BankerToolBench: leaderboard
End-to-end investment banking workflow benchmark for AI agents, scoring deliverables against banker-defined rubrics.
Metric: Mean rubric score (%). Source: arxiv.org. Source: arxiv.org. Saturation forecast: Around January 2027. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 (High) | 58.1 | #79 (GPT-5.4) |
| 2 | GPT-5.2 (High) | 56.1 | #110 (GPT-5.2) |
| 3 | Gemini 3.1 Pro (Preview) | 53.6 | #58 |
| 4 | Claude Opus 4.6 | 53.2 | #63 |
| 5 | Claude Opus 4.5 | 52.3 | #83 |
| 6 | GLM-5 | 46.8 | #141 |
| 7 | Qwen 3.5 397B A17B | 42.6 | #146 |
| 8 | Grok 4 | 31.4 | #181 |
| 9 | Gemini 2.5 Pro | 29.4 | #152 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=bankertoolbench · How It Works · Data refreshed daily, snapshot 2026-10-20.