LLM Reference

Deepgram Nova Models by Deepgram

DeepgramProprietaryAudio
3 models2023–2025

Details

ResearcherDeepgram
LicenseProprietary
Commercial useCommercial use: conditional
Models3
Released2023–2025

Capabilities

MultimodalAll models

Links

Website

About

Deepgram's Nova family contains general-purpose speech-to-text models optimized for high-accuracy transcription across batch and streaming workflows.

Current Variants

Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.

3 in view

Use when the workload needs speech recognition, multimodal inputs, and audio.

2025-02speech recognitionmultimodal inputsaudio

Use when the workload needs speech recognition, multimodal inputs, and audio.

2023-09speech recognitionmultimodal inputsaudio
DeepgramCurrent

Use when the workload needs speech recognition, multimodal inputs, and audio.

2023-01speech recognitionmultimodal inputsaudio

Release Timeline

3 release groups
2025-02
1 current
Deepgram Nova-3
speech recognitionmultimodal inputsaudio
Current
2023-09
1 current
Deepgram Nova-2
speech recognitionmultimodal inputsaudio
Current
2023-01
1 current
Deepgram
speech recognitionmultimodal inputsaudio
Current

Specifications(3 models)

Deepgram Nova model specifications comparison
ModelReleasedMultimodal
Deepgram Nova-32025-02Yes
Deepgram Nova-22023-09Yes
Deepgram2023-01Yes

Available From(1 provider)

Frequently Asked Questions

What is Deepgram Nova used for?
Deepgram Nova is used for audio, speech recognition, and vision and multimodal work. The family description and listed model capabilities point to those workloads as the best fit.
How does Deepgram Nova compare to MOSS-TTS?
Deepgram Nova by Deepgram is strongest where you need audio, while MOSS-TTS by MOSI AI is the closest related family to check for audio. Deepgram Nova has 3 listed variants, so compare the specs and pricing tables before choosing a production model.
Which Deepgram Nova model should I use?
If price is the main constraint, use the pricing table first because Deepgram Nova does not have complete provider pricing in the local data. For the most capable/latest local choice, evaluate Deepgram Nova-3 with multimodal inputs.