Text-to-speech

POST /v1/tts turns text into audio in one request. Use it for narration, notifications and pre-rendered lines. For live use, call say() inside a realtime session instead.

curl https://api.vaux.ng/v1/tts \
  -H "Authorization: Bearer $VAUX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "text": "We did it!", "voice": "maya", "emotion": "excited", "format": "mp3" }' \
  --output speech.mp3

Field	Type	Description
`text`	string, ≤ 5,000 chars	What to say
`voice`	string	Voice slug, name or id
`emotion`	string	Optional. See emotions
`speed`	0.6–1.5	Optional
`volume`	0.5–2	Optional
`language`	string	Optional, default `en`
`format`	`mp3` · `wav`	Default `mp3`

The response body is the audio. The X-Credits-Charged header shows the cost (billed per 1,000 characters).