Data as of Jul 25, 2026 · Based on 2,718,867 AI responses across 9,511 prompts · See how Parse measures this
PIGuard is a lightweight, open-source prompt guard model designed to detect prompt injection attacks against large language models. It introduces a training strategy called Mitigating Overdefense for Free (MOF) to reduce false positives on benign inputs containing trigger words.
Parse Score
