An unprecedented AI trading battle has erupted in the crypto space: nof1 brings six major AI models to compete against each other in trading capabilities with "real money" on-chain.

The rules of the game are simple and straightforward:

The platform collects real-time data on six mainstream cryptocurrencies, including BTC, ETH, and SOL, as well as historical charts, allowing six top models, including Grok, DeepSeek, Claude, and GPT, to independently analyze the market and execute trades.

Each participant starts with $10,000 and independently makes trading decisions such as placing orders, closing positions, and leveraging on the Hyperliquid perpetual contract platform. The one with the highest account balance at the end wins.

This is not a paper-based simulation; it's a real battle!

Here are a few of my personal insights:

1) The nof1 AI is positioned to use Alpha Arena as a benchmark to measure AI investment capabilities, so it primarily consists of general AI models rather than models specifically fine-tuned for trading. This allows for an objective reflection of the trading capabilities of AI LLMs.

2) There is no use of complex "ensemble learning" methods such as voting, weighting, or Multi Agents communication, because once that happens, it essentially becomes a competition of quantitative teams' strengths. The arena aims to let AI capabilities be fully autonomous.

3) The reason for conducting "fair competition" on HyperliquidX in real trading rather than through API connections to exchanges undoubtedly emphasizes the necessity of having all trades on-chain, fully transparent, and traceable. This avoids uncontrollable factors present in exchange environments. This is key to why Perp Dex will become a mainstream trading platform in the future; what if agents can replace human players in trading?

4) Public competitions of AI LLMs' trading capabilities are not rare. I know several cutting-edge AI + Crypto projects are doing similar research, such as having AI large models determine LP price ranges, allowing AI large models to autonomously execute DAO management, and even enabling AI large models to capture MEV opportunities, etc. A major trend in the future is to make AI capabilities as combinable as possible within crypto-native protocols like DeFi.

5) The market has long awaited killer applications in the DeFAI direction, and there is still much imaginative space for involving LLMs in on-chain games. For example, the recently discussed "prediction markets" are a great stage for AI training, and there have already been many cases of agents autonomously betting and making profits.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。