Skip to main content
POST
/
v3
/
minimax-speech-2.5-hd-preview
MiniMax Speech-2.5-hd-preview Synchronous Speech Synthesis
curl --request POST \
  --url https://api.highwayapi.ai/v3/minimax-speech-2.5-hd-preview \
  --header 'Authorization: <authorization>' \
  --header 'Content-Type: <content-type>' \
  --data '
{
  "text": "<string>",
  "voice_setting": {
    "speed": 123,
    "vol": 123,
    "pitch": 123,
    "voice_id": "<string>",
    "emotion": "<string>",
    "latex_read": true,
    "text_normalization": true
  },
  "audio_setting": {
    "sample_rate": 123,
    "bitrate": 123,
    "format": "<string>",
    "channel": 123
  },
  "pronunciation_dict": {
    "tone": [
      {}
    ]
  },
  "timbre_weights": [
    {
      "voice_id": "<string>",
      "weight": 123
    }
  ],
  "stream": true,
  "stream_options": {
    "exclude_aggregated_audio": true
  },
  "language_boost": "<string>",
  "output_format": "<string>",
  "voice_modify": {
    "pitch": 123,
    "intensity": 123,
    "timbre": 123,
    "sound_effects": "<string>"
  }
}
'
{
  "audio": "<string>",
  "status": 123
}
This API supports synchronous text-to-speech generation, with a maximum of 10,000 characters per text submission. It supports 100+ system voices and cloned voices for flexible selection; volume, pitch, speed, and output format adjustment; proportional voice mixing and fixed interval control; and multiple audio specifications and formats, including mp3, pcm, flac, and wav. Streaming output is also supported. After submitting a long-text speech synthesis request, note that the returned url is valid for 24 hours from the time the url is returned. Please download the information in time.
Suitable for scenarios such as short sentence generation, voice chat, and online social interaction. It has low latency, but the text length is limited to less than 10,000 characters. For long text, we recommend using asynchronous speech synthesis.

Request Headers

Content-Type
string
required
Enum value: application/json
Authorization
string
required
Bearer authentication format: Bearer {{API Key}}.

Request Body

text
string
required
The text to be synthesized. The length limit is less than 10,000 characters. Use line breaks to replace paragraph transitions. (If you need to control the interval time in the speech, add <#x#> between characters, where x is in seconds, supports 0.01-99.99, and allows up to two decimal places.) Custom speech intervals between texts are supported to achieve custom pause durations in synthesized speech. Note that the text interval must be set between two pieces of text that can be pronounced, and multiple consecutive intervals cannot be set.
voice_setting
object
required
audio_setting
object
pronunciation_dict
object
timbre_weights
object[]
Required if voice_id is not provided
stream
boolean
default:"false"
Whether to stream. Default is false, meaning streaming is not enabled.
stream_options
object
language_boost
string
default:"null"
Enhances recognition capability for specified low-resource languages and dialects. After setting this parameter, speech performance can be improved in the specified low-resource language/dialect scenario. If the low-resource language type is unclear, you can select “auto”, and the model will determine the low-resource language type on its own. The following values are supported:'Chinese', 'Chinese,Yue', 'English', 'Arabic', 'Russian', 'Spanish', 'French', 'Portuguese', 'German', 'Turkish', 'Dutch', 'Ukrainian', 'Vietnamese', 'Indonesian', 'Japanese', 'Italian', 'Korean', 'Thai', 'Polish', 'Romanian', 'Greek', 'Czech', 'Finnish', 'Hindi', 'Bulgarian', 'Danish', 'Hebrew', 'Malay', 'Persian', 'Slovak', 'Swedish', 'Croatian', 'Filipino', 'Hungarian', 'Norwegian', 'Slovenian', 'Catalan', 'Nynorsk', 'Tamil', 'Afrikaans', 'auto'
output_format
string
default:"hex"
Parameter that controls the output result format. Optional values are url and hex. The default value is hex. This parameter only takes effect in non-streaming scenarios; streaming scenarios only support returning data in hex format. The returned url is valid for 24 hours.
voice_modify
object
Voice effect settings. This parameter supports the following audio formats:
  • Non-streaming: mp3, wav, flac
  • Streaming: mp3

Response Information

audio
string
The synthesized audio clip, encoded in hex and generated according to the input-defined format (audio_setting.format) (mp3/pcm/flac). The return format depends on the definition of output_format. When stream is true, only the hex return format is supported.
status
number
Current audio stream status, returned only when stream is true. 1 indicates synthesis in progress, and 2 indicates synthesis has ended.