Speech-to-text that speaks pharma
Build accurate, reliable, and secure pharma voice applications with speech-to-text tuned for drug names and pharmaceutical terminology.




Leading voice teams build with Deepgram
Made for the way pharma talks
Get pharmaceutical terminology right, without sacrificing the performance, security, or flexibility you need.
Built for precision
Accurate on pharmaceutical terminology
General-purpose speech-to-text models often mistranscribe drug names and pharmaceutical terminology, risking errors in patient care and critical workflows. Deepgram Nova-3 Pharma accurately transcribes drug names, pharmaceutical terminology, and medical vocabulary, without sacrificing the transcription quality you expect from Nova-3.

Specialized training
Understands pharma from day one
Nova-3 Pharma is trained on Deepgram's medical speech corpus, with additional pharma-specific training to recognize the specialized terminology pharmaceutical workflows depend on.

Real-world applications
From IVR to pharmacovigilance, covered
Build pharmacy IVRs, prescription refill automation, payer voice agents, clinical documentation, and pharmacovigilance applications using one speech-to-text API, with real-time and batch transcription across every pharmaceutical workflow.

Easy integration
Seamless deployment
Deploy with a single model change (model=nova-3-pharma) using the same architecture and compute footprint as Nova-3 Medical, with support for streaming, batch, and eight English language variants.

Built for enterprise
Meet your compliance bar
Deploy in regulated healthcare environments through hosted API, self-hosted, or AWS SageMaker options, with HIPAA-eligible infrastructure and Zero Data Retention (ZDR) available for enterprise deployments.


Measured on the terms that matter most
Nova-3 Pharma measures drug-name recognition term by term with Keyword Recognition Rate (KRR), alongside overall Word Error Rate (WER), for a more complete view of pharmaceutical transcription accuracy.
- #1Drug name recognition
- 99thPercentile in transcription accuracy
What our pharma and healthcare customers say on G2
FAQ
Deepgram Nova-3 Pharma is a speech-to-text model optimized to accurately transcribe drug names, pharmaceutical terminology, and medical vocabulary while maintaining the high-quality general transcription performance of Nova-3.
Use Deepgram Nova-3 Pharma for pharmaceutical workflows including pharmacy IVRs, prescription refill automation, payer voice agents, clinical documentation, pharmacovigilance, and other applications where accurate recognition of drug names and pharmaceutical terminology is critical.
Deepgram Nova-3 Pharma is specifically optimized for drug names, pharmaceutical terminology, and medical vocabulary, while Nova-3 Medical is designed for broader clinical speech. It adds pharmaceutical specialization while maintaining the high-quality general transcription performance of Nova-3 and Nova-3 Medical.
Yes. Deepgram Nova-3 Pharma supports both real-time streaming and batch transcription for live voice applications as well as pre-recorded audio.
Access Deepgram Nova-3 Pharma through the API using (model=nova-3-pharma). It uses the same architecture and compute footprint as Nova-3 Medical, so no deployment or hardware changes are required.
Deepgram Nova-3 Pharma supports eight English variants: en, en-US, en-GB, en-AU, en-CA, en-IE, en-IN, and en-NZ.
Deepgram Nova-3 Pharma supports hosted API, self-hosted, and AWS SageMaker deployment options, with HIPAA-eligible infrastructure and Zero Data Retention (ZDR) available for enterprise deployments.




