Deepgram · Speech to text
nova-3
General speech recognition for real-world audio
Pricing
Charged by actual usage
See the app for this model’s price.
Lifetime and Team members get 40% off this model. See plans
About this model
Nova-3 is Deepgram’s general speech recognition model for meetings, multiple speakers, noisy recordings, and far-field audio. It provides multilingual transcription and stronger recognition of domain-specific terminology.
Use in Beav
- 01Download Beav and sign in
- 02Choose nova-3 in the model picker
- 03Use it for turning audio and video into text and subtitles
Similar models
View allAlibaba Cloud
fun-asr
Speech transcription across Chinese dialects and languages
4.26 points / minute
Alibaba Cloud
Qwen Audio 3.1 ASR Filetrans
Long-audio transcription across languages and Chinese dialects
Input 104 · Output 351 points / 1M tokens
Alibaba Cloud
qwen3-asr-flash-filetrans
Multilingual transcription for recorded audio
4.26 points / minute
