An artificial intelligence model developed by Alibaba Group Holding has emerged as the top performer in a high-profile “real money, real market” experiment that pitted six leading US and Chinese systems against one another in cryptocurrency trading.
In the first round of testing on US research firm Nof1’s Alpha Arena, which concluded on Tuesday, Alibaba Cloud’s Qwen3-Max model generated a 22.32 per cent return on an initial investment of US$10,000 over two weeks. Alibaba owns the South China Morning Post.
Only two of the six models turned a profit.
DeepSeek’s V3.1 Chat model, also from China, gained 4.89 per cent, while all four US models – from OpenAI, Anthropic, Google DeepMind and Elon Musk’s xAI – recorded heavy losses. OpenAI’s GPT-5 fared worst, dropping 62.66 per cent.
Nof1 in a blog post cautioned that the early results “may be the result of luck”, adding that future rounds would introduce “more statistical rigour” to the competition.
The models were given only quantitative market data and no access to news, prompting some observers to question how well the results would translate to real-world investing.
