AI News and Headlines

Google Health AI Releases MedASR: A Conformer Based Medical Speech to Text Model for Clinical Dictation

Google Health AI released MedASR, an open‑weights medical speech‑to‑text model built on the Conformer architecture. The 105‑million‑parameter model is trained on roughly 5,000 hours of de‑identified physician dictations and clinical conversations across radiology, internal and family medicine. Benchmarks show MedASR achieving word‑error rates as low as 4.6% on radiology dictation (with a six‑gram language model), outperforming Gemini 2.5 Pro and Whisper v3 Large. The model is hosted on Hugging Face and can be integrated via the Transformers pipeline or fine‑tuned for specific workflows.