ElevenLabs Text to Dialogue v3 API

POST /v1/tasks

All models are called through the Unified Async API POST /v1/tasks endpoint; only the input fields differ (see input parameters below).

Model summary

Model name	elevenlabs/text-to-dialogue-v3
Type	Audio generation (text-to-dialogue)
Endpoint	POST /v1/tasks
Pricing	See HiAPI Pricing (per character)

ElevenLabs Eleven v3 multi-speaker text-to-dialogue: turn scripted lines into natural, expressive conversation audio with a distinct voice per line, 70+ languages, billed per character.

Production guidance

For production, pass callback.url at the top level of the request body so HiAPI can notify your service when the task reaches a terminal state.
GET /v1/tasks/:id is better for local debugging, low-volume jobs, or fallback reconciliation if a callback is missed.
Use callback.when=final. Both success and fail are terminal states, so your service should deduplicate by taskId.
Split long scripts into multiple calls under the 5000-character limit; keep the same voice_id per character across calls for a consistent sound.

Best suited for

Audio dramas & audiobooks

Assign a voice per line and generate whole conversation passages in one call — Eleven v3 excels at expressive narration.

dialogue

Podcasts & interview intros

Host + guest two-voice dialogue for podcast openers, news reads, and interview simulations.

dialoguestability

Game NPCs & interactive stories

Pin one voice_id per character for a consistent sound and batch-generate quest dialogue.

dialogue

Multilingual voiceover

70+ languages with auto-detection by default; rewrite the text per language to produce localized versions.

language_code

Request parameters

POST https://api.hiapi.ai/v1/tasks

# Create an async task — returns taskId immediately (generation runs async)
# Replace YOUR_API_KEY with your account API key
curl -X POST "https://api.hiapi.ai/v1/tasks" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "elevenlabs/text-to-dialogue-v3",
  "input": {
    "dialogue": [
      {
        "text": "Welcome to HiAPI — one API for every AI model.",
        "voice": "EkK5I93UQWFDigLMpZcX"
      },
      {
        "text": "Really? Video, image and speech, all with the same key?",
        "voice": "Z3R5wn05IrDiVCyEkUrK"
      }
    ],
    "stability": 0.5
  }
}'

import requests, json

# Business params — fields vary by model (see the model page)
payload = json.loads(r'''
{
  "model": "elevenlabs/text-to-dialogue-v3",
  "input": {
    "dialogue": [
      {
        "text": "Welcome to HiAPI — one API for every AI model.",
        "voice": "EkK5I93UQWFDigLMpZcX"
      },
      {
        "text": "Really? Video, image and speech, all with the same key?",
        "voice": "Z3R5wn05IrDiVCyEkUrK"
      }
    ],
    "stability": 0.5
  }
}
''')

# Create an async task — returns taskId immediately (generation runs async)
resp = requests.post(
    "https://api.hiapi.ai/v1/tasks",
    headers={"Authorization": "Bearer YOUR_API_KEY", "Content-Type": "application/json"},
    json=payload,
)
print(resp.json())  # Returns { code, message, data: { taskId } }

// Create an async task — returns taskId immediately (generation runs async)
// Replace YOUR_API_KEY with your account API key
const resp = await fetch("https://api.hiapi.ai/v1/tasks", {
  method: "POST",
  headers: {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
  "model": "elevenlabs/text-to-dialogue-v3",
  "input": {
    "dialogue": [
      {
        "text": "Welcome to HiAPI — one API for every AI model.",
        "voice": "EkK5I93UQWFDigLMpZcX"
      },
      {
        "text": "Really? Video, image and speech, all with the same key?",
        "voice": "Z3R5wn05IrDiVCyEkUrK"
      }
    ],
    "stability": 0.5
  }
}),
});
console.log(await resp.json()); // Returns { code, message, data: { taskId } }

model string required

Fixed value elevenlabs/text-to-dialogue-v3.

example elevenlabs/text-to-dialogue-v3

input object required

Business parameters. Put ElevenLabs Text to Dialogue v3-specific configuration here.

