What happened?
Two Chinese AI models — DeepSeek and Alibaba’s Qwen — crushed Western rivals in a live crypto trading contest, turning $10,000 starting stakes into roughly $22,900 and $20,850 respectively in under two weeks. DeepSeek returned about 126% and Qwen about 108%, while OpenAI’s GPT-5 and Google’s Gemini 2.5 Pro lost around 57–60% of their portfolios. The contest uses identical real-time market data and runs through November 3, with models using different tactics like DeepSeek diversifying across assets and Qwen concentrating heavily on Ether.
Who does this affect?
Crypto traders and funds are directly affected because this shows AI-driven strategies can move real money fast and outperform some established models in volatile markets. AI developers and trading firms face pressure to adapt, partner, or compete with Chinese models that are proving effective in real-world trading. Exchanges, regulators, and retail investors also need to pay attention since bigger, faster algorithmic bets can amplify price swings and invite closer oversight.
Why does this matter?
If Chinese models keep outperforming, institutional flows may shift toward AI-driven trading strategies and increase volatility in the tokens those bots favor, like Ether. Bold forecasts from models like DeepSeek — for example projecting ETH to $12k–$15k or big gains for ADA and XRP — could become self-fulfilling as traders chase those calls and push prices. That dynamic would boost liquidity at times but also raise systemic risk, forcing exchanges and regulators to rethink rules around algorithmic crypto trading.