https://developer.voicemaker.in/voice/api

Authentication

The Voicemaker API requires authentication via API keys, which are sent in the HTTP Authorization request header as bearer tokens.

Your production API key should be included as follows:

Authorization: Bearer YOUR-API-KEY

Content Type

All requests should be JSON-encoded and versioned using the HTTP Content-Type request header.
Your Content-Type header should look like:

Content-Type: application/json;

Generate Audio

To generate audio, make a POST request to the main API endpoint. Here’s an example using cURL:

curl https://developer.voicemaker.in/voice/api \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "Engine": "neural",
    "VoiceId": "ai3-Jony",
    "LanguageCode": "en-US",
    "Text": "Welcome to Voicemaker API.",
    "OutputFormat": "mp3",
    "SampleRate": "48000",
    "Effect": "default",
    "MasterVolume": "0",
    "MasterSpeed": "0",
    "MasterPitch": "0"
  }'

By default, output files will be automatically removed from the server within one hour. However, you can configure the FileStore option to retain files for up to 240 hours (10 days).

Error Response

Status: 200 OK

{
    "success": true,
    "path": "https://developer.voicemaker.in/uploads/1605116292576-voicemaker.in-speech.mp3"
}

Status: 400 Bad Request

{
    "success": false,
    "message": "Error converting your text, please check tags or settings in text."
}

Attributes (*Required all)

Engine (*)
standard, neural

VoiceId (*)
ai2-Katie, ai2-Kathy, ai2-Isabella, ai2-Robert, ai2-Jerry
(Please see the list below for all available Voice IDs)

LanguageCode (*)
en-US, en-GB, en-AU (Please refer below to available Voices)
multi-lang - For Pro Voices.

AccentCode (Optional)
en-US, en-GB, en-AU, da-DK, de-DE, nl-NL, en-IN, en-GB-WLS, fr-FR, fr-CA, es-US (Please refer below to available Voices)
(The AccentCode feature is compatible with AI1, Pro1 (Multilingual) & Pro2 (Multilingual) voices. For Pro1 voices, you can use accent codes (which are the same as language codes) to make the voice speak in specific languages and accents. If you require support for additional languages, please contact us.)

Text (*)
The audio request text can be up to 10,000 characters per request. Requests exceeding 3,000 characters may take little longer to process and generate speech. If you need support for SSML tags, please let us know.

Welcome to the Air, The Future is Here.
Get The Complete Caller Experience.

Inline Tags which you can add in Text (Please test the voice before implementation.)

Inline Tags only supports on AI1, AI2, AI3 & AI4 voics only.

Hey, <emphasis level="strong">Everything is cool?</emphasis>
Hey, <emphasis level="moderate">How are you?</emphasis>
Hey, <emphasis level="reduced">How is going on ?</emphasis>

Testing 500 ms pause <break time="500ms"/> here.
Testing 2 seconds pause <break time="2s"/> here.

We are testing <prosody rate="slow"> Slow speed.</prosody> here.
We are testing <prosody rate="90%"> Slow speed.</prosody> here.
We are testing <prosody pitch="low"> Lower Pitch.</prosody> here.
We are testing <prosody pitch="-10%"> Lower Pitch.</prosody> here.
We are testing <prosody volume="soft"> Low Volume.</prosody> here.

प्रिय ग्राहक, हम <en-in>Voicemaker</en-in> से कॉल कर रहे हैं.
亲爱的朋友，我们<es-es>te invitamos a unirte a nuestra comunidad</es-es>，希望你能来.
Querido amigo, te estamos contactando desde <fr-fr>nuestra organización</fr-fr> para ofrecerte una promoción especial.
Sehr geehrter Kunde, wir haben ein spezielles Angebot von <it-it>nuestra empresa</it-it> für Sie.
Chers collègues, nous avons une réunion organisée par <de-de>notre équipe</de-de> demain.
Estimado cliente, tenemos noticias importantes de <zh-cn>nuestra empresa</zh-cn> para usted.
Olá, estamos entrando em contato da <en-us>nossa equipe</en-us> para uma atualização importante.

OutputFormat (*)
mp3, wav
(We recommend using mp3 format for real-time applications, as WAV files may take longer to convert the audio.)

ResponseType (Optional)
file, stream
(By default, the response type is set to "file" which provides a downloadable URL for the audio file. Alternatively, you can choose "stream" which returns the audio in an AudioBuffer format that can be saved to a buffer or a file. Using "stream" avoids an additional HTTP request to download the audio file via a URL.)

FileStore (Optional)
1 is the default
1 to 240
(You can set the duration for retaining your converted output audio file on our servers in hours, with a maximum limit of 240 hours (10 days). Note: this parameter is not supported when using the "stream" response type. )

CustomFileName (Optional)
yourfilename
(You can set a custom name to the Output audio file as specified. Note: this parameter is not supported when using the "stream" response type.

