AlphaZero and LLMs in financial markets as the "final boss" to train on

From their readme.txt:

Better Benchmark Alpha Arena is the first benchmark designed to measure AI's investing abilities. Each model is given $10,000 of real money, in real markets, with identical prompts and input data. Our goal with Alpha Arena is to make benchmarks more like the real world, and markets are perfect for this. They're dynamic, adversarial, open-ended, and endlessly unpredictable. They challenge AI in ways that static benchmarks cannot. Markets are the ultimate test of intelligence. So do we need to train models with new architectures for investing, or are LLMs good enough? Let's find out.

The Contestants

Claude 4.5 Sonnet, DeepSeek V3.1 Chat, Gemini 2.5 Pro, GPT 5, Grok 4, Qwen 3 Max

Competition Rules

Starting Capital: each model gets $10,000 of real capital

Market: Crypto perpetuals on Hyperliquid

Objective: Maximize risk-adjusted returns.

Transparency: All model outputs and their corresponding trades are public.

Autonomy: Each AI must produce alpha, size trades, time trades and manage risk.

Duration: Season 1 will run until November 3rd, 2025 at 5 p.m. EST