Deepgram
Voice AI API with sub-100ms Nova-3 STT, Aura-2 TTS, and a unified Voice Agent API - 1,300+ enterprise customers; $229M raised at $1.3B valuation.
Deepgram is a Voice AI API platform offering speech-to-text (Nova-3), text-to-speech (Aura-2), and a unified Voice Agent API that orchestrates STT + TTS + LLM in a single endpoint for real-time conversational agents. Founded in 2015 and publicly available since 2021, Deepgram reached a $1.3B valuation with its $130M Series C in January 2026 and reported $21.8M ARR in 2024 - with the company describing itself as cashflow-positive in 2025. Nova-3 supports 10+ languages with streaming latency under 100ms, while the Voice Agent API eliminates the multi-library orchestration complexity of building voice-first products. Self-hosted on-premise deployment is available for healthcare, financial services, and government compliance environments.
Key Features
- Nova-3 STT - $0.0048/min streaming, 10+ languages; Flux English at $0.0065/min for high-accuracy scenarios
- Aura-2 TTS - $0.030/1,000 characters with multiple pre-built voices and speed controls
- Voice Agent API - unified STT + TTS + LLM in one endpoint, eliminating multi-library orchestration for real-time voice agents
- Audio Intelligence API - summarization, topic detection, sentiment analysis, and intent recognition on transcripts
- Speaker Diarization - identifies and labels individual speakers in multi-participant audio recordings
- Self-hosted/on-premise deployment for compliance-sensitive industries including healthcare and financial services
- Redaction add-on - automatically removes PII and sensitive data from transcripts in real time
Use Cases
- Voice AI developers building real-time conversational agents who need sub-100ms STT latency to eliminate the pause that makes AI calls feel robotic
- Enterprise compliance teams deploying on-premise transcription for healthcare or financial calls where audio cannot leave their data environment
- Media companies transcribing large podcast and video archives with speaker-labeled, sentiment-analyzed output at $0.0077/min
- Contact center software vendors integrating real-time transcription and intent recognition directly into agent assistance platforms
Pros
- Voice Agent API combines STT + TTS + LLM in one call - eliminates the multi-API orchestration complexity that makes real-time voice agents slow and brittle
- $229M raised at $1.3B valuation with 1,300+ enterprise customers including NASA, Spotify, and Twilio - battle-tested at production scale
- $200 free credit with no expiry and no card required lets developers build and fully test a proof-of-concept before committing to paid usage
Cons
- Pure pay-per-use pricing with no flat monthly cap - production voice agent costs are difficult to predict without careful usage monitoring and rate limiting
- On-premise deployment requires a separate enterprise contract - there is no self-serve path for teams that need data residency
- $21.8M ARR against $229M raised means the company remains heavily venture-backed - long-term pricing stability depends on continued investor support
Deepgram Alternatives
Explore similar tools and alternatives
Looking for alternatives to Deepgram? Here are some similar tools you might like:
AssemblyAI
Speech AI API with best-in-class transcription, speaker diarization, sentiment analysis, and LeMUR LLM-over-audio.
Cartesia
Real-time voice AI platform - Sonic TTS achieves sub-90ms latency using State Space Models; $191M raised across seed through Series B.
ElevenLabs
AI voice generation platform for creating realistic text-to-speech, voice cloning, and multilingual dubbing.
Deepgram is also listed as an alternative to:
Ready to try Deepgram?
Visit the official website to explore all features and get started with Deepgram today.
Reviews
0 reviews for Deepgram
Based on 0 reviews
Share your experience
Log in to write a review for Deepgram
ElevenLabs
AI voice generation platform for creating realistic text-to-speech, voice cloning, and multilingual dubbing.
Murf AI
AI voice generator with 200+ voices in 35+ languages, 55ms API latency, and AI dubbing with lip sync.
Suno AI
AI music generator that creates full songs with vocals, instruments, and lyrics from a text prompt in seconds.
Have an AI Tool?
List your AI tool for free, or go featured for top placement in your category - and reach thousands of potential users.
Submit Your Tool