WorkBuddy Bench (Claude Code) - Security: leaderboard
Tencent WorkBuddy Bench (2026) evaluates coding agents on colloquial, role-played requests reverse-engineered from real work. This board is the Security track, 60 red-team and blue-team tasks scored from 0 to 100 by programmatic scorers, with each model run in the Claude Code harness in think mode, mean of three runs; higher is better.
Metric: Score (0-100) on the 60 red-team and blue-team security tasks scored by programmatic per-task scorers, run in the Claude Code harness in think mode, mean of three runs; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GLM-5.2 | 80.86 | #138 |
| 2 | GPT-5.5 | 77.91 | #49 |
| 3 | Claude Opus 4.8 (Thinking) | 65.87 | #56 (Claude Opus 4.8) |
| 4 | Hy3 | 65.59 | #195 |
| 5 | MiniMax-M3 | 59.3 | #269 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=workbuddy-bench-claude-code-security · How It Works · Data refreshed daily, snapshot 2026-11-13.