Speech to text

Semantria's Speech to Text endpoint will allow you to submit audio files and receive back transcribed text that has been enriched by Semantria. When you submit an audio file to the /speech endpoint you must use a configuration route via the --using parameter. You can also optionally specify a language ID but language identification will be performed if no language has been specified.

When you request results back, you will receive the complete transcription along with Semantria's NLP output. You can request the documents by sending a GET request to the /speech endpoint.

Speech to text uses a different set of API credits than document credits and is a feature that must be purchased separately from the text-only API. If a transcription is longer than the maximum text document length of your Semantria subscription, we will chunk the document into sections that fit into that length.