The most important element in voice synthesis is the voice ID that determines which voice will read the text. This document guides you on how to find the desired voice and how to check voice information needed for API calls.

โ€‹
1. What is a Voice ID?

voice_id is a value that identifies one unique character.

When synthesizing speech, you need to pass language, style, and model information along with this voice ID to generate accurate speech.

โ€‹
2. Three Ways to Find Voice IDs

โ€‹
2.1 Copy from Supertone Play

The fastest and most intuitive way is to copy the voice_id from the voice library in Supertone Play.

  • Click the Copy voice ID button within each voice item
  • The voiceโ€™s ID will be copied to your clipboard
  • Cloned voices can also be checked in a separate tab

Go to Supertone Play

Preset Voice Example:

Cloned Voice Example:

โ€‹
2.2 Get Complete Voice List with GET /v1/voices

To check the complete list of available voices without conditions through the API, use the GET /voices endpoint.

GET /v1/voices

The response includes the following information:

  • voice_id: ID to be used in calls
  • name: Character name (e.g., Agatha, Adam, etc.)
  • language: List of supported languages
  • styles: List of selectable styles
  • models: Available voice models
  • samples: Sample audio by language/style/model
  • thumbnail_image_url: Voice image

Response example:

{
    "voice_id": "91992bbd4758bdcf9c9b01",
    "name": "Adam",
    "description": "",
    "age": "young-adult",
    "gender": "male",
    "use_case": "meme",
    "language": [
       "ko",
       "en",
       "ja"
       ],
     "styles": [
       "neutral"
       ],
     "models": [
       "sona_speech_1"
       ],
     "samples": [
       {
        "language": "ja",
        "style": "neutral",
        "model": "sona_speech_1",
        "url": "https://d2hxfz33hpdvze.cloudfront.net/voice-library/resources/20250502_044710.wAA5PFQe7BHNQjrfMiNrRd.speech.wav"
        },
       {
        "language": "ko",
        "style": "neutral",
        "model": "sona_speech_1",
        "url": "https://d2hxfz33hpdvze.cloudfront.net/voice-library/resources/20250502_044844.tehBuHoUhZcWsCfUsr15As.speech.wav"
        },
        {
         "language": "en",
         "style": "neutral",
         "model": "sona_speech_1",
         "url": "https://d2hxfz33hpdvze.cloudfront.net/voice-library/resources/20250201_132654.rdMFfzDsaub9qXABwti65M.speech.wav"
         }
         ],
         "thumbnail_image_url": "https://d2hxfz33hpdvze.cloudfront.net/voice-library/resources/7wmvg5vEBeooDEB9xnut6s.png"
}

โ€‹
2.3 Filter by Conditions with GET /v1/voices/search

If you want to find only voices that match specific language, style, or purpose, try the GET /voices/search endpoint.

GET /v1/voices/search?language=ko&style=happy

Supported Filters:

  • language: Language code (ko, en, ja, etc.)
  • style: Emotional style (neutral, happy, etc.)
  • name, description, gender, age, use_case can also be searched

Sorting functionality is not supported, and accurate condition specification is important.

โ€‹
๐Ÿ“Œ Looking for Cloned Voices?

Searching with name=My Voice will show all cloned voices created in your current account.

GET /v1/voices/search?name=My Voice

A response example for the above search is as follows:

{
    "items": [
        {
            "voice_id": "opSGuRvHBe7EfZ4LQga1hE",
            "name": "My Voice 1",
            "description": "",
            "age": "",
            "gender": "",
            "use_case": "",
            "language": [
                "ja"
            ],
            "styles": [
                "jealous"
            ],
            "models": [
                "sona_speech_1"
            ],
            "samples": [
                {
                    "language": "ja",
                    "style": "jealous",
                    "model": "sona_speech_1",
                    "url": "https://d2hxfz33hpdvze.cloudfront.net/voice-library/resources/20250514_160801.bDwEhmuq8JALRMP3nSHzUD.speech.wav"
                }
            ],
            "thumbnail_image_url": ""
        },
        {
            "voice_id": "hZGrJcUTNSNwfZDzasAVUh",
            "name": "My Voice 2",
            "description": "",
            "age": "",
            "gender": "",
            "use_case": "",
            "language": [
                "en"
            ],
            "styles": [
                "active"
            ],
            "models": [
                "sona_speech_1"
            ],
            "samples": [
                {
                    "language": "en",
                    "style": "active",
                    "model": "sona_speech_1",
                    "url": "https://d2hxfz33hpdvze.cloudfront.net/voice-library/resources/20250514_161008.ekky3mQgtgVLWLa54ispy5.speech.wav"
                }
            ],
            "thumbnail_image_url": ""
        }
    ],
    "total": 2,
    "next_page_token": null
}

โ€‹
3. Check Sample Audio and Style Information

The samples field in the voice object contains sample audio based on language, style, and model combinations.

{
  "language": "ko",
  "style": "neutral",
  "model": "sona_speech_1",
  "url": "https://.../speech.wav"
}

Accessing this URL allows you to download actual speech samples. Since different characters support different style/language/model combinations, please check before use.

โ€‹
4. Next Steps

Once youโ€™ve found a voice ID, refer to the Text-to-Speech Guide to continue with text input, parameter settings, and audio output format specification.