voice_id as pipeline_config.tts_config.tts_voice_id when creating a Streaming Avatar session or saving a character.
There are two ways to create a voice: Clone Voice from reference audio you upload, or Design Voice from a text description. Both return a voice_id ready to use, and both count against the same max_voices quota.
For Clone Voice, upload the reference audio to the current workspace first: POST /v1/files/audios.
Response envelope. REST responses are wrapped in {code, message, data}: code = 0 means success. Field descriptions below refer to fields inside data.