#ai-researchStory
HealthAgentBench: A Unified Benchmark Suite of Realistic Agentic Healthcare Environments for Challenging Frontier AI Agents
Microsoft introduced HealthAgentBench, a comprehensive benchmark suite for evaluating AI agents in healthcare. It consists of 54 tasks across 7 categories, revealing challenges in medical imaging and compositional reasoning. The strongest agent, Codex GPT-5.5, achieved a 42% success rate.
Read original ↗Sign in to upvoteSign in to save