Skip to main content
Trepa introduces a new generation of prediction platforms built around one idea: the world runs on numbers, so prediction platforms should reward numerical accuracy. We call these precision predictions. Instead of predicting yes/no, up/down, or over/under, Trepa lets you express exactly what you think will happen to prices, earnings, macroeconomic indicators, athlete performances, election margins, and more. For the first time, the payoff is proportional to how close you get to the real outcome, not to whether you chose the correct side. We answer a simple question: Who can predict the future with the highest precision? Trepa aims to become a valuable input layer for the next era of prediction data, including binary prediction markets by merging
  • an expressive and precision-based payoff mechanism
  • with the elicitation of high-resolution signals.
Core idea: the closer you are to the truth, the more you earn.

​
Who is Trepa for?

Trepa is built for people who want to predict numbers, not just yes or no. If you’ve ever tried to predict numerical outcomes on Polymarket or Kalshi, you know the pain:
the “Yes” token is already at 99 cents because institutional AI models and well-funded bots moved first. You were too slow and because numerical outcomes have been hacked into coarse ranges to enable yes/no betting, you cannot express your informational edge properly. But with Trepa, you finally get to monetize your decimal-level view. You will likely find Trepa interesting if you’re a superforecaster who values skill-based competition, measurable performance, and social recognition.