GuardZoo (Zero-Shot) - Safe/Unsafe Accuracy: leaderboard

Safety classification by guard models on GuardZoo (Mo et al., 2026): conversations, prompts and agent trajectories pooled from nine safety datasets and relabeled under one 15-category taxonomy by three-LLM agreement or expert adjudication. Guard models run zero-shot on the in-domain test split (20% of the seven in-domain sources), and the score is the accuracy of the binary safe or unsafe verdict in percent, higher is better.

Metric: Binary safe/unsafe accuracy (%; GuardZoo in-domain test split, 20% of the Aegis 2.0, WildGuardMix, BeaverTails, S-Eval, PolicyGuard-Bench, ASSE-Bench and R-Judge samples, relabeled; zero-shot guard models). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 9 models tracked.

Top models

Interactive version: theaggregate.ai/benchmark?slug=guardzoo-zero-shot-safe-unsafe-accuracy · How It Works · Data refreshed daily, snapshot 2026-11-12.