Deepgram · Audio model

Deepgram Nova-3 / Aura-2

Production speech APIs, billed by the second.

Generate on Coinis Read the docs

One API key. Every model.

per minute
Example
AI audio
Commercial use included Verified May 26, 2026 Outputs are yours No training on your data
Endpoints

Start building with Deepgram Nova-3 / Aura-2.

One model. Three ways to call it. Same key, same bill.

Deepgram Nova-3 (STT)

Audio generation

One call. Same key. Same bill.

$0.01 / minute

Deepgram Aura-2 (TTS)

Audio generation

One call. Same key. Same bill.

$0.05 / 1k chars

Capabilities

What it does best.

Deepgram capabilities

Nova-3. Speech-to-Text

  • Handles 45+ languages across monolingual and multilingual variants[^1]
  • Multilingual variant detects language automatically within a single audio stream[^1]
  • Streaming and pre-recorded transcription share the same per-minute rate. No real-time premium[^2]
  • True per-second billing: a 14-second file is billed for exactly 14 seconds, not rounded up to 15 or 60[^2]
  • Speaker Diarization add-on identifies and labels individual speakers in multi-speaker recordings[^2]
  • Smart Formatting included at no extra charge: adds punctuation, correct casing, dates, and currency[^2]
  • Keyterm Prompting add-on boosts recognition accuracy for domain-specific vocabulary[^2]
  • PII Redaction add-on removes SSNs, credit card numbers, and phone numbers from transcripts automatically[^2]
  • Optimized for noisy environments, crosstalk, and far-field audio[^1]

Aura-2. Text-to-Speech

  • Supports 7 languages: English, Spanish, German, French, Dutch, Italian, and Japanese[^3]
  • 40+ named English voices spanning accents, age profiles, and use-case profiles including IVR, customer service, storytelling, and advertising[^3]
  • Select Spanish voices support codeswitching, alternating between English and Spanish mid-conversation[^3]
  • Low-latency synthesis designed for interactive voice response systems and conversational AI[^3]

Platform limits

  • Pre-recorded audio files: up to 2 GB per request[^4]
  • Processing timeout: 10 minutes per request[^4]
  • Concurrency: up to 100 concurrent REST requests and 150 concurrent WebSocket streams[^4]
API

Call Deepgram Nova-3 / Aura-2 in three lines.

One key. One base URL. Same SDK shape you already use. 

# 1. set your key
export COINIS_API_KEY="sk_live_..."

# 2. call the model
curl https://api.app.coinis.com/v1/audio/generate \
  -H "Authorization: Bearer $COINIS_API_KEY" \
  -d '{"prompt":"neon city, rain, tracking shot"}'
import { Coinis } from "@coinis/sdk";
const coinis = new Coinis(process.env.COINIS_API_KEY);

const job = await coinis.audio.generate({
  model: "models/deepgram/deepgram",
  prompt: "neon city, rain, tracking shot",
});
from coinis import Coinis
coinis = Coinis(os.environ["COINIS_API_KEY"])

job = coinis.audio.generate(
    model="models/deepgram/deepgram",
    prompt="neon city, rain, tracking shot",
)
Response 
{
  "id": "gen_8fa2c1",
  "status": "succeeded",
  "model": "models/deepgram/deepgram",
  "output": {
    "image_url": 
                "https://cdn.coinis.com/gen_8fa2c1.mp4"
              
              ,
    "format": "mp4"
  },
  "tokens_used": 10
}

Already on another provider's SDK? Change the host. Keep the call.

Pricing

Token pricing. No surprises.

One wallet across every model. No API accounts to juggle.

Deepgram Nova-3 / Aura-2 · Deepgram Nova-3 (STT)
0.1 tokens
per minute · $0.01
Speech to text
$0.01 / minute
One key. Every model. One invoice. 1 token = $0.10
1 minute ≈ 0 tokens ($0.01)
Start free. 15 tokens a week.

No credit card.

Why pay through Coinis
  • One wallet for every model. No API keys. No separate bills.
  • Generate ads. Launch to Meta. Track in one place.
  • On-brand output from your Brand Profile.

1 token = $0.10 pay-as-you-go. Less on a plan.

Standard vs Fast

Pick the run for the job.

Deepgram Nova-3 (STT)

Final renders, studios
Resolution —
Price $0.01 / minute

Deepgram Aura-2 (TTS)

Rapid tests, high volume
Resolution —
Price $0.05 / 1k chars
Use cases

Two buyers. One model.

For builders

Resell every model. One key. One bill.

Unified API across video, image, audio, and LLM.

Generate 500 variants overnight.

Async queue plus webhooks. Batch at scale.

White-label the output.

Ship it under your brand. Outputs are yours.

