Project AURIONAI Use-Cases Research & Innovation
LLM

MedGemma 27B text

27B text-only medical LLM built on Gemma 3 27B. Scores 87.7 on MedQA (0-shot with test-time scaling) per the MedGemma Technical Report.

GoogleGoogleUSAMedGemmaFamily
Google
Levelmodel
ClassLLM
Parameters27B
Contextnot in the notes
Released19 May 2025
Opennessopen weights restricted
Runs onnot in the notes
Statusnot in the notes

Benchmark Results

vendor-reported
AfriMed-QAdomain
vendor-reported
84
050100

Score in %. Higher is better.

Jul 2025 arxiv.org, huggingface.co
AgentClinic-MedQAdomain
vendor-reported
56.2
050100

Score in %. Higher is better.

Jul 2025 arxiv.org, huggingface.co
MedMCQAdomain
vendor-reported
74.2
050100

Score in %. Higher is better.

Jul 2025 arxiv.org, huggingface.co
MedQAdomain
27B, best-of-5vendor-reported
89.8
050100

Score in %. Higher is better.

Jul 2025 arxiv.org, huggingface.co
27B, 0-shot with test-time scalingvendor-reported
87.7
050100

Score in %. Higher is better.

Jul 2025 arxiv.org, huggingface.co
MMLU Meddomain
vendor-reported
87
050100

Score in %. Higher is better.

Jul 2025 arxiv.org, huggingface.co
PubMedQAdomain
vendor-reported
76.8
050100

Score in %. Higher is better.

Jul 2025 arxiv.org, huggingface.co
Where this comes from
Every score names who ran it: vendor-reported, maintainer-run or independent. A rank is shown with the population it was ranked in (gap rule 3).

Lineage

finetune
Gemma 3 27B to MedGemma 27B multimodal: finetune, high confidenceGemma 3 27B to MedGemma 27B text: finetune, high confidenceGemma 3 27BGemma 3 27BGoogleMedGemma 27B multimodalMedGemma 27B multimodalGoogleMedGemma 27B textMedGemma 27B textGoogle, May 2025
GoogleGemma 3 27B is the parent (finetune)high confidencemodel card
Where this comes from
Parent and child relations from model cards, Hugging Face base_model fields and the genealogy tables of the research notes. The tree shows two generations each way, 3 models. Confidence and evidence are the notes' own.

Industry Use Cases

Healthcare
Clinical LLM (open weights)

R&D Topics

Domain adaptation (vertical models)produced byOpen-weight frontier releases as research substrateused by

Identity

Developer
Google
Family
MedGemma
Released
19 May 2025 sourcesregistry date
Weights published
–
Record checked
10 Oct 2026
Where this comes from
Developer and dates from the research notes. Release date is the vendor's public announcement; weights date is when open weights were published.

Architecture and Size

Architecture
not in the notes
Total parameters
27B sources
Active parameters
–
Checkpoint parameters
–
Note
Hugging Face model card created 20 May 2025.
Context length
not in the notes
Input
text
Output
text
Where this comes from
Parameter counts as the developer states them. Where the checkpoint holds a different total, both are shown (register A16).

Openness and Access

Openness
open weights restricted
Licence
Health AI Developer Foundations terms of use sources
Access routes
open weights download
Where this comes from
The licence file governs over registry tags and aggregator labels.

Deployment

Runs on
not in the notes
Footprint
On premises or in the cloud (open weights).
Runtimes
not in the notes
Where this comes from
Deployment tiers and runtimes as named in the notes. Footprint is the notes' figure, not a measurement by AURION.

Indonesia

Indonesian support
not stated
Regional languages
not in the notes
Jakarta hosting
unknown
Languages
not in the notes
Where this comes from
Indonesia fields are carried on every model; unknown is an allowed value.

Sources

  1. MedGemma Technical Report arxiv.org https://arxiv.org/html/2507.05201v2
  2. google/medgemma-27b-text-it , Hugging Face huggingface.co https://huggingface.co/google/medgemma-27b-text-it

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "MedGemma 27B Text", https://aurion.semai.id/models/google-medgemma-27b-text/, records of October 2026.