Six leading AI large models compete in crypto trading, with DeepSeek and Grok consistently ranking top two in returns

Jinse Finance reported that the artificial intelligence research lab nof1, which focuses on financial markets, launched a large model trading test called Alpha Arena on the 18th. The test uses six mainstream AI large models (GPT-5, Gemini 2.5 Pro, Grok-4, Claude Sonnet 4.5, DeepSeek V3.1, Qwen3 Max), each of which is allocated $10,000 in real funds on Hyperliquid, with identical prompts and input data. As of press time, DeepSeek and Grok both have returns exceeding 14%, ranking in the top two, while Gemini 2.5 Pro has lost 42.57%.