dialogue object[] required

Array of dialogue lines, each { text, voice }, synthesized in order into one conversation audio. · Total characters across all text fields must not exceed 5000; requests over the limit return 400 (no charge) · Keep the same voice for the same character across lines for a consistent sound · All 67 preset voices are listed in the Preset voices table below

text string required

The spoken text of this line (at least 1 character).

voice string required

Voice ID for this line (67 presets, defaults to James). See the full Preset voices table below ↓

default EkK5I93UQWFDigLMpZcX

language_code enum optional

Language of the generated speech (ISO code). Leave empty to auto-detect. See the full Supported languages table below ↓

default "" enum: ""afarhyasazbebnbsbg +65

stability enum optional

Voice stability: 0 = more emotional and varied, 0.5 = balanced, 1 = most stable and consistent.

default 0.5 enum: 00.51

callback object optional

Optional callback configuration. When set, HiAPI notifies your service when the task reaches a terminal state.

url string required

Required when callback is set; HTTPS URL that receives terminal task notifications.

example https://your-domain.com/hiapi/callback

when enum optional

Callback trigger timing. Use final.

default final enum: final

Preset voices (67)

Pass a voice_id from this table in dialogue[].voice; keep one voice per character across lines for a consistent sound.

Show all 67 voices

Voice	Style	voice_id
James	Husky, Engaging	`EkK5I93UQWFDigLMpZcX`
Arabella	Mysterious, Emotional	`Z3R5wn05IrDiVCyEkUrK`
Bradford	Expressive, Articulate	`NNl6r8mD7vthiJatiJt1`
Xavier	Bold, Metallic Announcer	`YOq2y2Up4RgXP2HyXjE5`
Kuon	Cheerful, Clear, Steady	`B8gJV1IhpuegLxdpXFOE`
Monika Sogam	Deep, Natural	`2zRM7PkgwBPiau2jvVXc`
Mark	Casual, Relaxed	`1SM7GgM6IMuvQlz2BwM3`
Adeline	Feminine, Conversational	`5l5f8iK3YPeGga21rQIX`
Sam	Support Agent	`scOwDtmlUjD3prqpp97I`
Spuds Oxley	Wise, Approachable	`NOpBlnGInO9m6vDvFkFC`
Eve	Genuine, Lively, Happy	`BZgkqPqms7Kj9ulSkVzn`
Northern Terry	—	`wo6udizrrtpIxWGp2qJk`
British Football Announcer	—	`gU0LNdkMOQCOrPrwtbee`
Brock	Commanding Sergeant	`DGzg6RaUqxGRTHSBjfgF`
Nathan	Virtual Radio Host	`x70vRnQBMBu4FAYhjJbO`
Anika	Lively, Friendly, Engaging	`Sm1seazb4gs7RSlUVw7c`
Viraj	Rich, Soft	`P1bg08DkjqiVEzOn76yG`
Taksh	Calm, Serious, Smooth	`qDuRKMlYmrm8trt5QyBn`
Horatius	Energetic Character	`qXpMhyvQqiRxWQs4qSSB`
Liam	Energetic Creator	`TX3LPaxmHKxFdv7VOQHJ`
