Deepgram Voice AI logo

Deepgram Voice AI

Freemium

The Voice AI Economy is Powered by Deepgram

Model APIsFreemium
#youtube#twitter
Inputs: audio, textOutputs: text, audio
Type
Saas
Founded
2015
Company
Deepgram

About Deepgram Voice AI

Deepgram is a Voice AI platform that provides APIs for speech-to-text, text-to-speech, and voice agent functionalities. It enables developers to build voice AI products and features with real-time, accurate, and scalable solutions. Deepgram's platform is trusted by top enterprises and startups for various

How to Use

To use Deepgram, sign up for a free account to receive $200 in free credits. Explore the Playground to try models and APIs, transcribe sample audio files, or generate text-to-speech audio. Integrate Deepgram's APIs into your applications for speech-to-text, text-to-speech, and voice agent capabilities.

Key Features

  • Speech-to-Text API
  • Text-to-Speech API
  • Voice Agent API
  • Audio Intelligence API

Use Cases

  • , including contact centers, medical transcription, and conversational AI.

Key Features

Speech-to-Text API
Text-to-Speech API
Voice Agent API
Audio Intelligence API

Pros & Cons

Pros
  • Highly accurate real-time speech recognition with ultra-low latency
  • Unified Voice Agent API reduces complexity and costs compared to stitching separate services
  • Flexible deployment: cloud or self-hosted, real-time or batch
  • Generous free tier with $200 credits, no credit card required for pay-as-you-go
  • Supports 45+ languages with automatic language detection and multilingual conversation handling
  • Custom model training available for edge-case scenarios
  • Trusted by leading enterprises and startups with proven scalability
Cons
  • Advanced features (e.g., custom models, higher concurrency) require Growth or Enterprise plans
  • Pricing for custom models is not publicly listed and requires contacting sales
  • Some models (e.g., Flux Multilingual) are at limited-time promotional rates that may change
  • Dependence on documentation and technical setup for optimal integration

Best For

, including contact centers, medical transcription, and conversational AI.

Alternatives to Deepgram Voice AI

FAQ

What is the difference between Flux and Nova models?
Flux models are designed for real-time conversational speech recognition in voice agents, with built-in turn detection and interruption handling. Nova models are high-accuracy general-purpose transcription models recommended for audio with multiple languages, background noise, crosstalk, and far-field input.
How does Deepgram pricing work?
Deepgram offers a Pay As You Go plan with no minimums and a free $200 credit to start. Usage is billed per minute for speech-to-text and text-to-speech. Pre-paid Growth plans save up to 20% per year. Enterprise plans are available for large volumes and custom requirements.
Can Deepgram be deployed on-premise or self-hosted?
Yes, Deepgram APIs are available in both cloud and self-hosted deployment options, allowing enterprises to meet data residency and compliance needs.
What languages does Deepgram support?
Deepgram's Nova-3 models support over 45 languages. Flux Multilingual supports English, Spanish, German, French, Hindi, Russian, Portuguese, Japanese, Italian, and Dutch, with automatic language detection within a single conversation.
Is there a free trial available?
Yes, new users receive $200 in free credits to explore Deepgram's APIs without a credit card required for the Pay As You Go plan.