AI-Trader

Size

3 markets (Nasdaq-100 hourly, SSE-50 daily, 10 crypto pairs daily), 6 LLMs in the paper

Public split

Code and evaluation data on GitHub; live platform at ai4trade.ai

Grading

Cumulative return, Sortino ratio, max drawdown vs market index baseline

Licence

MIT (code, per README badge)

Submission

open

Products allowed

Yes

Latest version

arXiv v1 (2025-12-01); repo updated 2026-06-11

What it does not cover

Long horizons (paper windows are 2-5 weeks); the repo now presents itself as a trading platform more than a fixed benchmark

Each score shows who reported it: the benchmark's authors (quoted from their paper), the benchmark's official board, or Moneda (run with the benchmark's own grader and rules). We do not copy other aggregators.

1. 9.6MiniMax-M29.6
2. 8.4DeepSeek V3.18.4
3. 3.1Claude-3.7-Sonnet3.1
4. 1.6 gpt-51.6
5. 0.390Qwen3-Max0.390
6. -0.060Gemini 2.5 Flash-0.060

1. 1MiniMax-M29.69.6
2. 2DeepSeek V3.18.48.4
3. 3Claude-3.7-Sonnet3.13.1
4. 4 gpt-51.61.6
5. 5Qwen3-Max0.3900.390
6. 6Gemini 2.5 Flash-0.060-0.060

AI-Trader leaderboard (cumulative return %, higher is better)

Rank	Model	Organisation	cumulative return %	Setting	Source
1	MiniMax-M2	MiniMax	9.6	US market (Nasdaq-100, hourly), 2025-10-01 to 2025-11-07; QQQ baseline 1.87	Paper Table 1 Reported by the authors, 1 Dec 2025
2	DeepSeek V3.1	DeepSeek	8.4	US market, hourly	Paper Table 1 Reported by the authors, 1 Dec 2025
3	Claude-3.7-Sonnet	Anthropic	3.1	US market, hourly	Paper Table 1 Reported by the authors, 1 Dec 2025
4	gpt-5	OpenAI	1.6	US market, hourly	Paper Table 1 Reported by the authors, 1 Dec 2025
5	Qwen3-Max	Qwen	0.390	US market, hourly	Paper Table 1 Reported by the authors, 1 Dec 2025
6	Gemini 2.5 Flash	Google	-0.060	US market, hourly	Paper Table 1 Reported by the authors, 1 Dec 2025

• Paper
• Website
• Code

Last checked 1 Nov 2026. Something out of date or wrong? Send a correction.