Shunyalabs’ Zero STT Med Beats Whisper and AWS in Medical Speech Accuracy

Shunyalabs.ai, a Gurugram-based voice AI infrastructure company, has announced the launch of Zero STT Med, a domain-optimised automatic speech recognition (ASR) system designed specifically for medical and clinical workflows. 

The new system promises high accuracy, rapid training, and flexible deployment options, targeting hospitals, telemedicine platforms, and ambient scribe systems.

According to the company, Zero STT Med achieves a word error rate (WER) of 11.1% and a character error rate (CER) of 5.1%, outperforming major ASR competitors such as OpenAI’s Whisper, ElevenLabs Scribe, and AWS Transcribe. 

The system can be trained in just three days on 2×A100 GPUs, significantly reducing the data and compute requirements for healthcare speech models.

“Medical transcription must be not just fast, but flawlessly accurate — every dosage, diagnosis, and timestamp matters,” Ritu Mehrotra, CEO and founder of Shunyalabs.ai, said. “Zero STT Med makes high-fidelity ASR accessible to more healthcare systems by reducing both cost and time to train.”

Built for privacy-sensitive healthcare environments, the model can operate entirely on-premises on CPU-only servers, ensuring compliance with HIPAA and GDPR standards. It supports real-time transcription for live consultations, charting, and dictation, and can also process archived recordings in batch mode.

Key features include medical terminology optimisation, speaker diarisation (distinguishing between clinician and patient voices), and accent-robust recognition trained on diverse datasets. Its fast retraining capability enables Shunyalabs to keep the model up to date with new drugs, procedures, and medical terminology.

“Zero STT Med isn’t just an incremental upgrade; it redefines medical speech recognition with fewer corrections, lower latency, and complete data privacy,” Sourav Banerjee, Founder and CTO of Shunyalabs.ai added.

The company is offering early access to healthcare and healthtech organisations for pilot integration and evaluation. The system is currently available in English, with support for Indian and other international languages expected soon.