Callum	Husky Rogue	`N2lVS1w4EtoT3dr4eOWO`
Laura	Warm, Quirky	`FGY2WhTYpPnrIDTdsKH5`
Brittney	Fun, Young, Informative	`kPzsL2i3teMYv0FxEYQ6`
Mark	Natural Conversation	`UgBBYS2sOqTuMpoF3BR0`
Bella	Professional, Bright, Warm	`hpp4J3VqNfWAUOO0d1Us`
Brian	Deep, Resonant, Reassuring	`nPczCjzI2devNBz1zQrb`
Hope	Lively, Gossipy, Youthful	`uYXf8XasLslADfZ2MB4u`
Jeff	Elegant, Resonant, Strong	`gs0tAILXbY5DNrJrsM6F`
Jamahal	Young, Energetic, Natural	`DTKMou8ccj1ZaWGBiotd`
Finn	Young, Eager, Energetic	`vBKc2FfBKJfcZNyEt1n6`
Tom	Conversation & Books	`DYkrAHD8iwork3YSUBbs`
Cassidy	Crisp, Direct, Clear	`56AoDkrOh6qfVPDXZ7Pt`
Addison 2.0	Australian Audiobook	`eR40ATw9ArzDf9h3v7t7`
Jessica Anne Bogart	Chatty, Friendly	`g6xIsTj2HwM6VR4iXFCw`
Lucy	Fresh, Casual	`lcMyyd2HUfFzxdCaC4Ta`
Tiffany	Natural, Warm	`6aDn1KB0hjpdcocrUkmq`
Felix	Warm, Positive, Modern RP	`Sq93GQT4X1lKDXsQcixO`
Jessica Anne Bogart	Eloquent Villain	`flHkNRp1BlvT73UL6gyz`
Lutz	Giggly, Cheerful	`9yzdeviXkFddZ4Oz8Mok`
Emma	Cute, Upbeat	`pPdl9cQBQq4p6mRkZy2Z`
Edward	Loud, Confident, Cocky	`zYcjlYFOd3taleS0gkk3`
Marshal	Friendly, Witty Professor	`nzeAacJi50IvxcyDnMXa`
John Morgan	Rugged Cowboy	`ruirxsoakN0GWmGNIo04`
Aria	Sultry Villain	`TC0Zp7WVFzhA8zpTlRqV`
Viking Bjorn	Epic Medieval Raider	`ljo9gAlSqKOvF6D8sOsX`
Pirate Marshal	—	`PPzYpIqttlTYA83688JI`
Johnny Kid	Serious, Calm Narration	`8JVbfL6oEdmuxKn5DK2C`
Hope	Poetic, Romantic, Charming	`iCrDUkL56s3C8sCRl7wb`
Ana Rita	Smooth, Expressive, Bright	`wJqPPQ618aTW29mptyoc`
John Doe	Deep	`EiNlNiXeDU1pqqOPrYMO`
Burt Reynolds™	Deep, Smooth, Clear	`4YYIPFl9wE5c4L2eu2Gb`
Hank	Deep, Engaging Narration	`6F5Zhi321D3Oq7v1oNT4`
Wyatt	Wise Country Cowboy	`YXpFCvM1S3JbWEJhoskW`
Phil	Explosive, Passionate Announcer	`LG95yZDEHg6fCZdQjLqj`
Johnny Dynamite	Retro Radio DJ	`CeNX9CMwmxDxUF5Q2Inm`
Rachel M	Pro British Radio Host	`aD6riP1btT197c6dACmy`
Rex Thunder	Deep, Tough	`mtrellq69YZsNwzUSyXh`
Ed	Late-Night Announcer	`dHd5gvgSOzSfduK4CvEg`
Jean	Charming, Playful Femme Fatale	`eVItLK1UvXctxuaRV2Oq`
Britney	Cool, Calculating Villain	`esy0r39YPLQjOczyOib8`
Sven	Emotive, Kind	`Tsns2HvNFKfGiNjllgqo`
Viraj	Smooth, Gentle	`1U02n4nD6AdIZ9CjF053`
Nathaniel	Engaging, British, Calm	`AeRdCCKzvd23BpJoofzx`
Benjamin	Deep, Warm, Calm	`LruHrtVF6PSyGItzMNHS`
Allison	Calm, Soothing, Meditative	`1wGbFxmAM3Fgw63G1zZJ`
Theodore HQ	Serene, Composed	`hqfrgApggtO1785R4Fsn`
Leon	Soothing, Composed	`MJ0RnG71ty4LH3dvNfSd`

Supported languages (73)

All accepted language_code values; leave empty (default) to auto-detect.

