Deepgram Aura Models by Deepgram
4 models2024–2025From $30/1M input
Last refreshed 2026-06-29. Next refresh: weekly.
Details
ResearcherDeepgram
LicenseProprietary
Commercial useCommercial use: conditional
Models4
Released2024–2025
Links
WebsiteAbout
Deepgram's Aura family provides hosted text-to-speech voices, including Aura-2 models for low-latency speech generation.
Decision facts
- Best fit
- audiotext to speech
- Capability starting point
- Deepgram Aura-2
- Lowest tracked input
- Deepgram Aura-2 · $30/1M · Deepgram API
- Closest related family
- MOSS-TTS
Current Variants
Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.
4 in view
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| Deepgram Aura-2 | Use when the workload needs audio. | 2025-04 | audio | Current |
| Aura-2 EN | Use when the workload needs text to speech and audio. | 2025-04 | text to speechaudio | Current |
| Aura-2 ES | Use when the workload needs text to speech and audio. | 2025-04 | text to speechaudio | Current |
| Aura-1 | Use when the workload needs text to speech and audio. | 2024-01 | text to speechaudio | Current |
Release Timeline
2 release groups2025-04
3 current
2024-01
1 current
Aura-1
Currenttext to speechaudio
Specifications(4 models)
| Model | Released |
|---|---|
| Deepgram Aura-2 | 2025-04 |
| Aura-2 EN | 2025-04 |
| Aura-2 ES | 2025-04 |
| Aura-1 | 2024-01 |
Available From(1 provider)
Pricing
| Model | Provider | Input / 1M | Output / 1M | Type |
|---|---|---|---|---|
| Deepgram Aura-2 | Deepgram API | $30 | — | Serverless |