MedGemma-27B-text-it: benchmark results

Provider: Other. Access: Open.

Unified ELO 1562 ± 25, rank #536 of 1537 rated models, from 31 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
MIMIC-CDM - Pancreatitis68.2Accuracy (%)100
Open Portuguese LLM - ASSIN2 RTE94.85Macro F1 (%)99.1
Open Portuguese LLM - FaQuAD NLI85.38Macro F1 (%)98.6
Open Portuguese LLM - BLUEX75.38Accuracy (%)95.9
HealthBench Hard57Overall score (self-reported)93.5
MIMIC-CDM69.1Accuracy (%)92.9
MIMIC-CDM - Appendicitis89.1Accuracy (%)92.9
Open Portuguese LLM - ENEM79.78Accuracy (%)92.8
Open Portuguese LLM - OAB Exams60.41Accuracy (%)92.7
BRIDGE Medical Leaderboard - Few-Shot51.73Average Performance (%)91.7
MIMIC-CDM - Diverticulitis61.1Accuracy (%)89.3
BRIDGE Medical Leaderboard43.34Average Performance (%)85.2

Interactive version: theaggregate.ai/model?slug=medgemma-27b-text-it · How It Works · Data refreshed daily, snapshot 2026-09-25.