FinRiskAtlas - Applied Review - Risk Classification: leaderboard
FinRiskAtlas fixed-evidence review operation: assigning a case-level risk grade from evidence and the applicable risk-category set, scored by accuracy (FinRiskAtlas, arXiv 2608.25325v1, Appendix Table 9). Percent; higher is better. Reported for 33 archived model configurations under zero-shot direct-answer inference.
Metric: Accuracy (%). Source: arxiv.org. Saturation forecast: Not forecast. 33 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash (Preview) | 76.78 | #90 |
| 2 | Qwen 3.7 Max | 75.68 | #85 |
| 3 | Ling-2.6-1T | 75.68 | #307 |
| 4 | DeepSeek V4 Flash | 75.41 | #155 |
| 5 | Nemotron 3 Ultra | 75.41 | #245 |
| 6 | Claude Opus 4.7 | 75.14 | #51 |
| 7 | Ling-2.6-flash | 75 | #533 |
| 8 | Kimi K3 | 74.86 | #35 |
| 9 | DeepSeek V3 (0324) | 74.86 | #349 |
| 10 | Qwen 3 8B | 74.59 | #678 |
| 11 | GPT-5.6 Sol | 74.59 | #22 |
| 12 | GPT-5.4 | 74.04 | #83 |
| 13 | Qwen 3 32B | 73.77 | #449 |
| 14 | GPT-5.6 Terra | 73.77 | #40 |
| 15 | Qwen3-Next-80B | 73.77 | #321 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=finriskatlas-applied-review-risk-classification · How It Works · Data refreshed daily, snapshot 2026-11-05.