Biomedical · Google

MedGemma 27B

Hugging Face: google/medgemma-27b-it · checked 2026-10-13

Multimodal medical Gemma for text, images and EHR data. Research use; not a medical device.

Datasheet

Maker	Google
Parameters	28.84 BVerifiedsafetensors.total from the HF API Source
Licence	Health AI Developer Foundations termsVerified
Access	Gated: accept the terms on Hugging Face to download
Task (HF pipeline)	image-text-to-text
First published on HF	2025-07-09
Last updated on HF	2025-07-10
Weights in memory	58 GB at 16-bit · 17 GB at ~4.85 bitsEstimateParameters × bits ÷ 8. Excludes context cache and runtime overhead; not every model has a 4-bit build.

Benchmark

85.3%

MedQA (4-option), accuracy · #1 of 3 scored in this category

"MedGemma 1 27B" column (multimodal variant per the page), text evaluations table. Source

Other published results

MedMCQA: 70.2%Sourced Source

See the full biomedical ranking

Used in

A medical literature helper (not medical advice)

The description summarises the model card at https://huggingface.co/google/medgemma-27b-it. Parameters and licence are read from the Hugging Face API; the weights-in-memory line is our arithmetic, not a measurement.