API - Corsair

Every minimax.api.* operation is listed below with parameter shapes and return types from the plugin Zod schemas.

New to Corsair? See API access, authentication, and error handling.

Chat

completion

chat.completion Create a chat completion with a MiniMax text model Risk: read

await corsair.minimax.api.chat.completion({});

Input

Name	Type	Required	Description
`model`	`MiniMax-M3.1-Flash-Preview \| MiniMax-M3 \| MiniMax-M2.7 \| MiniMax-M2.7-highspeed \| MiniMax-M2.5 \| MiniMax-M2.5-highspeed \| MiniMax-M2.1 \| MiniMax-M2`	Yes	—
`messages`	`object[]`	Yes	—
`max_completion_tokens`	`number`	No	—
`max_tokens`	`number`	No	—
`temperature`	`number`	No	—
`top_p`	`number`	No	—
`tool_choice`	`none \| auto`	No	—
`tools`	`object[]`	No	—
`response_format`	`object`	No	—
`mask_sensitive_info`	`boolean`	No	—
`stream`	`false`	No	—

messages full type

{
  role: system | user | assistant | tool,
  name?: string,
  content?: string | {
    type?: string
  }[] | null,
  tool_calls?: {
    type?: string
  }[]
}[]

tools full type

{
  type: string,
  function: {
    name: string,
    description?: string,
    parameters?: {
    }
  }
}[]

response_format full type

{
  type: string
}

Output

Name	Type	Required	Description
`id`	`string`	No	—
`choices`	`object[]`	No	—
`created`	`number`	No	—
`model`	`string`	No	—
`object`	`chat.completion \| chat.completion.chunk`	No	—
`usage`	`object`	No	—
`base_resp`	`object`	No	—

choices full type

{
  finish_reason?: string,
  index?: number,
  message?: {
    role: system | user | assistant | tool,
    name?: string,
    content?: string | {
      type?: string
    }[] | null,
    tool_calls?: {
      type?: string
    }[]
  }
}[]

usage full type

{
  total_tokens?: number
}

base_resp full type

{
  status_code: number,
  status_msg?: string
}

createResponse

chat.createResponse Create a model response via the Responses API Risk: read

await corsair.minimax.api.chat.createResponse({});

Input

Name	Type	Required	Description
`model`	`string`	Yes	—
`input`	`object[]`	Yes	—
`instructions`	`string`	No	—
`max_output_tokens`	`number`	No	—
`temperature`	`number`	No	—
`top_p`	`number`	No	—
`service_tier`	`standard \| priority`	No	—
`tools`	`object[]`	No	—
`tool_choice`	`none \| auto`	No	—
`metadata`	`object`	No	—
`prompt_cache_key`	`string`	No	—
`text`	`object`	No	—
`reasoning`	`object`	No	—
`stream`	`false`	No	—

input full type

string | {
  type?: message | function_call | function_call_output | reasoning,
  role?: user | assistant | system | developer | tool,
  content?: string | {
    type?: string
  }[],
  call_id?: string,
  name?: string,
  arguments?: string,
  output?: string | {
    type?: string
  }[],
  summary?: {
  }[]
}[]

tools full type

{
  type: string
}[]

metadata full type

{
}

text full type

{
  format?: {
  }
}

reasoning full type

{
  effort?: string
}

Output

Name	Type	Required	Description
`id`	`string`	No	—
`object`	`string`	No	—
`created_at`	`number`	No	—
`model`	`string`	No	—
`status`	`completed \| incomplete \| failed`	No	—
`output`	`object[]`	No	—
`output_text`	`string`	No	—
`usage`	`object`	No	—
`error`	`object`	No	—
`incomplete_details`	`object`	No	—
`parallel_tool_calls`	`boolean`	No	—
`store`	`boolean`	No	—
`truncation`	`string`	No	—

output full type

{
  type?: string
}[]

usage full type

{
}

error full type

{
  code?: string,
  message?: string
}

incomplete_details full type

{
  reason?: string
}

estimateInputTokens

chat.estimateInputTokens Estimate the input token count of a request Risk: read

await corsair.minimax.api.chat.estimateInputTokens({});

