Core resources

Voices

Browse the text-to-speech voice catalog across providers, and preview a voice before assigning it to an assistant via config.ttsVoiceId.

GET/voices agents:read

List voices

The voice catalog. Filter by search term or provider.

Query parameters

q	string	Match name / language / tags.
provider	string	azure \| sarvam \| cartesia \| elevenlabs \| xai (or 'all').

bash

curl "https://api.bolchoai.in/v1/voices?provider=azure" -H "Authorization: Bearer $BOLCHO_API_KEY"

Response

json

{ "voices": [ { "id": "en-US-AvaMultilingualNeural", "name": "Ava", "provider": "azure", "language": "en-US", "gender": "female", "tags": ["multilingual"] } ], "providers": ["azure", "sarvam", "cartesia", "elevenlabs", "xai"] }

POST/voices/preview agents:read

Preview a voice

Synthesize a short sample clip. Parameters are sent as query string; the response body is the raw audio (audio/mpeg, or audio/wav for sarvam).

Query parameters

provider*	string	azure \| sarvam \| elevenlabs \| xai \| cartesia.
voiceId*	string	The voice to synthesize.
language	string	BCP-47 language (optional).

bash

curl -X POST "https://api.bolchoai.in/v1/voices/preview?provider=azure&voiceId=en-US-AvaMultilingualNeural" \
  -H "Authorization: Bearer $BOLCHO_API_KEY" --output preview.mp3