API

Every octen.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.

Answer

create

answer.create Decompose messages into sub-queries, search, and synthesize an answer Risk: read

await corsair.octen.api.answer.create({});

Input

Name	Type	Required	Description
`model`	openai/gpt-6.1-sol \| openai/gpt-6-astra \| openai/gpt-5.6-sol \| openai/gpt-5.6-terra \| openai/gpt-5.6-luna \| openai/gpt-5.4 \| anthropic/claude-fable-5.1 \| anthropic/claude-fable-5 \| anthropic/claude-opus-5.5 \| anthropic/claude-opus-5 \| anthropic/claude-opus-4.8 \| anthropic/claude-opus-4.6 \| anthropic/claude-sonnet-5.5 \| anthropic/claude-sonnet-5 \| anthropic/claude-sonnet-4.6 \| anthropic/claude-haiku-5.5 \| anthropic/claude-haiku-4.5 \| google/gemini-4-argon \| google/gemini-3.8-flash \| google/gemini-3.5-flash \| google/gemini-3.5-flash-lite \| google/gemini-3.1-pro-preview \| google/gemini-3.1-flash-lite \| google/gemini-3-flash-preview \| moonshotai/kimi-k3 \| moonshotai/kimi-k2.6 \| moonshotai/kimi-k2.5 \| minimax/minimax-m2.5 \| qwen/qwen3.6-plus \| deepseek/deepseek-v4-pro \| deepseek/deepseek-v4-flash-0731	No	—
`messages`	`object[]`	Yes	—
`mode`	`queries_only \| queries_and_search \| full`	No	—
`max_queries`	`number`	No	—
`web_search_options`	`object`	No	—

messages full type

{
  role: system | developer | user | assistant | tool
}[]

web_search_options full type

{
  count?: number,
  include_domains?: string[],
  exclude_domains?: string[],
  include_text?: string[],
  exclude_text?: string[],
  time_basis?: auto | published | crawled,
  time_range?: day | week | month | year | d | w | m | y,
  start_time?: string,
  end_time?: string,
  language?: ar | de | en | es | fr | hi | id | it | ja | ko | nl | pl | pt | ru | th | tr | vi | zh[],
  highlight?: {
    enable?: boolean,
    max_tokens?: number
  },
  full_content?: {
    enable?: boolean,
    max_tokens?: number
  },
  format?: markdown | text,
  safesearch?: off | strict,
  include_images?: boolean
}

Output

Name	Type	Required	Description
`request_id`	`string`	Yes	—
`object`	`chat.completion`	Yes	—
`created`	`number`	Yes	—
`model`	`string`	Yes	—
`choices`	`object[]`	Yes	—
`queries`	`string[]`	No	—
`search_results`	`object[]`	No	—
`meta`	`object`	No	—
`warning`	`string`	No	—

choices full type

{
  index?: number,
  finish_reason?: stop | length | tool_calls | content_filter | error,
  message?: {
    role?: assistant,
    content?: string | null
  }
}[]

search_results full type

{
  query?: string,
  results?: {
    title?: string,
    url?: string,
    highlight?: string,
    full_content?: string,
    authors?: string,
    time_published?: string,
    time_last_crawled?: string,
    favicon?: string,
    cover_image?: {
      url?: string,
      description?: string
    },
    images?: {
      url?: string,
      description?: string
    }[]
  }[],
  latency?: number
}[]

meta full type

{
  usage?: {
    num_search_queries: number,
    prompt_tokens: number,
    completion_tokens: number,
    total_tokens: number
  },
  latency?: number
}

Broad Search

post

broadSearch.post Expand one query into sub-queries and search each of them Risk: read

await corsair.octen.api.broadSearch.post({});

Input

Name	Type	Required	Description
`query`	`string`	Yes	—
`max_queries`	`number`	No	—
`search_options`	`object`	No	—

search_options full type

{
  count?: number,
  include_domains?: string[],
  exclude_domains?: string[],
  include_text?: string[],
  exclude_text?: string[],
  time_basis?: auto | published | crawled,
  time_range?: day | week | month | year | d | w | m | y,
  start_time?: string,
  end_time?: string,
  language?: ar | de | en | es | fr | hi | id | it | ja | ko | nl | pl | pt | ru | th | tr | vi | zh[],
  highlight?: {
    enable?: boolean,
    max_tokens?: number
  },
  full_content?: {
    enable?: boolean,
    max_tokens?: number
  },
  format?: markdown | text,
  safesearch?: off | strict,
  include_images?: boolean
}