Input

Name	Type	Required	Description
`model`	`string`	Yes	—
`input`	`object[]`	Yes	—
`instructions`	`string`	No	—
`tools`	`object[]`	No	—
`tool_choice`	`none \| auto`	No	—
`text`	`object`	No	—
`reasoning`	`object`	No	—
`stream`	`false`	No	—

input full type

string | {
  type?: message | function_call | function_call_output | reasoning,
  role?: user | assistant | system | developer | tool,
  content?: string | {
    type?: string
  }[],
  call_id?: string,
  name?: string,
  arguments?: string,
  output?: string | {
    type?: string
  }[],
  summary?: {
  }[]
}[]

tools full type

{
  type: string
}[]

text full type

{
  format?: {
  }
}

reasoning full type

{
  effort?: string
}

Output

Name	Type	Required	Description
`object`	`response.input_tokens`	Yes	—
`input_tokens`	`number`	Yes	—

Files

delete

files.delete Delete an uploaded file permanently [DESTRUCTIVE · IRREVERSIBLE] Risk: destructive · Irreversible

await corsair.minimax.api.files.delete({});

Input

Name	Type	Required	Description
`file_id`	`number`	Yes	—
`purpose`	`voice_clone \| prompt_audio \| t2a_async \| t2a_async_input \| video_generation`	Yes	—

Output

Name	Type	Required	Description
`file_id`	`number`	No	—
`base_resp`	`object`	No	—

base_resp full type

{
  status_code: number,
  status_msg?: string
}

list

files.list List uploaded files for a purpose Risk: read

await corsair.minimax.api.files.list({});

Input

Name	Type	Required	Description
`purpose`	`voice_clone \| prompt_audio \| t2a_async_input \| video_generation_input`	Yes	—

Output

Name	Type	Required	Description
`files`	`object[]`	Yes	—
`base_resp`	`object`	No	—

files full type

{
  file_id?: number,
  bytes?: number,
  created_at?: number,
  filename?: string,
  purpose?: string,
  download_url?: string
}[]

base_resp full type

{
  status_code: number,
  status_msg?: string
}

retrieve

files.retrieve Retrieve the metadata of an uploaded file Risk: read

await corsair.minimax.api.files.retrieve({});

Input

Name	Type	Required	Description
`file_id`	`number`	Yes	—

Output

Name	Type	Required	Description
`file`	`object`	Yes	—
`base_resp`	`object`	No	—

file full type

{
  file_id?: number,
  bytes?: number,
  created_at?: number,
  filename?: string,
  purpose?: string,
  download_url?: string
}

base_resp full type

{
  status_code: number,
  status_msg?: string
}

Image

generate

image.generate Generate images from a prompt, optionally with a subject reference Risk: write

await corsair.minimax.api.image.generate({});

Input

Name	Type	Required	Description
`model`	`image-01 \| image-01-live`	Yes	—
`prompt`	`string`	Yes	—
`subject_reference`	`object[]`	No	—
`aspect_ratio`	`1:1 \| 16:9 \| 4:3 \| 3:2 \| 2:3 \| 3:4 \| 9:16 \| 21:9`	No	—
`width`	`number`	No	—
`height`	`number`	No	—
`response_format`	`url \| base64`	No	—
`seed`	`number`	No	—
`n`	`number`	No	—
`prompt_optimizer`	`boolean`	No	—

subject_reference full type

{
  type: character,
  image_file: string
}[]

Output

Name	Type	Required	Description
`data`	`object`	No	—
`metadata`	`object`	No	—
`id`	`string`	No	—
`base_resp`	`object`	No	—

data full type

{
  image_urls?: string[],
  image_base64?: string[]
}

metadata full type

{
  success_count?: number,
  failed_count?: number
}

base_resp full type

{
  status_code: number,
  status_msg?: string
}

Models

list

models.list List available models (OpenAI-compatible) Risk: read

await corsair.minimax.api.models.list({});

