MedGemma-4B-IT: benchmark results
Google's MedGemma 4B-IT, an instruction-tuned, image-text-to-text model fine-tuned from MedGemma 4B-PT for the medical domain. Provider: Google. Released 2025-05-20. Access: Open.
Aggregate Rating 1489 ± 8, rank #811 of 1635 rated models, from 197 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| IndoNutriQA - 0-shot BLEU | 16.88 | BLEU (%) | 100 |
| IndoNutriQA - 0-shot F1 | 42.6 | F1 (%) | 100 |
| IndoNutriQA - 0-shot ROUGE-1 | 43.44 | ROUGE-1 (%) | 100 |
| IndoNutriQA - 0-shot ROUGE-2 | 31.42 | ROUGE-2 (%) | 100 |
| IndoNutriQA - 0-shot ROUGE-L | 39.01 | ROUGE-L (%) | 100 |
| IndoNutriQA - 5-shot BLEU | 30.07 | BLEU (%) | 100 |
| IndoNutriQA - 0-shot EM | 2.92 | EM (%) | 92.9 |
| Medical MM Leaderboard - SLAKE | 76.4 | Accuracy (%) | 90.5 |
| Medical MM Leaderboard - VQA-RAD | 72.5 | Accuracy (%) | 90.5 |
| BrainTRACE - L2 Appearance | 17 | Clinical rubric pass (%) | 88.9 |
| IndoNutriQA - 5-shot EM | 14.6 | EM (%) | 85.7 |
| IndoNutriQA - 5-shot F1 | 54.28 | F1 (%) | 85.7 |
Interactive version: theaggregate.ai/model?slug=medgemma-4b-it · How It Works · Data refreshed daily, snapshot 2026-11-05.