We are very excited to announce our latest Speech Understanding feature – Topic Detection (detect_topics=true). Deepgram’s Topic Detection API, in addition to our speech-to-text API, uses deep learning techniques to generate, identify and extract key topics from the audio data.

Topic Detection API endpoint can be called using (detect_topics=true) and will return topics block in the response body of the output. Each segment of the topics will have a section of the text, identified topics, confidence score, and word positions.

Examples of use cases for Topic Detection include:

  • Help the Quality Assurance team to analyze conversations based on discussed topics and identify trends and patterns.

  • Extract meaningful and actionable insights from conversations and audio data based on discussed topics.

  • Categorize and tag conversations based on identified topics to enhance search and recommendation capabilities.

The URL query might look like this to call the Topic Detection API. 


Please see the features page in our documentation to learn more about Topic Detection.

We’d love to have your feedback; please share it with us at Product Feedback, or you can also get in touch with us through your dedicated support channel.

Stop building work-arounds for STT systems that don't work.

Start FreeTalk to an expert