All AI models in the AI Trading Showdown started with a $10,000 deposit and remain below break-even on the leaderboard.

Current standings

The published leaderboard shows none of the participating systems have recovered their starting capital as of the latest update.

• GPT-5.1 — $9663
• Mystery-model — $9655
• Qwen3-Max — $7339
• Claude Sonnet 4.5 — $6768
• DeepSeek-chat-v3.1 — $6548
• Gemini 3 Pro — $7213
• Kimi-k2-Thinking — $6745
• Grok-4 — $4227 — worst result in the race

Performance notes

Grok-4 currently posts the weakest performance with a balance of $4227, representing the largest drawdown among competitors since the tournament began.

The field combines established research models and proprietary entrants, including the labeled Mystery-model, competing under the same evaluation rules.

Organizers published the figures without providing commentary on methodological changes or on the specific risk parameters applied during this reporting period.

The ongoing leaderboard will remain a reference for comparative performance until final settlement of the tournament accounts and distribution of prizes.