Filtering Profanity

    By default, the API will return an honest representation of the audio, meaning profanity will be present in the transcript if spoken in the audio. If you wish to filter profanity from the transcript, you can now do that!

    The returned transcript will not thoroughly scrub the word but show an unexplicit version of it. So, for example, you may see a sentence written like this:

    It was some tough s*** that they had to go through. But they did it; I mean, it blows my f****** mind every time I hear the story. 

    To filter profanity you need to add one additional parameter to your JSON payload—filter_profanity: true. Below is an example of a full request using CURL:

    curl --request POST \
      --url \
      --header 'authorization: YOUR-API-TOKEN' \
      --header 'content-type: application/json' \
      --data '{"audio_url": "", "filter_profanity": true'

    By setting filter_profanity: true, you will tell the model that for any known profane word, it should leave the first letter un-redacted for context and filter the remaining characters in the word.