medasr is Google's specialized automatic speech recognition model for the medical domain, with a focus on radiology reports and clinical dictation. It uses a CTC-based architecture fine-tuned on medical audio to achieve higher accuracy on specialized terminology than general-purpose ASR systems. The model is available on Hugging Face for research and integration into healthcare applications.
In the Voice, TTS & speech space, Medasr takes a focused approach. Accurately transcribing medical and radiological speech with domain-specific terminology. It is built as an open-source project for healthcare developers. Medasr is open source under the Apache-2.0 license. It runs on the web, the command line, and API.
Google builds and maintains Medasr, and the product first shipped in 2025. Development happens publicly on GitHub with 142 stars. Key capabilities include Medical ASR, radiology transcription, and CTC decoding. It exposes integrations via a public API.
Latest indexed changes and source events
google/medasr verified by the PulseGate indexer
Other apps tracked under the same category.
