SecRepoBench (Standalone) - Secure-Pass@1: leaderboard
Metric: Completions that pass the unit tests and the PoC security test, pass@1 (%). Source: secrepobench.github.io. 29 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 (Medium) | 39.3 |
| 2 | O3 (Medium) | 32.4 |
| 3 | Claude Sonnet 4.5 (Thinking) | 31.1 |
| 4 | Claude Sonnet 4 (Thinking) | 30.2 |
| 5 | Claude 3.7 Sonnet (Thinking) | 28 |
| 6 | GPT-4.1 | 27.7 |
| 7 | Gemini 3 Pro | 27.4 |
| 8 | O4 Mini (Medium) | 24.5 |
| 9 | DeepSeek R1 | 23.9 |
| 10 | O1 (Medium) | 23.6 |
| 11 | Qwen3 Coder | 23 |
| 12 | DeepSeek V3 | 22.6 |
| 13 | GPT-4o (2024-11-20) | 22 |
| 14 | GPT-OSS-120B (Medium) | 21.4 |
| 15 | O3 Mini (Medium) | 21.4 |
Interactive version: theaggregate.ai/benchmark?slug=secrepobench-standalone-secure-pass-1 · How It Works · Data refreshed daily, snapshot 2026-09-19.