The AI arena is free today

Open Superagent
DeepgramReleased on Oct 1, 2024

Nova 3: Pricing, Performance & Examples

Nova 3 is a speech-to-text model from Deepgram, released in October 2024.

Latest model with instant vocabulary updates and improved accuracy

Input
Audio
Output
Text

Nova 3 API

POST/v1/stt/transcribe

Any audio format up to 25 MB.

Missing Audio file

Run a request to see the response

Use it in your code

OpenAI-compatible endpoint through the LLM Stats gateway.

import requests

with open("audio.mp3", "rb") as f:
    response = requests.post(
        "https://gateway.llm-stats.com/v1/stt/transcribe",
        headers={"Authorization": "Bearer YOUR_API_KEY"},
        files={"file": f},
        data={"model_id": "nova-3"},
    )

print(response.json()["text"])

Need an API key? Create one above in the playground, or read the API documentation.

Nova 3 latency

Nova 3 time to first token, sustained output throughput, and failed-request rate from live API traffic over the trailing 7 days.

Nova 3 license

Nova 3 is a proprietary model available under its provider's product and API terms.

License
Proprietary
Hosted access

Proprietary license - usage restrictions apply

FAQ

Common questions about Nova 3.

Who created Nova 3?

Nova 3 was created by Deepgram.

What is the license for Nova 3?

Nova 3 is released under the Proprietary license.

Is Nova 3 multimodal?

Yes, Nova 3 is multimodal and can accept both text and images as input.

Where can I use Nova 3?

Nova 3 is available through 1 provider including Deepgram.