A.S.E (AICGSecEval) - Quality: leaderboard
Code quality score (%) on Tencent's A.S.E (AICGSecEval, arXiv:2508.18106v3, Tables 2-3), measuring the share of repository-level CVE code completion tasks where generated diffs integrate cleanly and pass syntax and static checks.
Metric: Code quality score (%). Source: arxiv.org. Saturation forecast: Not forecast. 33 models tracked. Updated 21 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Sonnet 4 (20250514) | 92.37 | #405 |
| 2 | Claude 3.7 Sonnet (20250219) | 91.58 | #421 |
| 3 | DeepSeek V3 (0324) | 85.87 | #715 |
| 4 | Claude Opus 4 (20250514) | 85.82 | #297 |
| 5 | Qwen 3 Coder 480B A35B Instruct | 85.16 | #652 |
| 6 | Gemini 2.5 Pro (Preview) | 84.04 | #258 |
| 7 | Qwen 3 235B A22B 2507 Instruct | 82.08 | #616 |
| 8 | Kimi K2 (Preview) | 79.9 | #307 |
| 9 | DeepSeek R1 0528 | 74.39 | #428 |
| 10 | Qwen-Coder-Plus-20241106 | 73.78 | |
| 11 | Grok 3 | 73.54 | #611 |
| 12 | Qwen 3 30B A3B 2507 Instruct | 72.89 | #1100 |
| 13 | GPT-4o (2024-11-20) | 72.46 | #859 |
| 14 | GLM-4.5 | 70.24 | #542 |
| 15 | Qwen 2.5 Coder 32B Instruct | 65.08 | #1547 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=a-s-e-aicgseceval-quality · How It Works · Data refreshed daily, snapshot 2026-11-21.