Input: empty object Output

Name	Type	Required	Description
`object`	`string`	No	—
`data`	`object[]`	Yes	—

data full type

{
  id: string,
  object?: string,
  created?: number,
  owned_by?: string
}[]

listAnthropic

models.listAnthropic List available models (Anthropic-compatible) Risk: read

await corsair.minimax.api.models.listAnthropic({});

Input

Name	Type	Required	Description
`limit`	`number`	No	—
`after_id`	`string`	No	—
`before_id`	`string`	No	—

Output

Name	Type	Required	Description
`data`	`object[]`	Yes	—
`first_id`	`string`	No	—
`has_more`	`boolean`	No	—
`last_id`	`string`	No	—

data full type

{
  id: string,
  created_at?: string,
  display_name?: string,
  type?: string
}[]

retrieve

models.retrieve Retrieve a model (OpenAI-compatible) Risk: read

await corsair.minimax.api.models.retrieve({});

Input

Name	Type	Required	Description
`model_id`	`string`	Yes	—

Output

Name	Type	Required	Description
`id`	`string`	Yes	—
`object`	`string`	No	—
`created`	`number`	No	—
`owned_by`	`string`	No	—

retrieveAnthropic

models.retrieveAnthropic Retrieve a model (Anthropic-compatible) Risk: read

await corsair.minimax.api.models.retrieveAnthropic({});

Input

Name	Type	Required	Description
`model_id`	`string`	Yes	—

Output

Name	Type	Required	Description
`id`	`string`	Yes	—
`created_at`	`string`	No	—
`display_name`	`string`	No	—
`type`	`string`	No	—

Music

coverPreprocess

music.coverPreprocess Preprocess reference audio for two-step music cover generation Risk: write

await corsair.minimax.api.music.coverPreprocess({});

Input

Name	Type	Required	Description
`model`	`music-cover`	Yes	—
`audio_url`	`string`	No	—
`audio_base64`	`string`	No	—

Output

Name	Type	Required	Description
`cover_feature_id`	`string`	No	—
`formatted_lyrics`	`string`	No	—
`structure_result`	`string`	No	—
`audio_duration`	`number`	No	—
`trace_id`	`string`	No	—
`base_resp`	`object`	No	—

base_resp full type

{
  status_code: number,
  status_msg?: string
}

generate

music.generate Generate a song from a prompt and lyrics Risk: write

await corsair.minimax.api.music.generate({});

Input

Name	Type	Required	Description
`model`	`music-3.0 \| music-2.6 \| music-cover \| music-3.0-free \| music-2.6-free \| music-cover-free`	Yes	—
`prompt`	`string`	No	—
`lyrics`	`string`	No	—
`output_format`	`url \| hex`	No	—
`audio_setting`	`object`	No	—
`lyrics_optimizer`	`boolean`	No	—
`is_instrumental`	`boolean`	No	—
`audio_url`	`string`	No	—
`audio_base64`	`string`	No	—
`cover_feature_id`	`string`	No	—

audio_setting full type

{
  sample_rate?: number,
  bitrate?: number,
  format?: mp3 | wav | pcm
}

Output

Name	Type	Required	Description
`data`	`object`	No	—
`base_resp`	`object`	No	—

data full type

{
  status?: number,
  audio?: string
}

base_resp full type

{
  status_code: number,
  status_msg?: string
}

generateLyrics

music.generateLyrics Generate or edit song lyrics Risk: write

await corsair.minimax.api.music.generateLyrics({});

Input

Name	Type	Required	Description
`mode`	`write_full_song \| edit`	Yes	—
`prompt`	`string`	No	—
`lyrics`	`string`	No	—
`title`	`string`	No	—

Output

Name	Type	Required	Description
`song_title`	`string`	No	—
`style_tags`	`string`	No	—
`lyrics`	`string`	No	—
`base_resp`	`object`	No	—

base_resp full type

{
  status_code: number,
  status_msg?: string
}

Speech

createAsyncTask

speech.createAsyncTask Create an asynchronous long-text speech synthesis task Risk: write

await corsair.minimax.api.speech.createAsyncTask({});

