by Google
MedGemma is Google's collection of open models for medical text and image comprehension. Developers use it as a starting point for healthcare AI applications.
APIFree
Step 1: Read the model card for training and usage details.
Step 2: Pick the size that fits your task, with the larger text model suited to most text use cases.
Step 3: Try prompt engineering and in-context learning first.
Step 4: Validate performance on your own use case.
Step 5: Fine-tune or add agentic orchestration if results fall short before deploying.
No. The page says it is not yet clinical-grade and needs validation and likely further adaptation before production.
Use prompt engineering with few-shot examples, fine-tune the model, or place it inside an agentic system with other tools.
For most medical text use cases the larger 27B model generally gives the best performance.
You may pursue any use case as long as it follows the Health AI Developer Foundations terms of use.
AWS HealthScribeAWS HealthScribe applies AI to generative AI healthcare API for transcribing patient-clinician conversations and producing preliminary clinical documentation.
Infermedica APIInfermedica API gives developers clinical AI triage they can add to an app, website or call center by chat, voice or form. It is certified as a Class IIb medical device under EU MDR.
Infermedica Symptom CheckerInfermedica Symptom Checker applies AI to AI clinical reasoning API for symptom assessment, triage, and care-navigation experiences in digital health products.