Output

Name	Type	Required	Description
`query`	`string`	No	—
`queries`	`string[]`	No	—
`search_results`	`object[]`	No	—

search_results full type

{
  query?: string,
  results?: {
    title?: string,
    url?: string,
    highlight?: string,
    full_content?: string,
    authors?: string,
    time_published?: string,
    time_last_crawled?: string,
    favicon?: string,
    cover_image?: {
      url?: string,
      description?: string
    },
    images?: {
      url?: string,
      description?: string
    }[]
  }[],
  latency?: number
}[]

Business Search

post

businessSearch.post Search business pages and return company and people cards Risk: read

await corsair.octen.api.businessSearch.post({});

Input

Name	Type	Required	Description
`query`	`string`	Yes	—
`count`	`number`	No	—
`include_domains`	`string[]`	No	—
`exclude_domains`	`string[]`	No	—
`time_basis`	`auto \| published \| crawled`	No	—
`time_range`	`day \| week \| month \| year \| d \| w \| m \| y`	No	—
`start_time`	`string`	No	—
`end_time`	`string`	No	—
`language`	`ar \| de \| en \| es \| fr \| hi \| id \| it \| ja \| ko \| nl \| pl \| pt \| ru \| th \| tr \| vi \| zh[]`	No	—
`highlight`	`object`	No	—
`full_content`	`object`	No	—
`entities`	`object`	No	—

highlight full type

{
  enable?: boolean,
  max_tokens?: number
}

full_content full type

{
  enable?: boolean,
  max_tokens?: number
}

entities full type

{
  enable?: boolean,
  count?: number,
  max_activities?: number,
  max_news?: number
}

Output

Name	Type	Required	Description
`query`	`string`	No	—
`results`	`object[]`	No	—
`entities`	`object[]`	No	—

results full type

{
  title?: string,
  url?: string,
  highlight?: string,
  full_content?: string,
  authors?: string,
  time_published?: string,
  time_last_crawled?: string,
  favicon?: string,
  cover_image?: {
    url?: string,
    description?: string
  },
  images?: {
    url?: string,
    description?: string
  }[]
}[]

entities full type

{
  type?: company | person,
  name?: string
}[]

Embedding

create

embedding.create Create text embeddings with an Octen embedding model Risk: read

await corsair.octen.api.embedding.create({});

Input

Name	Type	Required	Description
`input`	`string[]`	Yes	—
`model`	`octen-embedding-0.6b \| octen-embedding-4b \| octen-embedding-8b`	No	—
`dimension`	`number`	No	—
`input_type`	`query \| document`	No	—

Output

Name	Type	Required	Description
`results`	`object[]`	No	—
`model`	`string`	No	—

results full type

{
  index?: number,
  embedding?: number[]
}[]

Extract

post

extract.post Extract page content from up to 20 URLs Risk: read

await corsair.octen.api.extract.post({});

Input

Name	Type	Required	Description
`urls`	`string[]`	Yes	—
`mode`	`standard \| advanced \| auto`	No	—
`query`	`string`	No	—
`max_age_seconds`	`number`	No	—
`format`	`markdown \| text`	No	—
`timeout`	`number`	No	—
`include_images`	`boolean`	No	—
`include_videos`	`boolean`	No	—
`include_audio`	`boolean`	No	—
`include_links`	`object`	No	—

include_links full type

{
  scope?: prefer_internal | prefer_external,
  max_links?: number
}

Output

Name	Type	Required	Description
`results`	`object[]`	No	—

results full type

{
  url?: string,
  status?: success | failed,
  resolved_mode?: standard | advanced,
  title?: string | null,
  full_content?: string | null,
  highlights?: string[] | null,
  time_published?: string | null,
  time_last_crawled?: string,
  page_structure?: {
    primary?: Index Page | Content Page | No Main Content,
    secondary?: string | null
  } | null,
  category?: {
    primary?: string,
    secondary?: string | null
  } | null,
  favicon?: string | null,
  cover_image?: {
    url?: string
  } | null,
  images?: {
    url?: string
  }[],
  videos?: {
    url?: string
  }[],
  audio?: {
    url?: string
  }[],
  links?: {
    url?: string,
    anchor_text?: string,
    is_external?: boolean
  }[],
  error_message?: string
}[]

Grounded Generation

create

groundedGeneration.create Queue a search-grounded image or video generation task Risk: write

await corsair.octen.api.groundedGeneration.create({});

Input

Name	Type	Required	Description
`modality`	`image \| video`	Yes	—
`prompt`	`string`	Yes	—
`model`	`google/gemini-3-pro-image \| google/gemini-3.1-flash-image \| minimax/hailuo-3`	No	—
`aspect_ratio`	`16:9 \| 1:1 \| 9:16`	No	—
`duration`	`number`	No	—

