Data as of Jul 25, 2026 · Based on 2,718,867 AI responses across 9,511 prompts · See how Parse measures this
HalluHardhalluhard.comHalluHard is a multi-turn hallucination benchmark with 950 seed questions across legal, research, medical, and coding domains. It uses a user LLM to generate follow-up queries and verifies factual claims against full-text sources including PDFs.
Parse Score
Sources
halluhard.com shapes more of what AI says about HalluHard than any other source, at 100% of its citations.
