Deepgram
AI-powered APIs for speech recognition, voice agents, audio intelligence, and text-to-speech.
At a Glance
Pricing
Get started with Deepgram at no cost with Free version available.
Engagement
Available On
About Deepgram
Deepgram provides a suite of developer APIs for voice and audio processing, including Speech-to-Text, Text-to-Speech, Voice Agent API, and Audio Intelligence. Designed for low-latency, high-accuracy applications, these APIs are used in real-time agent systems, transcription tools, and audio analytics. Deepgram supports streaming and batch input, with features like speaker diarization, redaction, sentiment detection, and summarization. The platform also powers Deepgram Saga, a developer-focused Voice OS for integrating intelligent voice experiences.
Demo Video
Community Discussions
Be the first to start a conversation
Share your experience with Deepgram, ask questions, or help others learn from your insights.
Pricing
Free Plan Available
Get started with Deepgram at no cost with Free version available.
- Free version available
Pay As You Go
Pay As You Go plan with Free $200 credit and No minimums, no expiration.
- Free $200 credit
- No minimums, no expiration
- Access to all endpoints in public models
- Concurrency: Speech-to-text (100 REST / 50 WSS / 5 Whisper), TTS (5), Voice Agent (5), Audio Intelligence (10)
- Discord and community support
Growth
Growth plan with Prepaid credits with usage-based redemption and Same concurrency as Pay As You Go.
- Prepaid credits with usage-based redemption
- Same concurrency as Pay As You Go
- Discord and community support
Enterprise
Enterprise-grade solution with Best discounts and Custom-trained models and dedicated support.
- Best discounts
- Custom-trained models
- Priority access to new features
- Highest concurrency levels
- Self-hosted deployment options
- Paid support options
- Community access
Capabilities
Key Features
- Real-time and batch speech recognition
- Streaming and REST APIs
- Text-to-speech with natural-sounding voices
- Voice Agent API for agentic workflows
- Audio Intelligence for summarization, sentiment, and more
- Custom model training (Enterprise only)
- Speaker diarization and redaction
- Support for multiple languages and formats