Input

Name	Type	Required	Description
`model`	`speech-2.8-hd \| speech-2.8-turbo \| speech-2.6-hd \| speech-2.6-turbo \| speech-02-hd \| speech-02-turbo \| speech-01-hd \| speech-01-turbo`	Yes	—
`text`	`string`	No	—
`text_file_id`	`number`	No	—
`voice_setting`	`object`	Yes	—
`audio_setting`	`object`	No	—
`pronunciation_dict`	`object`	No	—
`language_boost`	`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`	No	—
`voice_modify`	`object`	No	—

voice_setting full type

{
  voice_id: string,
  speed?: number,
  vol?: number,
  pitch?: number,
  emotion?: happy | sad | angry | fearful | disgusted | surprised | calm | fluent | whisper,
  english_normalization?: boolean,
  latex_read?: boolean
}

audio_setting full type

{
  audio_sample_rate?: number,
  bitrate?: number,
  format?: mp3 | pcm | flac | wav | pcmu_raw | pcmu_wav | opus,
  channel?: number
}

pronunciation_dict full type

{
  tone?: string[]
}

voice_modify full type

{
  pitch?: number,
  intensity?: number,
  timbre?: number,
  sound_effects?: spacious_echo | auditorium_echo | lofi_telephone | robotic
}

Output

Name	Type	Required	Description
`task_id`	`string`	Yes	—
`file_id`	`number`	No	—
`task_token`	`string`	No	—
`usage_characters`	`number`	No	—
`base_resp`	`object`	No	—

base_resp full type

{
  status_code: number,
  status_msg?: string
}

queryAsyncTask

speech.queryAsyncTask Query the status of an asynchronous speech synthesis task Risk: read

await corsair.minimax.api.speech.queryAsyncTask({});

Input

Name	Type	Required	Description
`task_id`	`string \| number`	Yes	—

Output

Name	Type	Required	Description
`task_id`	`string \| number`	No	—
`status`	`success \| failed \| expired \| processing \| Success \| Failed \| Expired \| Processing`	Yes	—
`file_id`	`number`	No	—
`base_resp`	`object`	No	—

base_resp full type

{
  status_code: number,
  status_msg?: string
}

t2a

speech.t2a Synthesize speech from text (synchronous T2A) Risk: write

await corsair.minimax.api.speech.t2a({});

Input

Name	Type	Required	Description
`model`	`speech-2.8-hd \| speech-2.8-turbo \| speech-2.6-hd \| speech-2.6-turbo \| speech-02-hd \| speech-02-turbo \| speech-01-hd \| speech-01-turbo`	Yes	—
`text`	`string`	Yes	—
`voice_setting`	`object`	No	—
`audio_setting`	`object`	No	—
`pronunciation_dict`	`object`	No	—
`timbre_weights`	`object[]`	No	—
`language_boost`	`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`	No	—
`text_normalization_mode`	`basic \| quality`	No	—
`voice_modify`	`object`	No	—
`subtitle_enable`	`boolean`	No	—
`subtitle_type`	`sentence \| word \| word_streaming`	No	—
`output_format`	`url \| hex`	No	—

voice_setting full type

{
  voice_id: string,
  speed?: number,
  vol?: number,
  pitch?: number,
  emotion?: happy | sad | angry | fearful | disgusted | surprised | calm | fluent | whisper,
  text_normalization?: boolean,
  latex_read?: boolean
}

audio_setting full type

{
  sample_rate?: number,
  bitrate?: number,
  format?: mp3 | pcm | flac | wav | pcmu_raw | pcmu_wav | opus,
  channel?: number,
  force_cbr?: boolean
}

pronunciation_dict full type

{
  tone?: string[]
}

timbre_weights full type

{
  voice_id: string,
  weight: number
}[]

voice_modify full type

{
  pitch?: number,
  intensity?: number,
  timbre?: number,
  sound_effects?: spacious_echo | auditorium_echo | lofi_telephone | robotic
}

