Model API

Eleven v4 Turbo Dialogue

Fast, lower-cost multi-speaker dialogue with ElevenLabs Eleven v4 Turbo. Write the conversation turn by turn, give each turn a voice, and get one audio track with natural timing, emotion tags such as [laughing] or [whispering] and up to 10 voices. For podcasts, ads, explainers and game scenes.

Model id

Use this in every request

eleven-v4-turbo-dialogue

Run it

Quote the input, start the generation with the quote, then poll the job. Set RAYWAKE_API_KEY first; see Authentication.

# 1. Price the exact input (quotes are valid for up to 10 minutes)
curl -s https://api.raywake.com/v1/quotes \
  -H "Authorization: Bearer $RAYWAKE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "eleven-v4-turbo-dialogue",
  "input": {
    "prompt": "A red ceramic cup on a wooden table, soft morning light"
  }
}'

# 2. Start the generation with the quote_id from step 1
curl -s https://api.raywake.com/v1/generate \
  -H "Authorization: Bearer $RAYWAKE_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $(uuidgen)" \
  -d '{"model": "eleven-v4-turbo-dialogue", "quote_id": "QUOTE_ID", "input": {"prompt":"A red ceramic cup on a wooden table, soft morning light"}}'

# 3. Poll until the status is terminal
curl -s https://api.raywake.com/v1/jobs/JOB_ID \
  -H "Authorization: Bearer $RAYWAKE_API_KEY"

Input

Input fields for eleven-v4-turbo-dialogue
Name	Type	Description
`prompt`required	string	Required. A JSON list of turns, each with a voice and the text it speaks: [{"voice": "sarah", "text": "[excited] Did you hear the news?"}, {"voice": "george", "text": "No, what happened?"}]. Voices: adam, alice, bella, bill, brian, callum, charlie, chris, daniel, eric, george, harry, jessica, laura, liam, lily, matilda, river, roger, sarah, will. Up to 2,000 characters of spoken text and 10 different voices. Audio tags such as [whispering], [laughing] or [excited] set the delivery. Priced per character of this field.
`outputFormat`	string (enum)	Required. Default: `mp3_44100_128` One of: `mp3_22050_32`, `mp3_44100_32`, `mp3_44100_64`, `mp3_44100_96`, `mp3_44100_128`
`languageCode`	string (enum)	Optional. Auto lets the model detect the language from the text. Choosing one makes it speak and read numbers and dates in that language. Default: `auto`
`timestamps`	string (enum)	Optional. When true, a JSON file is returned along with the audio, with the start and end time of every word, which voice speaks it, and the name of each voice. Default: `false` One of: `false`, `true`
`applyTextNormalization`	string (enum)	Optional. Spells out numbers, dates and abbreviations before speaking them, e.g. "3/4" becomes "three quarters". Auto lets the model decide. Default: `auto` One of: `auto`, `on`, `off`
`stability`	number	Optional. Controls how consistent the delivery is. Lower values are more expressive, higher values keep a steadier delivery. Default: `0.5`
`similarityBoost`	number	Optional. How closely the output follows the reference voice. Higher values match it more closely but can sound less natural. Default: `0.75`
`seed`	number	Optional. The same seed with the same settings gives a near identical result. Leave empty for a random result.

Output

A succeeded job lists its results under outputs, stored on our side. See Jobs for every field.

Job excerpt

"outputs": [
  { "type": "audio", "url": "https://…", "mime_type": "audio/mpeg", "thumbnail_url": null }
]

Pricing

1,000 characters$0.07/ request

The exact charge for your input comes from a quote. How charges settle