Google opens access to AI models for medical imaging and speech, unveils MedGemma 1.5 and MedASR: All you need to know

Google has launched two new artificial intelligence models focused on healthcare, MedGemma 1.5 and MedASR, as part of its expanding efforts in medical AI. Unlike some rivals that offer healthcare AI tools primarily as paid enterprise services, Google has opted for a more open approach by releasing both models publicly for the wider research and developer community.

According to Google Research, the updated version brings improved multimodal reasoning and better performance when dealing with complex medical imagery. It is also designed to be more flexible, allowing researchers to fine-tune it for specialised datasets and specific study requirements.

The model supports multiple forms of medical imaging, including radiology scans and other clinically relevant visuals. Google said MedGemma 1.5 is intended for uses such as image-based question answering, report drafting, and structured data extraction. The company stressed that it is not meant to provide diagnoses or treatment advice and should only be used as a support tool in research and development settings.

Google said the model aims to reduce transcription errors that often occur when general-purpose speech recognition systems are used in medical contexts. Potential use cases include transcribing doctor-patient discussions, creating clinical notes, and converting dictated reports into text.

The company added that MedASR can be adapted for different healthcare settings and fine-tuned to match specific clinical workflows or documentation standards.