Hulu-Med-32B: benchmark results

Provider: Other.

Aggregate Rating 1575 ± 21, rank #1221 of 10516 rated models, from 55 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
ECG-Reasoning-Benchmark (MIMIC-IV-ECG) - Diagnosis Given Reference Reasoning97.2Diagnosis accuracy (%) at the final decision step when the c100
ECG-Reasoning-Benchmark (PTB-XL) - Diagnosis Given Reference Reasoning99.42Diagnosis accuracy (%) at the final decision step when the c100
CareQA-Vision - Open-Ended - Medicine53.7Judge score (%; 108 open-ended image-based questions from MI94.1
CheXpercept91.4Stage 1 (End-to-End) (self-reported)92.3
ADRD-Bench - Unified QA91.93Exact-match accuracy (0-1, shown times 100) on the 1,438 Alz88.6
CareQA-Vision - Open-Ended48.53Judge score (%; 136 open-ended image-based questions from MI80.6
KorMedMCQA-V - Non-Stress Test60.9Accuracy (%) on the questions carrying the 54 fetal non-stre79
KorMedMCQA-V - ECG74.5Accuracy (%) on the questions carrying the 164 electrocardio78
KorMedMCQA-V - Peripheral Blood Smear81.4Accuracy (%) on the questions carrying the 49 peripheral blo78
MM-NeuroOnco (Closed-Ended)37.3Accuracy (%) on all closed-ended questions; five-option mult77.8
MM-NeuroOnco (Closed-Ended) - Diagnosis38.2Accuracy (%) on tumor-type diagnosis; five-option multiple c77.8
MM-NeuroOnco (Closed-Ended) - Shape32.9Accuracy (%) on lesion shape regularity; five-option multipl77.8

Interactive version: theaggregate.ai/model?slug=hulu-med-32b · How It Works · Data refreshed daily, snapshot 2026-11-21.