Transcribing audio files

The V‑Cloud API consists of two core methods:

  • The /transcribe endpoint accepts POST requests containing an audio file or a zip file less than 250MiB containing one or more audio files to transcribe. Successful /transcribe requests return a requestid to identify results for retrieval and tracking.

  • The /transcribe/result endpoint accepts GET requests containing a valid requestid . Successful /transcribe/result requests return a URL with the results as a zip or JSON file. The format of your results depends on the options specified in the initial /transcribe request.

    Note:

    V-Cloud limits the /transcribe endpoint to 128 concurrent connections per IP address and the /transcribe/result endpoint to 32 concurrent connections per IP address.

The following sections provide reference information for and examples of using each of these methods, plus an example of setting up a simple callback server and using it with the /transcribe endpoint.

Passing transcription parameters

V‑Cloud transcription parameters, often referred to as stream tags, enable you to customize how each audio file will transcribe. Transcription parameters are key-value pairs passed in the body of the request. The syntax of a parameter is name=value , as shown in the following example:

curl -F file=@sample1.wav \
     -F token=your-token-here \
     -F emotion=true \
     https://vcloud.vocitec.com/transcribe

Refer to V‑Cloud transcription parameters for more information.