Replace a saved voice's audio
Replace a saved voice's audio and its transcript together. Without text the transcript is cleared and the next generation transcribes the new audio.
voice_id*Voice Idapplication/json- body
audio_base64*Audio Base64Base64-encoded replacement audio (WAV)
text?|Transcript of the new audio. Omit it to clear the transcript. The next generation then transcribes the new audio.
word_timestamps?array<>|Word-level timestamps for text
transcript_manual?|True if the user wrote or corrected text
Successful Response
application/json- response
success*SuccessWhether the operation succeeded
message?MessageHuman-readable result message
""curl -X POST 'http://127.0.0.1:8100/api/v1/voices/string/audio' \ -H 'Content-Type: application/json' \ -d '{ "audio_base64": "UklGRi4AAABXQVZFZm10IBAAAA..."}'{ "success": true, "message": ""}Persist a transcript for a saved voice POST
Store a transcript on a saved voice and invalidate its baked clone prompt.
Align a saved voice's transcript to its audio POST
Find word timings for a saved voice's transcript, transcribing the audio when the voice has none. Waits until the voice is aligned. A voice that is already aligned, or holds no speech, returns at once unless `force` is set.