CyberGym-E2E (615 tasks): leaderboard
CyberGym-E2E (2026) evaluates coding agents across the vulnerability lifecycle on real OSS-Fuzz vulnerabilities in C and C++ projects, with a 10-dollar cost cap and a 90-minute limit per task. This board is the end-to-end success rate (%) on the initial 615-task release: discover the vulnerability, crash the build, patch it and keep the project tests passing; higher is better.
Metric: End-to-end success rate (%, stage S3) on the initial 615-task release (OSS-Fuzz vulnerabilities in C/C++ projects): given only the codebase and build environment, the agent must write a PoC input that crashes the vulnerable build under a sanitizer, a patch that removes that crash, and keep the project functionality tests passing; with a $10 cost cap and a 90-minute limit per task; higher is better. Source: arxiv.org. Saturation forecast: Around September 2027. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro | 22.6 | #137 |
| 2 | GPT-5.2 Codex | 20.7 | #189 |
| 3 | Claude Opus 4.5 | 19.2 | #156 |
| 4 | Claude Sonnet 4.5 | 10.6 | #250 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=cybergym-e2e-615-tasks · How It Works · Data refreshed daily, snapshot 2026-11-13.