MedGemma 4B IT
Multimodal
Zero-eval
#1MIMIC CXR
#1DermMCQA
#1PathMCQA
+4 more
by Google
About
MedGemma 4B IT is a multimodal language model developed by Google. The model shows competitive results across 7 benchmarks. It excels particularly in MIMIC CXR (88.9%), DermMCQA (71.8%), PathMCQA (69.8%). As a multimodal model, it can process and understand text, images, and other input formats seamlessly. Released in 2025, it represents Google's latest advancement in AI technology.
Timeline
AnnouncedMay 20, 2025
ReleasedMay 20, 2025
Specifications
Capabilities
Multimodal
License & Family
License
Health AI Developer Foundations terms of use
Benchmark Performance Overview
Performance metrics and category breakdown
Overall Performance
7 benchmarks
Average Score
58.5%
Best Score
88.9%
High Performers (80%+)
1Top Categories
general
59.5%
vision
56.1%
Benchmark Performance
Top benchmark scores with normalized values (0-100%)
Ranking Across Benchmarks
Position relative to other models on each benchmark
MIMIC CXR
Rank #1 of 1
#1MedGemma 4B IT
88.9%
DermMCQA
Rank #1 of 1
#1MedGemma 4B IT
71.8%
PathMCQA
Rank #1 of 1
#1MedGemma 4B IT
69.8%
SlakeVQA
Rank #1 of 1
#1MedGemma 4B IT
62.3%
VQA-Rad
Rank #1 of 1
#1MedGemma 4B IT
49.9%
All Benchmark Results for MedGemma 4B IT
Complete list of benchmark scores with detailed information
MIMIC CXR MIMIC CXR benchmark | general | text | 0.89 | 88.9% | Self-reported |
DermMCQA DermMCQA benchmark | general | text | 0.72 | 71.8% | Self-reported |
PathMCQA PathMCQA benchmark | general | text | 0.70 | 69.8% | Self-reported |
SlakeVQA SlakeVQA benchmark | vision | multimodal | 0.62 | 62.3% | Self-reported |
VQA-Rad VQA-Rad benchmark | vision | multimodal | 0.50 | 49.9% | Self-reported |
CheXpert CXR CheXpert CXR benchmark | general | text | 0.48 | 48.1% | Self-reported |
MedXpertQA MedXpertQA benchmark | general | text | 0.19 | 18.8% | Self-reported |
Resources