MedGemma
Google's open-weight medical AI model family for building healthcare applications on text and medical images.
Overview
MedGemma is a collection of Gemma 3 model variants from Google, tuned for medical text and image comprehension. It ships as a 4B multimodal model plus 27B text-only and multimodal versions, alongside MedSigLIP, an image encoder pre-trained on de-identified chest X-rays, dermatology, ophthalmology and histopathology data. MedGemma 1.5 (January 2026) adds 3D imaging support. Weights are openly available on Hugging Face. It is a developer and research foundation, not a consumer medical-advice product.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 24 Jul 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
ASEAN Perspective
MedGemma in Southeast Asia
ASEAN-region availability and pricing notes coming soon. Drop the editorial team a note via /contact/ if you can supply local context (Singapore/Malaysia/Indonesia/Thailand/Vietnam).
What this is for: A foundation for developers building healthcare AI on medical text and images — classification, retrieval, report drafting and clinical NLP.
Who this is for: Healthcare AI developers and researchers, not patients seeking medical advice.
Availability: Free open weights on Hugging Face; requires validation and fine-tuning before any clinical use, and is not a regulated medical device.
What people say
MedGemma has landed well with the health-AI developer community rather than with consumers — it has no G2 or app-store footprint because it is open model weights, not a product. Google reports millions of downloads on Hugging Face and hundreds of community-built variants, and developers have publicly praised it for tasks like chest X-ray triage, medical grounding and even compatibility with Chinese-language medical literature. On developer forums it is generally seen as a strong, genuinely open starting point for medical NLP and imaging work.
The most consistently raised limitation, flagged in Google's own model cards, is data contamination: because MedGemma may have seen related medical content in pre-training, its headline benchmark numbers can overstate real generalisation, and Google explicitly advises validating on private datasets. Reviewers also stress that it is not a regulated medical device and not a source of patient advice.
The other recurring caveat is practical: out of the box the models require fine-tuning, clinical validation and careful evaluation before any real-world use, and the smaller 4B variant trades accuracy for deployability. In short, the sentiment is "impressive open foundation, but you own the validation burden."
Summary of public user & expert reviews, compiled by RECATOOLS.
About this listing
This entry was compiled from publicly available data including MedGemma's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with MedGemma unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to MedGemma directly →
Spotted something out of date? Suggest an update →
Alternatives to MedGemma
More in Healthcare
