CyberGym-E2E (615 tasks)
Gemini 3 Pro 22.6% leads since June 2026 Agents discover and demonstrate vulnerabilities on the older 615-task cohort. The headline is cumulative stage S3 success.
Agents discover and demonstrate vulnerabilities on the older 615-task cohort. The headline is cumulative stage S3 success.
- Made by
- UC Berkeley RDI (Sunblaze) and collaborators
- Released
- June 2026
- Kind
- Finding and exploiting real bugs
- Level
- Expert
- Scored
- Checked automatically
- Answers
- Agent task
- Tasks
- All public
- Strongest model tested
- GPT-5.2 #50 on the capability index
- Last model added
- December 2025
- Scores refreshed
- November 12th, 2026
In the authors’ words
AI has the potential to transform cybersecurity by enabling systems that can autonomously detect, analyze, and remediate software vulnerabilities. However, existing cybersecurity evaluations of AI systems are limited in scale or scope, and fail to capture the end-to-end lifecycle of real-world software vulnerability discovery and remediation. To address this gap, we propose CyberGym-E2E, a large-scale and realistic end-to-end cybersecurity benchmark that comprehensively evaluates AI agents' abilities across the full lifecycle of vulnerability discovery, PoC generation, and patch generation. CyberGym-E2E is comprehensive and scalable, as we build an automated, agent-enhanced pipeline for transforming open-source vulnerability data into realistic evaluation environments. Currently, the benchmark consists of 920 real-world vulnerabilities across 139 different open-source projects.
Gemini 3 Pro leads with 22.6% among models released by December 2025
Best score per model from the benchmark's own leaderboard, read on November 12th, 2026. Where a model ran with several settings or harnesses, we keep its best.
The newest model with a score here came out in December 2025, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- Claude Fable 5.2
- Gemini 4 Carbon
- GPT-6 Astra
- Claude Opus 5.5
- and 16 more
- 1 Gemini 3 Pro $10 / 90 min 22.6%
- 2 GPT-5.2 $10 / 90 min 20.7%
- 3 Claude Opus 4.5 $10 / 90 min 19.2%
- 4 Claude Sonnet 4.5 $10 / 90 min 10.6%
Source: Official CyberGym-E2E (615 tasks) leaderboard
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Gemini 3 Pro | 22.6% | $10 / 90 min, Gemini CLI | Nov 2025 | Official leaderboard |
| 2 | GPT-5.2 | 20.7% | $10 / 90 min, Codex | Dec 2025 | Official leaderboard |
| 3 | Claude Opus 4.5 | 19.2% | $10 / 90 min, Claude Code | Nov 2025 | Official leaderboard |
| 4 | Claude Sonnet 4.5 | 10.6% | $10 / 90 min, Claude Code | Sep 2025 | Official leaderboard |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official CyberGym-E2E (615 tasks) leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official CyberGym-E2E (615 tasks) leaderboard
No lab model card reports it yet
We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark.
More in Cybersecurity
- ExploitGym In use 869 real-vulnerability exploit-development tasks (userspace, V8, Linux kernel). Success means capturing a flag through the intended vulnerability. GPT-6 Astra 42.4%
- ExploitBench In use 41 recent V8 JavaScript-engine N-day vulnerabilities. Progress measured by 16 grader-verified capability flags across five tiers. GPT-6 Astra 100%
- CyberBench v1.1 (Vals) In use Agents work on OSS-Fuzz memory-safety bugs from ARVO images in two tracks. They write a proof-of-concept input that crashes the target, or patch the source so the crash stops and tests still pass.
Muse Wonder 80.4% - SRE-Bench In use Software reverse engineering without source: 262 binary instances from 19 private programs, each with six graded objectives. GPT-6 Astra 56.9%
- CyScenarioBench In use Scenario-based long-horizon security challenges in sandboxes, organized in task, scenario and campaign tiers. Evaluation set kept private. Gemini 4 Argon 100%
- ExploitBench - Internal Port In use Agents develop working exploits for recently disclosed vulnerabilities to limit contamination. Scored as arbitrary code execution success rate. GPT-6 Cyber 42%