Google released two new AI models built for medicine. These are Med-GEMMA & Med-SigLIP

Google Releases Two Free Medical AI Models That Could Change How We Use Technology in Healthcare

Google released two new AI models built for medicine. These are Med-GEMMA & Med-SigLIP. The models are designed to read medical records, answer clinical questions, and even interpret chest X-rays & MRI/& CT-Scans with precision. And they’re available to use without restrictions.

It means, free for practitioners, medical educators, students etc to use. One of the models, Med-GEMMA 27B, was tested using real U.S medical licensing exam questions & got an accuracy score of 87.7%. Even its smaller versions reached over 81%, which puts them in a range where they could be genuinely helpful to medical teams. The second model, Med-SigLIP, was trained using data that included input from certified radiologists. What id different about it is that it does not just process images. It can look at an X-ray/scans and understand it alongside patient medical/clinical history and clinical notes. That’s a hard thing to do well; and it’s something doctors spend years learning to do with care and accuracy.

Now, anyone can try these models. There is no payment attachedto using it, no special license. They’re open-source and can run locally. That means a hospital in Boston, a teaching lab in Nairobi, a theater in Enugu Nigeria or a clinic in Manila could all explore the same technology without asking for permission or paying a any fee. They are available right now on platforms like Hugging Face, GitHub, and Kaggle. You can download the models, study how they were trained, tweak them, or even build something new on top of them.

This does not mean to replace doctors. These tools are not perfect, and they are not meant to make final decisions. But they can offer support especially in places where access to specialists is limited or where time and resources are also tight and limited.

This kind of release doesn’t happen often. Making high-level medical AI free and editable is a step toward wider access and shared innovation. And it puts serious tools into the hands of people who are usually left waiting.