Output

Name	Type	Required	Description
`data`	`object`	No	—
`trace_id`	`string`	No	—
`extra_info`	`object`	No	—
`base_resp`	`object`	No	—

data full type

{
  audio?: string,
  status?: number,
  subtitle_file?: string,
  subtitle?: string,
  subtitles?: {
  }[]
}

extra_info full type

{
  audio_length?: number,
  audio_sample_rate?: number,
  audio_size?: number,
  bitrate?: number,
  audio_format?: string,
  audio_channel?: number,
  usage_characters?: number,
  word_count?: number
}

base_resp full type

{
  status_code: number,
  status_msg?: string
}

Video

cancelOrDeleteV2

video.cancelOrDeleteV2 Cancel a queued task or delete a finished task record permanently [DESTRUCTIVE · IRREVERSIBLE] Risk: destructive · Irreversible

await corsair.minimax.api.video.cancelOrDeleteV2({});

Input

Name	Type	Required	Description
`task_id`	`string`	Yes	—

Output

Name	Type	Required	Description
`task_id`	`string`	No	—
`action`	`cancelled \| deleted`	No	—
`status`	`cancelled \| deleted`	No	—

contextIr

video.contextIr Create an H3-Context-IR task that enriches a video prompt Risk: write

await corsair.minimax.api.video.contextIr({});

Input

Name	Type	Required	Description
`model`	`MiniMax-H3`	Yes	—
`content`	`object[]`	Yes	—
`duration`	`number`	Yes	—
`ratio`	`adaptive \| 21:9 \| 16:9 \| 4:3 \| 1:1 \| 3:4 \| 9:16`	No	—
`callback_url`	`string`	No	—

content full type

{
  type: text | image_url | video_url | audio_url,
  text?: string,
  image_url?: {
    url: string
  },
  video_url?: {
    url: string
  },
  audio_url?: {
    url: string
  },
  role?: first_frame | last_frame | reference_image | reference_video | reference_audio
}[]

Output

Name	Type	Required	Description
`task_id`	`string`	Yes	—

generate

video.generate Create a Hailuo video generation task (text, image, first/last frame or subject reference) Risk: write

await corsair.minimax.api.video.generate({});

Input

Name	Type	Required	Description
`model`	`MiniMax-Hailuo-2.3 \| MiniMax-Hailuo-2.3-Fast \| MiniMax-Hailuo-02 \| T2V-01-Director \| T2V-01 \| I2V-01-Director \| I2V-01-live \| I2V-01 \| S2V-01`	Yes	—
`prompt`	`string`	No	—
`first_frame_image`	`string`	No	—
`last_frame_image`	`string`	No	—
`subject_reference`	`object[]`	No	—
`prompt_optimizer`	`boolean`	No	—
`fast_pretreatment`	`boolean`	No	—
`duration`	`number`	No	—
`resolution`	`512P \| 720P \| 768P \| 1080P`	No	—
`callback_url`	`string`	No	—

subject_reference full type

{
  type: character,
  image: string[]
}[]

Output

Name	Type	Required	Description
`task_id`	`string`	Yes	—
`base_resp`	`object`	No	—

base_resp full type

{
  status_code: number,
  status_msg?: string
}

generateV2

video.generateV2 Create an H3 multimodal video generation task Risk: write

await corsair.minimax.api.video.generateV2({});

Input

Name	Type	Required	Description
`model`	`MiniMax-H3 \| MiniMax-H3-Max`	Yes	—
`content`	`object[]`	Yes	—
`resolution`	`480P \| 768P \| 2K`	Yes	—
`duration`	`number`	Yes	—
`ratio`	`adaptive \| 21:9 \| 16:9 \| 4:3 \| 1:1 \| 3:4 \| 9:16`	No	—
`extra`	`object`	No	—
`callback_url`	`string`	No	—

content full type

{
  type: text | image_url | video_url | audio_url,
  text?: string,
  image_url?: {
    url: string
  },
  video_url?: {
    url: string
  },
  audio_url?: {
    url: string
  },
  role?: first_frame | last_frame | reference_image | reference_video | reference_audio
}[]

extra full type

{
  prompt_expansion_mode?: string
}

Output

Name	Type	Required	Description
`task_id`	`string`	Yes	—

listV2

video.listV2 List H3 video tasks from the last 7 days Risk: read

await corsair.minimax.api.video.listV2({});

