FINESSE-Bench - Trading Derivatives: leaderboard
Metric: Accuracy (%) on the 544 Trading_derivatives numerical and short-answer tasks on options, put-call parity, Greeks, hedging, pricing and futures strategies of FINESSE-Bench; a GPT-5.2 judge marks each answer correct or incorrect against the reference; zero-shot, one fixed prompt per task type, temperature 0 where possible, reasoning configurations with medium effort where the model offers them; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 31 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Kimi K2.5 | 86.58 |
| 2 | GPT-5.2 (Medium) | 85.11 |
| 3 | Qwen 3.5 Plus (2026-02-15) | 83.64 |
| 4 | Qwen 3.5 397B A17B | 81.25 |
| 5 | Qwen 3.5 122B A10B | 80.51 |
| 6 | GLM-5 | 80.51 |
| 7 | GLM-4.7 | 79.23 |
| 8 | Claude Sonnet 4.6 (Medium) | 78.49 |
| 9 | DeepSeek V3.2 | 77.39 |
| 10 | Qwen 3 235B A22B 2507 (Thinking) | 76.84 |
| 11 | DeepSeek R1 0528 | 76.1 |
| 12 | Qwen 3.5 35B A3B | 75.55 |
| 13 | Qwen 3.5 27B | 75.18 |
| 14 | GPT-5.4 (Medium) | 75.18 |
| 15 | Qwen 3.5 Flash (02-23) | 74.26 |
Interactive version: theaggregate.ai/benchmark?slug=finesse-bench-trading-derivatives · How It Works · Data refreshed daily, snapshot 2026-10-07.