Feature Overview | Deepgram's Docs

To learn how to get up and running with Speech-To-Text Intelligence, read the Audio Intelligence getting started guide.

Speech-To-Text Intelligence features

Feature	Pre-recorded?	Streaming?	Language(s)
Sentiment Analysis	Yes	No	English (all available regions)
Intent Recognition	Yes	No	English (all available regions)
Topic Detection	Yes	No	English (all available regions)
Summarization	Yes	No	English (all available regions)
Entity Detection	Yes	Yes*	English (all available regions)

*Entity Detection for streaming is supported on Nova, Nova 2, Nova 3, and Enhanced models. It is not available for Base models or Flux.

Rate Limits

For information on Deepgram’s Concurrency Rate Limits, refer to our API Rate Limits Documentation.

Deepgram Self-Hosted

Having challenges with performance and latency? Check out Deepgram’s Self-Hosted Solution to get the benefits of running your own hosted instance of Deepgram.