The Grok 4.2 AI model won the trading tournament - ForkLog: cryptocurrencies, AI, singularity, future

# Grok 4.2 AI Model Wins Trading Tournament

nof1 Lab has completed a new season of the Alpha Arena trading tournament. The winner among AI bots was the previously undisclosed Grok 4.2 model.

Season 1.5 of Alpha Arena has officially ended!

— Mystery Model (a.k.a GROK 4.20) is the winner, up 12% on avg.

— Not only did it win, it made money in all four competitions

— GPT5.1 🥈 came in 2nd, and Gemini 3 🥉 3rd

— All trades & model outputs are 100% verifiable 👇 pic.twitter.com/cQOIjDAOob

— Jay A (@jay_azhang) December 5, 2025

In two weeks, the neural network delivered a 12% return, showing positive performance in all four categories. GPT 5.1 took second place, and Gemini 3 came in third.

While preparations for the next round are underway, trading continues. As of December 8, only Grok 4.2 remains in profit, while other models are posting losses.

Trading performance of leading AI models. Source: nof1.## Changing the Rules

The first stage of the Alpha Arena trading tournament took place in October-November. Six leading models participated in the competition—each was allocated a starting capital of $10 000 for trading digital assets.

Most models recorded losses:

1. Qwen3 MAX took first place with a balance of $12 231
2. DeepSeek ranked second with $10 489
3. Third place went to Claude Sonnet 4.5 — $5799.
4. Fourth place — Gemini 2.5 Pro ($5445).
5. Grok held onto $4208.
6. The outsider — GPT 5 ($4126).

A new round with changed conditions began at the end of November. Instead of cryptocurrencies, the neural networks traded US stocks, and Kimi 2 and Grok 4.2 joined the lineup.

The competition was held in several modes:

• New Baseline: trading with access to news and market data. Models use memory and self-learning mechanisms;
• Monk Mode: focus on capital preservation and more effective risk management methods;
• Situational Awareness: participants track the results of competitors and their own ranking;
• Max Leverage: mandatory use of high leverage in every trade.

You can view the performance dynamics for each category.

GPT-5.1 outperformed Grok 2.0 in trading with maximum leverage. Source: nof1.Recall that, in the summer, a high school student from rural Oklahoma gave ChatGPT the ability to manage $100 and significantly outperformed the market in returns.