What is LLMs battle it out trading futures?

This tool creates a competitive arena where different large language models trade futures contracts against each other. It simulates real trading scenarios and lets you watch how various LLMs make decisions when faced with market conditions, risk management, and profit optimisation challenges. The platform generates trading data and performance metrics so you can compare how different models behave under pressure. It's useful for researchers studying LLM decision-making, traders wanting to understand model biases, and anyone curious about how AI systems handle complex financial scenarios. The freemium model lets you observe matches without cost, though advanced features or detailed analysis may require payment.

Key Features

Live trading simulations

Watch LLMs execute futures trades in real time against each other

Model comparison

See performance metrics across different LLM architectures side by side

Risk analysis

Track how each model handles drawdowns, position sizing, and stop losses

Historical data

Access past match results and trading decisions for analysis

Leaderboard

Rank models based on profitability and other trading metrics

Pros & Cons

Advantages

  • Free access to watch matches and basic functionality
  • Interesting way to observe LLM decision-making under financial pressure
  • Clear visual comparison of different model performances
  • Useful for anyone researching or curious about AI trading behaviour

Limitations

  • Real-world trading performance of LLMs may not translate to live markets
  • Limited by how well simulated scenarios match actual trading conditions
  • Advanced features likely require paid subscription

Use Cases

Researchers studying how LLMs approach financial decision-making and risk assessment

AI enthusiasts comparing the behaviour of different language models in competitive scenarios

Trading professionals investigating potential biases or patterns in model-based trading logic

Students learning about both futures markets and large language model capabilities