• audio_type: one of [“mp3”, “wav”, “flac”, “webm”]
• return_type: one of [“text”, “json”]
• audio_content: base64 encoded audio file.

audio_type should match base64 encoded audio file supplied in audio_content
If return_type: text, then the transcription response will be in string format. If return_type: json, then the transcription response will include a list of all words, corresponding confidence scores, start and end fields indicating the timestamp in which the word has started and ended.

Example bash script for sending STT requests;
export DATA=$(cat payload.wav | base64)
curl --request POST \ --url https://ardic-speech-to-text-service.p.rapidapi.com/stt \ --header 'X-RapidAPI-Host: ardic-speech-to-text-service.p.rapidapi.com' \ --header 'X-RapidAPI-Key: YOUR_RAPIDAPI_KEY' \ --header 'content-type: application/json' \ --data '{ "audio_type": "wav", "return_type": "text", "audio_content": $DATA }'

Disclaimer; Currently, the limit for file size is 1 MB. To be able to transcribe longer files with single request, we recommend using mp3 over wav since the compression let us create longer files with smaller file size.