The Clarify API analyzes media and exposes the data as insights. The following is a list of available insights:


Classification includes the detected language and acoustic environment.

Spoken Words

Statistics on the speech and conversation dynamics in the audio, including word count, speaking speed, duration of speech, crosstalk, and interruptions.

Spoken Keywords

Keywords list the most relevant terms spoken in the audio. Also includes named entities which are people, places, organizations, products, and dates etc. mentioned.

Spoken Topics

Topics are the subjects being spoken about in the audio and are listed by categories and related terms.

PCI Data

PCI data lists the audio segments that contain credit cards, verification numbers and expiry dates.

Fork me on GitHub