Benchmark RadarAI BENCHMARK PROFILE
HiMed
HiMed is a Hindi medical dataset and benchmark suite covering Western and Indian medicine. It evaluates medical reasoning in Hindi across multiple tasks, with data and evaluation scripts publicly released on GitHub.
- Released
- 2026-05-23
- Readiness
- Runnable
- Primary field
- Health & Life Sciences
Why it matters
Hindi remains underrepresented in medical LLMs; this benchmark enables systematic evaluation of cross-lingual medical reasoning, helping reduce performance gaps between English and Hindi in healthcare applications.
Motivation
Medical large language models hold promise for reducing healthcare disparities, yet Hindi remains severely underrepresented.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.