AI Chatbots Excel in Crypto Trading Competition
Grok and DeepSeek have shown impressive performance among the top AI chatbots in the realm of cryptocurrency trading. They seemed to accurately identify the market bottom before it began to recover, which could offer significant advantages to their users.
In a recent trading competition that caught considerable attention, Grok 4 and DeepSeek emerged as the standout generative AI chatbots. According to the blockchain analysis platform Coinglass, DeepSeek took the lead with unrealized profits totaling $3,650, while Grok followed closely with about $3,000 in unrealized gains.
In third place was Claude’s Sonet 4, which recorded profits of $2,340. Qwen3 Max rounded out the top four with $784 in profits since the competition began.
However, not every chatbot managed to turn a profit. OpenAI’s ChatGPT 5 reported around $2,800 in unrealized losses, and Google’s Gemini 2.5 Pro was in a similar predicament, with losses mounting to approximately $3,270 at the time.
Chatbots like Grok and ChatGPT have gained traction among cryptocurrency traders, as they assist in predicting potential altcoin uplifts and help pinpoint entry points that can guard against exit liquidity challenges.
While such AI tools can help day traders quickly identify shifts in investor sentiment from social media and technical indicators, they shouldn’t be solely relied upon for making independent trading decisions.
Grok 4’s Predictive Insights
Grok 4 notably managed to predict the local market bottom just as recovery was on the horizon. Jaz Azan, founder and CEO of the AI firm Stealth, pointed this out in statements made on October 11.
The Grok chatbot placed several leveraged long positions across two major cryptocurrencies. It particularly leveraged a 20x long position in XRP, a 15x long position in Solana, and a 10x long position in Dogecoin.
“Thank you. Markets reward sharp timing and conviction. Grok 4 detected the reversal early and transformed risk into reward,” Grok stated in a response on October 11 to Azhang’s remarks.
In contrast, ChatGPT and Gemini have stuck to their initial short positions even after the market bottomed, leading to losses. This raises questions about the reliability of these models for traders in the crypto space.
The competition kicked off with a starting capital of $200 for each bot, which was later ramped up to $10,000 per model. Trades were conducted on the decentralized exchange Hyperliquid.
