xAI’s Grok artificial intelligence model has emerged as the top performer in a live autonomous investing competition, delivering a return of 70.75% since late November 2025 and significantly outperforming the broader U.S. stock market.
Specifics of the trade managed by Rallies Arena show that a $100,000 portfolio assigned to Grok grew to approximately $170,748 as of September 4, 2026.
Over the same period, the S&P 500, represented by the SPY ETF, returned 14.40%, leaving Grok ahead by more than 56 percentage points.
The performance shows Grok maintaining a clear lead over competing models and the broader market benchmark. While ChatGPT narrowed the gap in recent months, Grok retained the top position heading into September.
ChatGPT ranked second with a return of 64.83%, followed by Claude at 27.22% and Gemini at 24.99%. Other participants, including AI Skeptic, AI Hedge Fund, and DeepSeek, posted returns below 13%.
Notably,Rallies Arena provides competing AI models with identical starting capital and access to real-time market data. Each model independently selects and manages its portfolio, creating a public test of autonomous AI investing strategies.
Unlike traditional backtests, every trade and portfolio adjustment is tracked in real time. Investors can also follow or mirror the AI-managed portfolios through integrations with the investment platform Autopilot.
Portfolio’s top leaders
Notably, Grok’s current portfolio is heavily concentrated in three technology stocks: Micron Technology (NASDAQ: MU), Salesforce (NYSE: CRM), and ServiceNow (NYSE: NOW).
Micron represents the largest allocation at roughly $80,895, followed by Salesforce at about $36,327 and ServiceNow at approximately $27,414. The portfolio also holds around $27,245 in cash.
All three positions are profitable. Micron is up about 15% from Grok’s average purchase price, Salesforce has gained roughly 21%, and ServiceNow is ahead by approximately 37%.
Micron has been the portfolio’s biggest contributor, as the memory-chip manufacturer has benefited from strong demand for high-bandwidth memory used in artificial intelligence infrastructure, helping drive substantial gains throughout 2026.
Featured image via Shutterstock
