MedGemma-1.5-4B-it: benchmark results
Provider: Other.
Aggregate Rating 1459 ± 22, rank #3382 of 10916 rated models, from 40 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| KorMedMCQA-V - Non-Stress Test | 50 | Accuracy (%) on the questions carrying the 54 fetal non-stre | 58 |
| KorMedMCQA-V - Peripheral Blood Smear | 55 | Accuracy (%) on the questions carrying the 49 peripheral blo | 56 |
| ECG-Reasoning-Benchmark (MIMIC-IV-ECG) - Diagnosis Given Reference Reasoning | 84.52 | Diagnosis accuracy (%) at the final decision step when the c | 55.6 |
| ECG-Reasoning-Benchmark (PTB-XL) - Diagnosis Given Reference Reasoning | 85.02 | Diagnosis accuracy (%) at the final decision step when the c | 50 |
| MMBU - Grounded Classification (Masks) | 42.8 | Closed-ended classification of a region marked by a segmenta | 50 |
| ShatterMed-QA (English, Easy) | 81.85 | Accuracy (%) on the 5,923 English Easy multiple-choice clini | 50 |
| KorMedMCQA-V - Single Image | 48.1 | Accuracy (%) on the 1,069 questions with one image (KorMedMC | 40 |
| ShatterMed-QA (English, Hard) | 76.26 | Accuracy (%) on the 1,692 English Hard multiple-choice clini | 40 |
| MMBU - Ungrounded Classification | 40.3 | Closed-ended (multiple-choice) ungrounded classification of | 37.5 |
| KorMedMCQA-V - ECG | 49.3 | Accuracy (%) on the questions carrying the 164 electrocardio | 36 |
| KorMedMCQA-V - MRI | 48.1 | Accuracy (%) on the questions carrying the 40 MRI images (Ko | 36 |
| KorMedMCQA-V - X-Ray | 44.3 | Accuracy (%) on the questions carrying the 586 X-ray images | 36 |
Interactive version: theaggregate.ai/model?slug=medgemma-1-5-4b-it · How It Works · Data refreshed daily, snapshot 2026-11-08.