FinRiskAtlas - Domain Knowledge - Overall: leaderboard
Unweighted macro-average over 42 financial-risk domain-knowledge task families (4,679 instances) covering concepts, regulations, obligations and risk mechanisms behind financial review, from FinRiskAtlas (arXiv 2608.25325v1, Appendix Table 8). Percent; higher is better; 33 archived model configurations answering zero-shot.
Metric: Macro-Averaged Accuracy (%). Source: arxiv.org. Saturation forecast: Around February 2027. 33 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash (Preview) | 80.69 | #90 |
| 2 | Kimi K3 | 80.68 | #35 |
| 3 | Qwen 3.7 Max | 80.41 | #85 |
| 4 | Claude Opus 4.7 | 80.19 | #51 |
| 5 | GPT-5.6 Sol | 79.21 | #22 |
| 6 | GPT-5.4 | 78.66 | #83 |
| 7 | Kimi K2.5 | 78.56 | #152 |
| 8 | GPT-5.6 Terra | 78.34 | #40 |
| 9 | Ling-2.6-1T | 77.65 | #307 |
| 10 | GPT-5.4 Mini | 77.42 | #228 |
| 11 | Qwen3-Next-80B | 77.19 | #321 |
| 12 | GLM-5.2 | 77.13 | #80 |
| 13 | Seed 2.0 Lite | 77.01 | #161 |
| 14 | Qwen 3 235B A22B | 76.9 | #323 |
| 15 | GLM-5 | 76.26 | #149 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=finriskatlas-domain-knowledge-overall · How It Works · Data refreshed daily, snapshot 2026-11-05.