Batch

Send a complete WAV recording in one request when you do not need results while it plays.

Send a file #

Send the WAV file as the body of POST /v1/transcriptions?mode=mt_asr&expected_num_speakers=auto, with your key as for any request (Authentication). The service runs it as a normal mt_asr session and collects that session's events. The audio format is the same as for streaming (Limits).

Target-speaker files #

ts_asr files are not accepted here: enrollment audio in a request header is fragile and sensitive. Stream the enrollment and the file through /v1/stream instead (Streaming).