Input

Name	Type	Required	Description
`pageNum`	`number`	No	—
`pageSize`	`number`	No	—
`status`	`queued \| running \| succeeded \| failed \| cancelled`	No	—
`taskIds`	`string[]`	No	—
`model`	`string`	No	—
`taskType`	`generation \| h3_context_ir \| regeneration`	No	—

Output

Name	Type	Required	Description
`items`	`object[]`	Yes	—
`total`	`number`	No	—

items full type

{
  id?: string,
  model?: string,
  status?: queued | running | succeeded | failed | cancelled,
  error?: {
    code?: string,
    message?: string
  },
  created_at?: number,
  updated_at?: number,
  content?: {
    url?: string,
    prompt?: string
  },
  resolution?: string,
  duration?: number,
  usage?: {
  },
  ratio?: string,
  task_type?: generation | h3_context_ir | regeneration,
  modality?: video | text
}[]

query

video.query Query a Hailuo video generation task Risk: read

await corsair.minimax.api.video.query({});

Input

Name	Type	Required	Description
`task_id`	`string`	Yes	—

Output

Name	Type	Required	Description
`task_id`	`string`	No	—
`status`	`Preparing \| Queueing \| Processing \| Success \| Fail`	No	—
`file_id`	`string`	No	—
`video_width`	`number`	No	—
`video_height`	`number`	No	—
`base_resp`	`object`	No	—

base_resp full type

{
  status_code: number,
  status_msg?: string
}

queryV2

video.queryV2 Query an H3 video task by id Risk: read

await corsair.minimax.api.video.queryV2({});

Input

Name	Type	Required	Description
`task_id`	`string`	Yes	—

Output

Name	Type	Required	Description
`task`	`object`	Yes	—

task full type

{
  id?: string,
  model?: string,
  status?: queued | running | succeeded | failed | cancelled,
  error?: {
    code?: string,
    message?: string
  },
  created_at?: number,
  updated_at?: number,
  content?: {
    url?: string,
    prompt?: string
  },
  resolution?: string,
  duration?: number,
  usage?: {
  },
  ratio?: string,
  task_type?: generation | h3_context_ir | regeneration,
  modality?: video | text
}

regenerate

video.regenerate Create a 2K video regeneration task Risk: write

await corsair.minimax.api.video.regenerate({});

Input

Name	Type	Required	Description
`model`	`MiniMax-H3`	Yes	—
`source_task_id`	`string`	No	—
`content`	`object[]`	No	—
`resolution`	`2K`	Yes	—
`callback_url`	`string`	No	—
`aigc_watermark`	`boolean`	No	—

content full type

{
  type: text | image_url | video_url | audio_url,
  text?: string,
  image_url?: {
    url: string
  },
  video_url?: {
    url: string
  },
  audio_url?: {
    url: string
  },
  role?: base_video | first_frame | last_frame | reference_image | reference_video | reference_audio
}[]

Output

Name	Type	Required	Description
`task_id`	`string`	Yes	—

templateGenerate

video.templateGenerate Create a Video Agent task from a template Risk: write

await corsair.minimax.api.video.templateGenerate({});

Input

Name	Type	Required	Description
`template_id`	`string`	Yes	—
`text_inputs`	`object[]`	No	—
`media_inputs`	`object[]`	No	—
`callback_url`	`string`	No	—

text_inputs full type

{
  value: string
}[]

media_inputs full type

{
  value: string
}[]

Output

Name	Type	Required	Description
`task_id`	`string`	Yes	—
`base_resp`	`object`	No	—

base_resp full type

{
  status_code: number,
  status_msg?: string
}

templateQuery

video.templateQuery Query a Video Agent template generation task Risk: read

await corsair.minimax.api.video.templateQuery({});

Input

Name	Type	Required	Description
`task_id`	`string`	Yes	—

Output

Name	Type	Required	Description
`task_id`	`string`	No	—
`status`	`Preparing \| Queueing \| Processing \| Success \| Fail`	No	—
`video_url`	`string`	No	—
`base_resp`	`object`	No	—

base_resp full type

{
  status_code: number,
  status_msg?: string
}

Voice

clone

voice.clone Clone a voice from uploaded audio Risk: write

await corsair.minimax.api.voice.clone({});