SampleRate (Optional)
48000, 44100, 24000, 22050, 16000, 8000
(We recommend always using a sample rate of 48000, as it provides the best audio output for all voices.)

Effect (Optional)
default
breathing, soft, whispered - Only supports standard Engine ai1 voices.
conversational - Only support on neural Engine en-US - ai1-Joanna, ai1-Matthew, ai3-Aria, ai3-Jenny, cmn-CN - ai3-cmn-CN-Xiaoxiao.
news - Only support on neural Engine en-GB - ai1-Amy, en-US - ai3-Jony, ai3-Aria, ai3-Jenny, es-US - ai1-es-US-Lupe
customersupport - Only support on neural Engine en-US - ai3-Aria, ai3-Jenny, cmn-CN - ai3-cmn-CN-Yunyang.
assistant - Only support on neural Engine en-US - ai3-Aria, ai3-Jenny.
happy - Only support on neural Engine en-US - ai3-Aria, cmn-CN - ai3-cmn-CN-Yunye, ai3-cmn-CN-Xariyah, ai3-cmn-CN-Xiomara, ai3-cmn-CN-Carissa, ai3-cmn-CN-Xylia, ai3-cmn-CN-Xander.
empathic - Only support on neural Engine en-US - ai3-Aria.
clam - Only support on neural Engine cmn-CN - ai3-cmn-CN-Yunye.
whispered, happy, sad, angry, excited, friendly, hopeful, shouting, terrified, unfriendly:
ai3-Aria, ai3-Jenny, ai3-Jony, ai3-en-US-Alexander, ai3-en-US-Madison, ai3-en-US-Jayden, ai3-en-US-Ashley, ai3-en-US-Joshua

Please check "Get List of All Voices" below to access all voices and it's voice effects.

MasterSettings (Optional)
advanced_v1 is the default
advanced_v1, advanced_v2

advanced_v1 - Directly connects to AI engines, with characters being counted on each conversion.
advanced_v2 - If you only adjust `MasterVolume`, `MasterSpeed`, or `MasterPitch` without changing the `VoiceId` or `Text`, characters will not be counted. However, this will add an additional 200ms delay to the initial request.

MasterVolume (Optional)
0 is the default
-20 to 20

MasterSpeed (Optional)
0 is the default
-100 to 100

MasterPitch (Optional)
0 is the default
-100 to 100
If using MasterSettings as advanced_v1 then `MasterPitch` is not supported for neural Engine AI1 voices only.

Stability (Optional, available only for ProPlus and Custom Pro voices.)
50 is the default
-0 to 100
0 - Voice will become more expressive with the varying output in each convert.
100 - Voice will become more stable and make it more consistent with every convert. For longer text, we suggest using this side.

Similarity (Optional, available only for ProPlus and Custom Pro voices.)
80 is the default
-0 to 100
0 - Voice will become slightly robotic.
100 - Voice will closely match the original voice talent's voice by matching emotional and tonal values.

ProEngine (Optional, available only for ProPlus and Custom Pro voices.)
highres is the default
turbo, highres
turbo - Turbo will charge 3x characters and is designed for real-time applications like chatbots and low-latency use cases, though it may slightly trade off accuracy and stability.
highres - HighRes will charge 6x characters and is designed for high-quality, accurate outputs, making it perfect for content creation.

Get List of All Voices

To get a list of all voices, make a POST request to the voice list API endpoint.

curl https://developer.voicemaker.in/voice/list \
  -H "Authorization: Bearer YOUR-API-KEY" \
  -H "Content-Type: application/json" \
  -d '{"language": "en-US" }' \
  > '{ "success": true, "data": { "voices_list": [ { "Engine": "neural", "VoiceId": "ai1-Joanna", "VoiceGender": "Female", "VoiceWebname": "Joanna", "Country": "US", "Language": "en-US", "LanguageName": "English, US" }, { "Engine": "neural", "VoiceId": "ai1-Kendra", "VoiceGender": "Female", "VoiceWebname": "Kendra", "Country": "US", "Language": "en-US", "LanguageName": "English, US" } ] } }'

For retrieving the voice list via the API, we do not collect or count the requests.

Available Voices on API

Pro Voices

Pro Voices are billed at a higher character rate due to their advanced capabilities:

• ProPlus (Turbo model): Counts as 3x per character
• ProPlus (High-Res model): Counts as 6x per character
• Pro2: Counts as 3x per character
• Pro1: Counts as the standard of 1x per character. Only with Pro1, any Chinese, Japanese, or Korean character will be counted as 2x per character.

Custom Pro Voices

Default Voices

Default Voices (AI1, AI2, AI3, AI4, AI5, AI6, AI13) are counts at the standard of 1x per character. With default voices, any Chinese, Japanese, or Korean character will be counted as 2x per character.