deepgram

JSON twin: https://www.healthaidb.com/software/deepgram.json

Company Name

Deepgram

Product URL

https://deepgram.com/product/speech-to-text

Company URL

https://deepgram.com

Categories

Summary

Deepgram provides an enterprise-grade, HIPAA-capable speech-to-text API (Nova-3 Medical) optimized for accurate, real-time medical transcription and voice AI in clinical workflows.

Description

Deepgram is a voice AI platform offering medical-focused speech-to-text models (Nova-3 Medical) and related APIs for real-time and batch clinical transcription, diarization, formatting, analytics, and customizable domain adaptation with on‑prem and cloud deployment options to meet HIPAA/security requirements.

Api Available

yes

Certifications

Company Founding

2015

Company Offices

Compliance

Customers

Data Residency

Multi-region cloud hosting (US and EU regions) and option for self-hosted / on-prem deployment (customer-controlled region)

Data Standards

Deployment Model

Features

Id

P0457

Integration Partners

Integrations

Languages Supported

Last Updated

2025-09-07

License

commercial proprietary

Links

Market Segment

Optional Modules

Os Platforms

Pricing Details

Free tier and developer pricing available for basic API use; enterprise/pricing for medical models and on‑prem deployments: contact vendor for custom quote

Pricing Model

enterprise_quote

Privacy Features

Ratings

Regions Available

Release Year

2015

Security Features

Specialties

Support Channels

System Requirements

API-first (SaaS). For self-hosted: Linux servers, Docker/Podman, Kubernetes (EKS/GKE/AKS), access to cloud object storage (S3-compatible).

