The Fin AI, with Stevens Institute of Technology, Columbia, Harvard, University of Manchester, Université de Montréal, University of Florida, Georgia Tech, PAAL AI, DeepKin, 2025

Agent Market Arena (AMA)

Live, ongoing trading by LLM agent frameworks on stocks and crypto. Four agent designs are crossed with several LLM backbones on verified market data and expert-checked news.

Best reported

40.8

InvestorAgent / GPT-4.1

cumulative return % (TSLA)

Reported by the authors, 30 Oct 2025

About this benchmark

Size
4 assets (TSLA, BMRN, BTC, ETH), 4 agents x 5 backbones plus a vote ensemble; paper period 2025-08-01 to 2025-09-30
Public split
Live dashboard only; no code or dataset release linked
Grading
Cumulative return, annualised return, volatility, Sharpe, max drawdown, per asset vs buy-and-hold
Licence
Not stated
Submission
maker-only
Products allowed
Not stated
Latest version
arXiv v2 (2025-10-30); live board ongoing
What it does not cover
Broad portfolios (only 4 single assets) and external agent submissions

Results

10 models
Leaderboard

What this does not measure: not investment performance. Returns come from a test setting over a fixed period, not from managing real money.

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.

Agent Market Arena (AMA) arXiv v2 (2025-10-30); live board ongoing

cumulative return % (TSLA), higher is better. 10 models.

Updated 30 Oct 2025

  1. InvestorAgent / GPT-4.140.8
  2. InvestorAgent / GPT-4o33.0
  3. InvestorAgent / Vote29.1
  4. InvestorAgent / Claude-sonnet-428.9
  5. TradeAgent / Gemini-2.0-flash21.9
  6. DeepFundAgent / GPT-4o13.9
  7. InvestorAgent / Claude-3.5-haiku11.6
  8. InvestorAgent / Gemini-2.0-flash10.6
  9. DeepFundAgent / Vote8.6
  10. TradeAgent / Vote-0.360

Agent Market Arena (AMA) arXiv v2 (2025-10-30); live board ongoing

cumulative return % (TSLA), higher is better. 10 models.

Updated 30 Oct 2025

  1. 1InvestorAgent / GPT-4.140.8
  2. 2InvestorAgent / GPT-4o33.0
  3. 3InvestorAgent / Vote29.1
  4. 4InvestorAgent / Claude-sonnet-428.9
  5. 5TradeAgent / Gemini-2.0-flash21.9
  6. 6DeepFundAgent / GPT-4o13.9
  7. 7InvestorAgent / Claude-3.5-haiku11.6
  8. 8InvestorAgent / Gemini-2.0-flash10.6
  9. 9DeepFundAgent / Vote8.6
  10. 10TradeAgent / Vote-0.360
Agent Market Arena (AMA) leaderboard (cumulative return % (TSLA), higher is better)
1InvestorAgent / GPT-4.140.8live 2025-08-01 to 2025-09-30, TSLA; buy-and-hold 46.88Paper Table 1
Reported by the authors, reported 30 Oct 2025 · checked 1 Nov 2026
2InvestorAgent / GPT-4o33.0live, TSLAPaper Table 1
Reported by the authors, reported 30 Oct 2025 · checked 1 Nov 2026
3InvestorAgent / Vote29.1live, TSLAPaper Table 1
Reported by the authors, reported 30 Oct 2025 · checked 1 Nov 2026
4InvestorAgent / Claude-sonnet-428.9live, TSLAPaper Table 1
Reported by the authors, reported 30 Oct 2025 · checked 1 Nov 2026
5TradeAgent / Gemini-2.0-flash21.9live, TSLAPaper Table 1
Reported by the authors, reported 30 Oct 2025 · checked 1 Nov 2026
6DeepFundAgent / GPT-4o13.9live, TSLAPaper Table 1
Reported by the authors, reported 30 Oct 2025 · checked 1 Nov 2026
7InvestorAgent / Claude-3.5-haiku11.6live, TSLAPaper Table 1
Reported by the authors, reported 30 Oct 2025 · checked 1 Nov 2026
8InvestorAgent / Gemini-2.0-flash10.6live, TSLAPaper Table 1
Reported by the authors, reported 30 Oct 2025 · checked 1 Nov 2026
9DeepFundAgent / Vote8.6live, TSLAPaper Table 1
Reported by the authors, reported 30 Oct 2025 · checked 1 Nov 2026
10TradeAgent / Vote-0.360live, TSLAPaper Table 1
Reported by the authors, reported 30 Oct 2025 · checked 1 Nov 2026

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