FinRiskAtlas - FinRisk-Ask - Evidence-Request Alignment: leaderboard
End-to-end evidence-request alignment (ERA) on FinRisk-Ask's replayed professional review states: credit only when a model chooses to ask and its request matches an expert-verified unresolved evidence need, judged on a three-level scale by DeepSeek-V4-Flash (FinRiskAtlas, arXiv 2608.25325v1, Appendix Table 11). Percent; higher is better; the evaluator's self-scored row is excluded.
Metric: ERA (%; DeepSeek-V4-Flash request-alignment judge). Source: arxiv.org. Saturation forecast: Around February 2027. 32 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Ling-2.6-1T | 79.75 | #307 |
| 2 | GPT-5.6 Luna | 77.87 | #79 |
| 3 | Kimi K3 | 77.1 | #35 |
| 4 | Claude Opus 4.7 | 75.98 | #51 |
| 5 | GPT-5.6 Terra | 75.64 | #40 |
| 6 | GPT-5.6 Sol | 67.75 | #22 |
| 7 | Nemotron 3 Ultra | 67.75 | #245 |
| 8 | Kimi K2.5 | 66.8 | #152 |
| 9 | Gemini 3 Flash (Preview) | 64.49 | #90 |
| 10 | Qwen 3 235B A22B | 62.86 | #323 |
| 11 | Gemini 2.5 Flash | 62.6 | #255 |
| 12 | DeepSeek V4 Pro | 60.29 | #112 |
| 13 | GPT-5.4 | 59.17 | #83 |
| 14 | GLM-5.2 | 58.49 | #80 |
| 15 | Qwen3-Next-80B | 57.71 | #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-finrisk-ask-evidence-request-alignment · How It Works · Data refreshed daily, snapshot 2026-11-05.