor of roughly seventeen to one this year alone. Predict Fun has carved out a legitimate niche, offering DeFi-powered yield on open positions to solve the “idle capital” problem that many prediction markets struggle with, and it’s backed by serious institutional money, having raised a strategic round from YZi Labs and Susquehanna earlier this year. It’s simply operating at a fundamentally different scale than the market leader, at least among the platforms whose numbers can actually be checked.
Third Place Among Verifiable Platforms Belongs To A Familiar Name
Rounding out the top three, still strictly within the on-chain, officially verifiable category, is PancakeSwap Prediction, the built-in prediction market feature inside the PancakeSwap ecosystem on BNB Chain. According to its official DefiLlama revenue page at https://defillama.com/protocol/pancakeswap-prediction, PancakeSwap Prediction has generated approximately $3 million in cumulative revenue so far in 2026, built from quarterly totals of roughly $1.37 million in Q1, $1.01 million in Q2, and just over $600,000 so far in the current quarter.
What I find genuinely interesting about PancakeSwap Prediction’s model is what happens to that revenue once it’s collected. Every dollar of protocol revenue is automatically routed into buying back and burning CAKE, PancakeSwap’s native token, which DefiLlama’s own methodology documentation confirms directly. There’s no separate treasury allocation or team discretion involved, the mechanism is fully automated and transparent, and the cumulative total funneled through that buyback program since inception now sits above $16.7 million. It’s a smaller platform than Polymarket or Predict Fun in absolute terms, but the revenue-to-token-holder pipeline is about as direct as it gets in this sector.
What This Concentration Actually Tells Us
Stacking these three platforms side by side, the picture that emerges is one of extreme concentration at the very top of the on-chain prediction market sector specifically. Among the more than one hundred prediction market protocols DefiLlama tracks with public, verifiable revenue data, Polymarket alone accounts for the overwhelming majority of all measurable revenue generated in 2026, with Predict Fun and PancakeSwap Prediction trailing well behind in a distant second and third place, and the rest of the tracked protocols splitting what little remains.
I think that concentration reflects something real about how this specific corner of the market has evolved. Polymarket built enormous trust and liquidity over years of operating without fees, expanded aggressively into sports and political markets, and secured genuine regulatory standing in the United States through its CFTC-designated Polymarket US arm. It also landed a $2 billion combined strategic investment from Intercontinental Exchange, the parent company of the New York Stock Exchange, a deal that gave the platform both capital and institutional legitimacy that smaller on-chain competitors simply don’t have access to. When Polymarket finally flipped its fee switch on, all of that accumulated trust and volume converted into revenue almost immediately, at a scale nothing else in the on-chain sector could match.
Why Kalshi Isn’t In This Ranking, And What That Does And Doesn’t Mean
I want to underline this point clearly, because it’s the most important caveat in the entire piece. Kalshi is excluded from this comparison, not because its revenue is necessarily lower than Predict Fun’s or PancakeSwap Prediction’s, but because it isn’t measurable through the same official, on-chain channel the other three platforms use. Kalshi frequently leads the broader prediction market sector in raw notional trading volume, and given its scale, particularly in sports markets, its actual revenue could plausibly be far higher than any of the three platforms discussed here.
The honest summary is this: among prediction markets whose revenue is transparently and officially verifiable on-chain, Polymarket leads by a wide margin, with Predict Fun and PancakeSwap Prediction a distant second and third. That is a genuinely different claim from ranking every prediction market on earth by total revenue, and I don’t want this piece read as making that broader claim. Where Kalshi would land in an all-inclusive ranking simply isn’t something I can state with a verifiable, official source behind it right now.
Where This Leaves The Sector
Pulling all of this together, I think the clearest takeaway is that among the on-chain, transparently verifiable segment of the prediction market industry, this is increasingly a story about one platform rather than a genuinely competitive field. Polymarket’s roughly $117.85 million in 2026 revenue isn’t just a lead over Predict Fun and PancakeSwap Prediction within this category, it’s larger than the combined total of literally every other on-chain prediction market DefiLlama tracks. For a platform that spent years deliberately forgoing revenue to build market share, that’s about as clear a validation of the strategy as you could ask for, within the specific, checkable data available.
Whether that gap narrows as competitors like Predict Fun continue raising capital and building out their own fee infrastructure is worth watching. But based on the official, on-chain numbers available right now, anyone can pull up, Polymarket isn’t just winning the on-chain prediction market revenue race in 2026. It’s not particularly close.
Disclosure: This is not trading or investment advice. Always do your research before buying any cryptocurrency or investing in any services.
Follow us on Twitter @themerklehash to stay updated with the latest Crypto, NFT, AI, Cybersecurity, and Metaverse news!