Generate speech

Inference (JSON + base64)

POSThttps://api.lucidweights.ai/v1/tts/inference

Converts text into speech and returns base64-encoded audio in JSON. Only speakers present in the registry are accepted.

Field	Type	Default	Description
text	string	required	Text to convert to speech
speaker	string	required	Registered speaker name
language	string	default	ISO 639-1 language code
output_format	string	pcm_24000	See output formats on the overview page
speed	float	1.0	Range 0.1–2.0
expressive_level	integer	30	Range 0–100; optimal 25–65
custom_pronunciations	object	{"words":{},"phonemes":{}}	Per-request pronunciation overrides
translate_text	boolean	false	Translate input before synthesis

Response

{
  "audio": "<base64-encoded audio>"
}

NOTE

If the speaker is not registered, the request returns 400 with a list of valid speaker names in detail.

Example

curl -X POST "https://api.lucidweights.ai/v1/tts/inference" \
  -H "accept: application/json" \
  -H "x-api-key: lw_YOUR_SECRET" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "The quick brown fox jumps over the lazy dog.",
    "speaker": "emily",
    "language": "en",
    "output_format": "pcm_24000",
    "speed": 1,
    "expressive_level": 30
  }'

File inference

POSThttps://api.lucidweights.ai/v1/tts/file-inference

Same request body as inference. Returns an audio file (audio/wav or audio/mpeg depending on output_format).

Example

curl -X POST "https://api.lucidweights.ai/v1/tts/file-inference" \
  -H "x-api-key: lw_YOUR_SECRET" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "The quick brown fox jumps over the lazy dog.",
    "speaker": "emily",
    "language": "en",
    "output_format": "pcm_24000"
  }' \
  --output output.wav