Home / What Is TTS / Deepgram Aura-2
What Is Deepgram Aura-2?
Ultra-low latency TTS with 90ms time-to-first-byte, built for conversational AI.
Deepgram Aura-2 is a text-to-speech API purpose-built for real-time conversational AI. It delivers industry-leading latency with approximately 90ms time-to-first-byte, making it the fastest option for voice bots, virtual assistants, and interactive voice response systems. Aura-2 offers 93+ voices across 7 languages.
Official: Deepgram website
ProviderDeepgram
TierPremium
SpeedFast
Quality
Supported Languages (7)
en
es
fr
de
it
pt
nl
Strengths
- ~90ms latency
- 93+ voices
- Conversational AI optimized
- Streaming output
- Real-time synthesis
Limitations
- Does not support voice cloning
- Premium tier — requires credits
- Commercial API — higher per-character cost
How to Use Deepgram Aura-2 via WhatIsTTS
1. Create a free account and get your credits.
2. Go to Text to Speech and select Deepgram Aura-2.
3. Enter text, choose a voice, and click Generate. We route to Deepgram.
No API key management needed. We handle authentication, routing, and billing through our unified platform.
Compare Deepgram Aura-2 vs...
Frequently Asked Questions
Deepgram Aura-2 is a text-to-speech API purpose-built for real-time conversational AI. It delivers industry-leading latency with approximately 90ms time-to-first-byte, making it the fastest option for voice bots, virtual assistants, and interactive voice response systems. Aura-2 offers 93+ voices across 7 languages.
Real-time conversational AI, voice bots, IVR systems
No, Deepgram Aura-2 does not currently support voice cloning. For voice cloning, try ElevenLabs Multilingual v2, PlayHT, or Cartesia Sonic through our platform.
Deepgram Aura-2 supports 7 languages including English. Language availability may vary by voice. Check the voice catalog on our platform for the full list of supported languages and voices.
Visit our Text to Speech page, select Deepgram Aura-2 from the model dropdown, choose a voice, enter your text, and click Generate. You can also use our REST API: POST to /api/v1/tts/ with model="deepgram-aura-2" and your text. No provider API key needed — we handle routing and billing.
Deepgram Aura-2 is a premium-tier model costing 5 credits per 1,000 characters on WhatIsTTS. Premium models offer the highest quality and advanced features. Credits are included with all plans.
Deepgram Aura-2 is one of 20+ TTS models available on WhatIsTTS. Compare it with providers like ElevenLabs, OpenAI, Google Cloud, Azure, Amazon Polly, PlayHT, Deepgram, and Cartesia — all accessible through one platform with unified billing. Use our side-by-side comparison to find the best fit.
Yes. Deepgram Aura-2 is a commercial API from Deepgram that permits commercial use of generated audio. This includes podcasts, videos, apps, IVR systems, and more. Review Deepgram's terms of service for specific requirements.
Through WhatIsTTS, Deepgram Aura-2 output is available in MP3 (default), WAV, OGG, and FLAC formats. MP3 is ideal for web playback, WAV for further audio processing. Select your preferred format before generating.
Deepgram Aura-2 has fast generation speed. Optimized for real-time applications like voice chat and interactive experiences.
Yes. WhatIsTTS provides a unified REST API for all providers including Deepgram Aura-2. Generate an API key in your account, then POST to /api/v1/tts/ with model="deepgram-aura-2". We provide code examples in Python, JavaScript, cURL, and Go. No need to manage a separate Deepgram API key.
Deepgram Aura-2 by Deepgram offers: ~90ms latency, 93+ voices, Conversational AI optimized, Streaming output. Available as a premium-tier model on WhatIsTTS with 7 language support.
No. WhatIsTTS handles the Deepgram integration for you. You only need a WhatIsTTS account and credits. We route your requests to Deepgram's API and handle authentication, billing, and error handling.
Yes, Deepgram Aura-2 supports streaming output for low-latency applications. Audio chunks are delivered as they are generated, enabling real-time playback.