4 Projects Remixing the Prediction Market Model

Polymarket's trailblazing efforts brought onchain forecasting into the mainstream, turning prediction markets from a niche corner of crypto into a widely recognized category.

But the story doesn’t end with incumbents like Polymarket and Kalshi. A new wave of projects, both emerging and established, is experimenting with how far the model can be stretched in different directions. Today, let's spotlight four of them. 👇

🎮 Frenzy Finance

Most prediction markets ask you to forecast events that resolve over weeks or months. Frenzy Finance asks you to forecast the next 10 seconds.

Built on Base and launched in early March, Frenzy is an onchain prediction protocol where you click cells on a trading grid to bet on where a price will land in the next price tick window. Correct predictions mint USDC-backed credits to your wallet, and incorrect ones burn them.

There are some interesting wrinkles here, like how the FrenzyPool vault is used to collateralize predictions. Dig into the docs to learn more about these intricacies. Overall, though, what I like most about this project is how it feels more like an arcade game than a complex finance app. I can envision it becoming popular with casual users.

Current status: Frenzy is on Base and Base Sepolia, but the frontend isn't open to U.S. traders. If you want to try Frenzy without real stakes first, the app's testnet mode demo lets you use free test USDC so you can start immediately learning the ropes.

🗳️ Fact Machine

"Opinion markets" are trending lately, with platforms like Confetti and Cue making the case that there's appetite for betting on subjective outcomes rather than objective ones. A newer entrant worth noting is Fact Machine, which is taking a fresh approach.

That's because opinion markets so far have had to contend with whale users, who can time their liquidity in such a way so as to dominate outcomes regardless of what people actually believe. As an antidote for this dynamic, Fact Machine splits out its participants into two roles: bettors and voters.

In other words, bettors trade on what they think the majority opinion will be, while voters (who must pass proof-of-personhood for sybil resistance) determine the actual market outcomes via polls. Then the side with the most votes wins, and bettors who predicted the majority correctly earn payouts. For their part, voters are incentivized via points, which can be rolled back into market bets or redeemed for cash at the end of each week.

Whale manipulation has always struck me as the biggest specter hanging over prediction markets in general, so it'll be interesting to see how Fact Machine fares with this new defensive mechanism in the opinions niche. Godspeed.

Current status: Fact Machine has an invite-only alpha live on Solana centered around paper trading competitions with real cash prizes. It's worth having on your radar for when the alpha or the beta opens up more.

🤖 Context V2

I wrote about Context back in October as one of the more compelling prediction market upstarts on Base. Since then, the team has shipped its V2, and the vision has sharpened considerably: they're going all-in on AI agent users.

Specifically, the new V2 landed yesterday, and it brought with it a full agentic stack, featuring a suite of 20+ tools, a CLI with 50+ commands, and a Claude Code plugin. The idea is that any AI agent should be able to create markets and manage a prediction portfolio through a single API.

There are some other notable changes here, too, like how the V2 launch marked a switch from an AMM model to an order book system. But either way, the question remains: are agents the main future of prediction markets? If so, Context is giving us an early example of what that could look like.

Current status: Context V2 is open to try on Base right now, and humans can dive in too. Explore the docs to consider the various possibilities available to you.

⚡ Hyperliquid HIP-4

HIP-4 isn't a standalone platform, but it is a new primitive being built into Hyperliquid's core infrastructure, and the prediction market implications are big.

Hyperliquid's existing builder-deployed perps (HIP-3) can't handle the instant step-function updates that event contract resolutions require. HIP-4, dubbed Outcomes, will solve for this via fully collateralized binary contracts that settle cleanly to 0 or 1, and without leverage, liquidations, or continuous oracle dependencies.

Accordingly, Hyperliquid traders will soon be able to hedge and speculate on event predictions without ever leaving the Hyperliquid platform, an attractive addition to Hyperliquid's growing suite of onchain money legos.

Current status: The first BTC/HYPE binary market went live on testnet last week, suggesting the mainnet rollout isn't far off now. Keep your eyes peeled for official news on the full release.