Multispeaker speech to text

Identify up to eight speakers and transcribe English, Hindi, Marathi, Kannada, or Telugu. File speaker detection supports up to 60 seconds. Live microphone transcription is available without speaker labels.

Developer guide

Client API keyThe key stays in this page’s memory.

Audio file

WAV, MP3, FLAC or OGG · up to 60 seconds and 25 MiB

Live microphone

Pause briefly between sentences. Transcripts arrive after each utterance.

Ready.

Last response