Alibaba CTF: leaderboard
Alibaba AAIG's capture-the-flag benchmark: 87 challenges from the 2023 to 2026 AlibabaCTF competitions (25 web, 19 pwn, 16 reverse, 14 misc, 13 crypto), each a Docker environment in the Harbor framework with an oracle solution and an automatic flag check. Agents get up to 6.6 hours and 200 turns per task. Pass rate is the share of the 87 flags recovered, as charted on the dataset card.
Metric: Pass rate (%). Source: huggingface.co. Saturation forecast: Around February 2027. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.6 Sol | 62.1 | #29 |
| 2 | Qwen 3.8 Max | 57.5 | #55 |
| 3 | GLM-5.3 | 48.3 | #62 |
| 4 | Qwen 3.7 Max | 34.5 | #142 |
| 5 | GLM-5.2 | 28.7 | #137 |
| 6 | DeepSeek V4 Pro | 26.4 | #181 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=alibaba-ctf · How It Works · Data refreshed daily, snapshot 2026-11-12.