For creatives

Ship a Reel before lunch.

Prompt to platform-native clip in minutes.

Same product. Ten formats.

One generation, every aspect ratio.

Commercial UGC without a creator.

Authentic selfie-style ads, on brand.

Real-time voice agents and conversational AI Nova-3 streaming powers live transcription for voice bots and virtual assistants.[^1] Aura-2 handles synthesis in the same pipeline, so you can build both transcription and speech generation into a single voice agent flow.[^3]

Call center transcription with diarization Nova-3's Speaker Diarization add-on labels each speaker in a recorded customer call.[^2] Pair it with Smart Formatting to get clean, readable transcripts ready for QA review or CRM logging.

Compliance and PII-sensitive transcription Nova-3's PII Redaction add-on strips SSNs, credit card numbers, and phone numbers from transcripts automatically.[^2] Regulated industries get a clean output without building a separate scrubbing layer.

Multilingual voice products Nova-3 Multilingual detects languages automatically in a single audio stream.[^1] Aura-2 adds TTS output across 7 languages for end-to-end voice interfaces in globally deployed applications.[^3]

IVR and voice-assistant front-ends Aura-2's 40+ English voices are tuned for interactive voice response systems and conversational AI front-ends.[^3] Pick a voice profile that matches your brand tone and deploy it at scale with per-character billing.

Renders in seconds. Set a seed. Get the same frame back.

Outputs are yours. Sell them.

Safe for paid ads.

Your prompts are never used for training.

FAQ

Deepgram Nova-3 / Aura-2 FAQs

How much does Deepgram Nova-3 cost per minute on Coinis?

Deepgram Nova-3 (STT) is priced at $0.013 per minute on Coinis. Deepgram's direct Pay-As-You-Go rate for Nova-3 pre-recorded transcription is $0.0077 per minute. The Coinis retail price includes platform access, unified billing, and a single API key across all catalog models.

What is the difference between Deepgram Nova-3 and Aura-2?

Nova-3 is a speech-to-text model. It transcribes audio into text, supporting 45+ languages, noisy audio, and multi-speaker recordings. Aura-2 is a text-to-speech model. It converts written text into spoken audio across 7 languages with 40+ named English voices.

Does Deepgram support real-time streaming transcription, and is it priced differently?

Yes, Nova-3 supports both real-time WebSocket streaming and pre-recorded REST transcription. On Coinis, both modes are billed at the same $0.013 per minute rate. There is no separate premium for the low-latency streaming path.

Which languages do Nova-3 and Aura-2 support?

Nova-3 supports 45+ languages across its monolingual variants. The Multilingual variant detects multiple languages automatically within a single audio stream. Aura-2 TTS supports 7 languages: English (with American, British, Australian, Irish, and Filipino accents), Spanish (with multiple regional accents), German, French, Dutch, Italian, and Japanese.

Is there a Deepgram API I can call directly from Coinis?

Yes. Send a POST request to https://api.app.coinis.com/v1/audio/generate with your Coinis API key. You get the same Nova-3 and Aura-2 models, the same pricing, and no separate Deepgram account required. Full endpoint documentation is on the Deepgram API sub-page.

What are the file size and concurrency limits for Nova-3?

Pre-recorded audio files are capped at 2 GB per request, and each request times out after 10 minutes of processing. Pay-As-You-Go accounts support up to 100 concurrent REST requests and 150 concurrent WebSocket streams. For multichannel audio, billing is based on total processed duration: a 10-minute stereo file is billed as 20 minutes.

How does Deepgram Aura-2 compare to ElevenLabs for text-to-speech?

Aura-2 on Coinis is billed at $0.051 per 1,000 characters. It covers 7 languages and 40+ named English voices, with select Spanish voices supporting codeswitching. ElevenLabs is available as a separate model on Coinis. Compare both on the Coinis audio catalog to find the voice and price that fits your workflow.

Can I use Deepgram for HIPAA-compliant or PII-redacted transcription?

Nova-3's PII Redaction add-on automatically removes sensitive identifiers such as SSNs, credit card numbers, and phone numbers from transcripts. HIPAA compliance with a signed BAA is available only on Deepgram's Enterprise plan. If HIPAA is a hard requirement, contact the Coinis team to discuss Enterprise routing options.

Related

Pair it with.

ElevenLabs (Multilingual / Turbo / Music)

Model

OpenAI Whisper

Model

Sync Labs Lipsync 1.9 / 2

Model

All audio models

Catalog
Upscale this Edit a frame Add a voiceover

Start free

Your wallet. Every model. One call away.

Start free. 15 tokens a week. No card.

Generate on Coinis

No credit card.

Pricing and capabilities verified 2026-05-26. Read the docs .