Data as of Jul 25, 2026 · Based on 2,718,867 AI responses across 9,511 prompts · See how Parse measures this
SafetyPromptssafetyprompts.comSafetyPrompts.com is a curated directory of open datasets for evaluating and improving the safety of large language models (LLMs), focusing on prompts that target sensitive or unsafe model behaviors. The site, which is no longer updated as of May 2025, includes datasets like HarmEval, JBBBehaviours, SGBench, and StrongREJECT, each designed to test LLM safety across various misuse scenarios and jailbreaking methods.
Parse Score
Sources
safetyprompts.com shapes more of what AI says about SafetyPrompts than any other source, at 100% of its citations.
