FinEval - Financial Security Knowledge: leaderboard
FinEval's financial security section: 1,640 proprietary questions on application, cryptographic, memory, network, system, software and web security, malware analysis, penetration testing, reverse engineering and vulnerability detection in financial settings. Scored as accuracy (%) under zero-shot prompting; higher is better.
Metric: Accuracy (%, zero-shot). Source: github.com. Saturation forecast: Estimated already saturated. 19 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4o | 81.8 | #741 |
| 2 | Qwen 2.5 72B Instruct | 81.8 | #1093 |
| 3 | GPT-4o Mini | 79.1 | #1627 |
| 4 | Claude 3.5 Sonnet | 78.1 | #747 |
| 5 | Gemini 1.5 Pro | 77.8 | #1053 |
| 6 | Gemini 1.5 Flash | 77.5 | #1671 |
| 7 | Yi 1.5 34B Chat | 76 | #2513 |
| 8 | internlm2.5-20B-chat | 74.1 | #2563 |
| 9 | GLM-4 9B Chat | 73.1 | #2782 |
| 10 | internlm2-chat-20B | 73.1 | #2896 |
| 11 | Qwen 2.5 7B Instruct | 71.7 | #2583 |
| 12 | Yi-1.5-9B Chat | 71.4 | #4060 |
| 13 | Baichuan2-13B-Chat | 61.6 | #6615 |
| 14 | chatglm3-6B | 48.2 | #8269 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=fineval-financial-security-knowledge · How It Works · Data refreshed daily, snapshot 2026-11-12.