Target Users

Training Options

Type

product

User Reviews

Version

1.0

Canonical JSON

{
  "company_name": "Deepgram",
  "company_url": "https://deepgram.com",
  "company_offices": [
    "United States"
  ],
  "company_founding": "2015",
  "product_url": "https://deepgram.com/product/speech-to-text",
  "categories": [
    "clinical",
    "developer",
    "analytics",
    "speech-recognition",
    "clinical documentation"
  ],
  "market_segment": [
    "enterprise",
    "smb",
    "developer"
  ],
  "links": [
    "https://deepgram.com",
    "https://deepgram.com/product/speech-to-text",
    "https://deepgram.com/solutions/medical-transcription",
    "https://developers.deepgram.com/documentation/",
    "https://deepgram.com/pricing",
    "https://developers.deepgram.com/support/",
    "https://playground.deepgram.com/",
    "https://www.g2.com/products/deepgram/reviews"
  ],
  "summary": "Deepgram provides an enterprise-grade, HIPAA-capable speech-to-text API (Nova-3 Medical) optimized for accurate, real-time medical transcription and voice AI in clinical workflows.",
  "description": "Deepgram is a voice AI platform offering medical-focused speech-to-text models (Nova-3 Medical) and related APIs for real-time and batch clinical transcription, diarization, formatting, analytics, and customizable domain adaptation with on‑prem and cloud deployment options to meet HIPAA/security requirements.",
  "target_users": [
    "clinicians",
    "medical transcriptionists/scribes",
    "healthcare administrators",
    "EHR vendors/integrators",
    "clinical researchers",
    "developers building healthcare apps",
    "health systems/hospitals"
  ],
  "specialties": [
    "primary care",
    "emergency medicine",
    "radiology",
    "cardiology",
    "oncology",
    "psychiatry",
    "surgery"
  ],
  "regions_available": [
    "United States",
    "Canada",
    "United Kingdom",
    "European Union",
    "Australia",
    "India",
    "Global"
  ],
  "languages_supported": [
    "English",
    "Spanish",
    "French",
    "German",
    "Portuguese",
    "Chinese (Mandarin)",
    "Japanese",
    "Italian"
  ],
  "pricing_model": "enterprise_quote",
  "pricing_details": "Free tier and developer pricing available for basic API use; enterprise/pricing for medical models and on‑prem deployments: contact vendor for custom quote",
  "license": "commercial proprietary",
  "features": [
    "Real-time streaming speech-to-text API",
    "Pre-recorded (batch) transcription",
    "Medical / Nova medical models optimized for clinical terminology",
    "Low-latency, high-throughput transcription",
    "Speaker diarization",
    "Punctuation and formatting",
    "Timestamps and word-level confidence",
    "Custom model training / tuning",
    "Entity detection and formatting",
    "Redaction (PHI/PII) and configurable redaction",
    "Summarization and audio intelligence (topic detection, intent, sentiment)",
    "Language detection and multi-language support",
    "Streaming/websocket and REST API support",
    "Batch job processing and S3 integration",
    "On-prem/self-hosted deployment options (K8s/VMs)"
  ],
  "integrations": [
    "Amazon Connect",
    "AWS (S3, Marketplace, ECS/EKS/Kubernetes deployments)",
    "Genesys",
    "Zapier",
    "n8n",
    "Pipedream",
    "Merge (unified API platform)",
    "GitHub (examples / SDKs)",
    "Third-party orchestration via S3/webhooks"
  ],
  "deployment_model": [
    "SaaS",
    "self_hosted/on_prem",
    "hybrid"
  ],
  "os_platforms": [
    "Web (API/console)",
    "Linux (server/Kubernetes/Docker)",
    "macOS",
    "Windows"
  ],
  "optional_modules": [
    "Medical transcription model (Nova)",
    "Custom model training / private models",
    "Voice Agent API (real-time agents)",
    "Text-to-Speech (TTS) add-on",
    "Enhanced audio intelligence (summarization/topic detection)"
  ],
  "data_standards": [],
  "api_available": "yes",
  "system_requirements": "API-first (SaaS). For self-hosted: Linux servers, Docker/Podman, Kubernetes (EKS/GKE/AKS), access to cloud object storage (S3-compatible).",
  "compliance": [
    "HIPAA (supports HIPAA workflows/BAA)",
    "SOC 2 Type II",
    "GDPR",
    "CCPA"
  ],
  "certifications": [
    "SOC 2 Type II"
  ],
  "security_features": [
    "TLS encryption in transit",
    "Encryption at rest",
    "Role-based access control (RBAC)",
    "Single sign-on (SSO/SAML/OAuth)",
    "Two-factor authentication (2FA)",
    "Audit logging"
  ],
  "privacy_features": [
    "BAA available",
    "Configurable PHI/PII redaction",
    "Anonymization/redaction APIs",
    "Data minimization controls"
  ],
  "data_residency": "Multi-region cloud hosting (US and EU regions) and option for self-hosted / on-prem deployment (customer-controlled region)",
  "customers": [
    "Sharpen",
    "Red Box",
    "Five9",
    "Amazon Web Services (AWS)",
    "Leading medical transcription platform"
  ],
  "user_reviews": [
    "Impressively fast and quite accurate.",
    "Best performance I've seen for voice agents so far — very impressed.",
    "Accuracy is terrible in my tests; not suitable for production.",
    "Nova-2 model is great for transcribing English audio — very fast.",
    "Good documentation and the best latency I've found so far.",
    "Keyword boost didn't improve accuracy for me.",
    "Outperforms OpenAI Whisper for English transcription in my experience.",
    "Useful for telephony when using a model tuned for phone audio."
  ],
  "ratings": [
    "G2 — Named High Performer (Winter 2022); 91% users think company is heading in right direction; 89% likely to recommend (as reported by Deepgram)",
    "AWS Marketplace — generally positive marketplace reviews for Deepgram Speech Recognition listings"
  ],
  "support_channels": [
    "email",
    "chat",
    "ticketing",
    "community",
    "24x7"
  ],
  "training_options": [
    "documentation",
    "webinars",
    "live_online",
    "certification"
  ],
  "release_year": "2015",
  "integration_partners": [
    "AWS / Amazon Web Services",
    "Five9",
    "Red Box",
    "Sharpen",
    "CYPHER (marketplace listing referenced)",
    "AWS Marketplace"
  ],
  "id": "P0457",
  "slug": "deepgram",
  "type": "product",
  "version": "1.0",
  "last_updated": "2025-09-07",
  "links_json": {
    "self": "https://www.healthaidb.com/software/deepgram.json"
  }
}