Show all 73 languages

Language	Code
Auto-detect (default)	`""`
Afrikaans	`af`
Arabic	`ar`
Armenian	`hy`
Assamese	`as`
Azerbaijani	`az`
Belarusian	`be`
Bangla	`bn`
Bosnian	`bs`
Bulgarian	`bg`
Catalan	`ca`
Cebuano	`ceb`
Nyanja	`ny`
Croatian	`hr`
Czech	`cs`
Danish	`da`
Dutch	`nl`
English	`en`
Estonian	`et`
Filipino	`fil`
Finnish	`fi`
French	`fr`
Galician	`gl`
Georgian	`ka`
German	`de`
Greek	`el`
Gujarati	`gu`
Hausa	`ha`
Hebrew	`he`
Hindi	`hi`
Hungarian	`hu`
Icelandic	`is`
Indonesian	`id`
Irish	`ga`
Italian	`it`
Japanese	`ja`
Javanese	`jv`
Kannada	`kn`
Kazakh	`kk`
Kyrgyz	`ky`
Korean	`ko`
Latvian	`lv`
Lingala	`ln`
Lithuanian	`lt`
Luxembourgish	`lb`
Macedonian	`mk`
Malay	`ms`
Malayalam	`ml`
Chinese	`zh`
Marathi	`mr`
Nepali	`ne`
Norwegian	`no`
Pashto	`ps`
Persian	`fa`
Polish	`pl`
Portuguese	`pt`
Punjabi	`pa`
Romanian	`ro`
Russian	`ru`
Serbian	`sr`
Sindhi	`sd`
Slovak	`sk`
Slovenian	`sl`
Somali	`so`
Spanish	`es`
Swahili	`sw`
Swedish	`sv`
Tamil	`ta`
Telugu	`te`
Thai	`th`
Turkish	`tr`
Ukrainian	`uk`
Urdu	`ur`
Vietnamese	`vi`
Welsh	`cy`

Example requests

Two-voice product chat

Two characters alternate lines (James × Arabella) at the balanced default stability.

Request body

{
  "model": "elevenlabs/text-to-dialogue-v3",
  "input": {
    "dialogue": [
      {
        "text": "Welcome to HiAPI — one API for every AI model.",
        "voice": "EkK5I93UQWFDigLMpZcX"
      },
      {
        "text": "Really? Video, image and speech, all with the same key?",
        "voice": "Z3R5wn05IrDiVCyEkUrK"
      },
      {
        "text": "Exactly, and it is usage-based pricing.",
        "voice": "EkK5I93UQWFDigLMpZcX"
      }
    ],
    "stability": 0.5
  }
}

Podcast opener (pinned language)

Host + guest dialogue with language_code pinned to en and stability 1 for a steadier read.

Request body

{
  "model": "elevenlabs/text-to-dialogue-v3",
  "input": {
    "dialogue": [
      {
        "text": "Welcome to today's podcast.",
        "voice": "x70vRnQBMBu4FAYhjJbO"
      },
      {
        "text": "Happy to be here — today I want to talk about the future of AI voice.",
        "voice": "uYXf8XasLslADfZ2MB4u"
      }
    ],
    "language_code": "en",
    "stability": 1
  }
}

Getting the result

The response returns a taskId immediately without waiting for generation to finish.
In production, prefer waiting for callback.url to receive the terminal notification. For local debugging, poll GET /v1/tasks/:id.
When status=success, download the generated audio from output[].url.
When status=fail, fix the request based on the returned error instead of retrying the same invalid payload.

FAQ

How is it billed?

Billing is per character: the total character count across all text fields, multiplied by the per-1000-characters rate. See the live pricing page for the current rate. View live pricing

How much text can one call synthesize?

Up to 5000 characters total across all lines per call. Requests over the limit return 400 without charging. Split long scripts into multiple calls, ideally by paragraph.

How do I pick voices?

Pass a preset voice_id in each line (67 presets covering narration, characters, male and female voices). The full list is in the Preset voices table on this page; keep one voice_id per character. See the preset voices table

Which languages are supported?

70+ languages. The language is auto-detected by default; pass language_code (e.g. en / zh / ja) to pin it. Keep one language per conversation for best results.

What is the output format?

The result is MP3 audio (audio/mpeg). After status=success, download it from output[].url; result URLs expire, so copy them to your own storage promptly.