Input

Name	Type	Required	Description
`file_id`	`number`	Yes	—
`voice_id`	`string`	Yes	—
`clone_prompt`	`object`	No	—
`text`	`string`	No	—
`model`	`speech-2.8-hd \| speech-2.8-turbo \| speech-2.6-hd \| speech-2.6-turbo \| speech-02-hd \| speech-02-turbo \| speech-01-hd \| speech-01-turbo`	No	—
`language_boost`	`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`	No	—
`text_validation`	`string`	No	—
`accuracy`	`number`	No	—
`need_noise_reduction`	`boolean`	No	—
`need_volume_normalization`	`boolean`	No	—
`aigc_watermark`	`boolean`	No	—

clone_prompt full type

{
  prompt_audio: number,
  prompt_text: string
}

Output

Name	Type	Required	Description
`demo_audio`	`string`	No	—
`input_sensitive`	`object`	No	—
`extra_info`	`object`	No	—
`base_resp`	`object`	No	—

input_sensitive full type

{
  type?: number
}

extra_info full type

{
}

base_resp full type

{
  status_code: number,
  status_msg?: string
}

delete

voice.delete Delete a generated voice permanently [DESTRUCTIVE · IRREVERSIBLE] Risk: destructive · Irreversible

await corsair.minimax.api.voice.delete({});

Input

Name	Type	Required	Description
`voice_type`	`voice_cloning \| voice_generation`	Yes	—
`voice_id`	`string`	Yes	—

Output

Name	Type	Required	Description
`voice_id`	`string`	No	—
`created_time`	`string`	No	—
`base_resp`	`object`	No	—

base_resp full type

{
  status_code: number,
  status_msg?: string
}

design

voice.design Design a custom voice from a description Risk: write

await corsair.minimax.api.voice.design({});

Input

Name	Type	Required	Description
`prompt`	`string`	Yes	—
`preview_text`	`string`	Yes	—
`voice_id`	`string`	No	—

Output

Name	Type	Required	Description
`voice_id`	`string`	Yes	—
`trial_audio`	`string`	No	—
`base_resp`	`object`	No	—

base_resp full type

{
  status_code: number,
  status_msg?: string
}

list

voice.list List system, cloned and generated voices Risk: read

await corsair.minimax.api.voice.list({});

Input

Name	Type	Required	Description
`voice_type`	`system \| voice_cloning \| voice_generation \| all`	Yes	—

Output

Name	Type	Required	Description
`system_voice`	`object[]`	No	—
`voice_cloning`	`object[]`	No	—
`voice_generation`	`object[]`	No	—
`base_resp`	`object`	No	—

system_voice full type

{
  voice_id: string,
  voice_name?: string,
  description?: string[],
  created_time?: string
}[]

voice_cloning full type

{
  voice_id: string,
  voice_name?: string,
  description?: string[],
  created_time?: string
}[]

voice_generation full type

{
  voice_id: string,
  voice_name?: string,
  description?: string[],
  created_time?: string
}[]

base_resp full type

{
  status_code: number,
  status_msg?: string
}

​Chat

​completion

​createResponse

​estimateInputTokens

​Files

​delete

​list

​retrieve

​Image

​generate

​Models

​list

​listAnthropic

​retrieve

​retrieveAnthropic

​Music

​coverPreprocess

​generate

​generateLyrics

​Speech

​createAsyncTask

​queryAsyncTask

​t2a

​Video

​cancelOrDeleteV2

​contextIr

​generate

​generateV2

​listV2

​query

​queryV2

​regenerate

​templateGenerate

​templateQuery

​Voice

​clone

​delete

​design

​list

Chat

completion

createResponse

estimateInputTokens

Files

delete

list

retrieve

Image

generate

Models

list

listAnthropic

retrieve

retrieveAnthropic

Music

coverPreprocess

generate

generateLyrics

Speech

createAsyncTask

queryAsyncTask

t2a

Video

cancelOrDeleteV2

contextIr

generate

generateV2

listV2

query

queryV2

regenerate

templateGenerate

templateQuery

Voice

clone

delete

design

list