.png)
In the rapid evolution of the crypto world, predictive markets are being redefined.
In the past five years, the total value locked (TVL) of global predictive markets has grown from approximately $120 million (2020 Q4) to over $900 million (2024 Q3), with a compound annual growth rate (CAGR) of more than 80%.
At the same time, the on-chain deployment of AI and zero-knowledge (zk) technologies has risen from zero cases in 2022 to more than 150 public implementations in 2024, with over 70% adopting zk-SNARK/zk-STARK for data privacy and verifiability.
These macro trends are driving predictive markets from “speculation” to “cognition”:
Mainly focused on single events such as political elections and sports competitions, with fragmented liquidity and severe information asymmetry.
AI models capture on-chain behavior, social signals, and news flows in real time, combined with zk-proofs to achieve privacy-preserving multidimensional information pricing, enabling markets to become generators of “intelligent consensus.”
From early event betting and information trading to today’s deep integration with AI and zk technologies, the market is moving from “speculation” to “cognition” and from “outcome pricing” to “intelligent consensus.”
Merin was born precisely at this key turning point as a next-generation predictive market infrastructure project, aiming to reshape the industry’s underlying logic through technological innovation.
From Speculation to Cognition, From Price to Consensus
Merin’s core vision is to build an open, secure, and scalable Intelligent Predictive Network, enabling individuals, institutions, and even AI agents to participate in information pricing with higher efficiency and privacy protection.
Predictive markets are not only a branch of financial innovation but also a new way for humans and machines to understand the future together.
Merin’s system architecture consists of four key layers:
Smart contract layer
Responsible for event creation and automatic settlement, ensuring transparent rules and decentralized execution for all predictive markets;
Oracle layer
Adopts zk-proofs and a multi-node adjudication mechanism to ensure that external data is accurately brought on-chain without relying on a single trusted source;
Liquidity layer
Uses an adaptive market-making model (zk-AMM) and cross-market capital reallocation mechanisms to solve the liquidity shortage and fragmented depth issues common in traditional predictive markets;