Output

Name	Type	Required	Description
`request_id`	`string`	No	—
`generation_id`	`string`	Yes	—
`modality`	`image \| video`	Yes	—
`status`	`queued`	Yes	—
`created_at`	`number`	Yes	—

get

groundedGeneration.get Get a grounded generation task and its output URL Risk: read

await corsair.octen.api.groundedGeneration.get({});

Input

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

Output

Name	Type	Required	Description
`request_id`	`string`	No	—
`generation_id`	`string`	Yes	—
`modality`	`image \| video`	Yes	—
`status`	`queued \| running \| succeeded \| failed`	Yes	—
`output_url`	`string`	No	—
`references`	`object[]`	No	—
`search_results`	`object[]`	No	—
`error`	`object`	No	—
`created_at`	`number`	Yes	—
`updated_at`	`number`	Yes	—
`usage`	`object`	No	—

references full type

{
  source_id?: string,
  url?: string,
  title?: string,
  cover_url?: string
}[]

search_results full type

{
  query?: string,
  results?: {
    type?: image | video,
    title?: string,
    url?: string,
    source_page?: string,
    description?: string,
    cover_url?: string
  }[]
}[]

error full type

{
  code?: MM_GENERATION_CONTENT_POLICY | MM_GENERATION_INPUT_IMAGE_PRIVACY_BLOCKED | MM_GENERATION_CONTENT_NOT_SUPPORTED | MM_GENERATION_NO_REFERENCE_MATERIAL | MM_GENERATION_REJECTED | MM_GENERATION_TIMEOUT | MM_GENERATION_FAILED,
  message?: string
}

usage full type

{
  image_count?: number,
  video_count?: number,
  video_seconds?: number
}

Image Search

post

imageSearch.post Search images by text query or by a reference image Risk: read

await corsair.octen.api.imageSearch.post({});

Input

Name	Type	Required	Description
`inputs`	`object[]`	Yes	—
`count`	`number`	No	—
`include_domains`	`string[]`	No	—
`exclude_domains`	`string[]`	No	—
`safesearch`	`off \| strict`	No	—

inputs full type

{
  type: text | image,
  url?: string,
  data?: string
}[]

Output

Name	Type	Required	Description
`results`	`object[]`	No	—

results full type

{
  title?: string,
  url?: string,
  source_page?: string,
  description?: string,
  width?: number,
  height?: number,
  thumbnail?: string
}[]

News Search

post

newsSearch.post Search news articles, optionally grouped into events Risk: read

await corsair.octen.api.newsSearch.post({});

Input

Name	Type	Required	Description
`query`	`string`	Yes	—
`count`	`number`	No	—
`include_domains`	`string[]`	No	—
`exclude_domains`	`string[]`	No	—
`time_basis`	`auto \| published \| crawled`	No	—
`time_range`	`day \| week \| month \| year \| d \| w \| m \| y`	No	—
`start_time`	`string`	No	—
`end_time`	`string`	No	—
`language`	`ar \| de \| en \| es \| fr \| hi \| id \| it \| ja \| ko \| nl \| pl \| pt \| ru \| th \| tr \| vi \| zh[]`	No	—
`highlight`	`object`	No	—
`full_content`	`object`	No	—
`subjects`	`object`	No	—

highlight full type

{
  enable?: boolean,
  max_tokens?: number
}

full_content full type

{
  enable?: boolean,
  max_tokens?: number
}

subjects full type

{
  enable?: boolean,
  count?: number,
  max_sub_news?: number
}

Output

Name	Type	Required	Description
`query`	`string`	No	—
`results`	`object[]`	No	—
`subjects`	`object[]`	No	—

results full type

{
  title?: string,
  url?: string,
  summary?: string,
  highlight?: string,
  full_content?: string,
  authors?: string,
  time_published?: string,
  time_last_crawled?: string,
  favicon?: string,
  cover_image?: {
    url?: string,
    description?: string
  },
  images?: {
    url?: string,
    description?: string
  }[]
}[]

subjects full type

{
  name?: string,
  summary?: string,
  time_start?: string,
  time_latest?: string,
  top_news?: {
    title?: string,
    url?: string,
    summary?: string,
    highlight?: string,
    full_content?: string,
    authors?: string,
    time_published?: string,
    time_last_crawled?: string,
    favicon?: string,
    cover_image?: {
      url?: string,
      description?: string
    },
    images?: {
      url?: string,
      description?: string
    }[]
  },
  sub_news?: {
    title?: string,
    url?: string,
    summary?: string,
    highlight?: string,
    full_content?: string,
    authors?: string,
    time_published?: string,
    time_last_crawled?: string,
    favicon?: string,
    cover_image?: {
      url?: string,
      description?: string
    },
    images?: {
      url?: string,
      description?: string
    }[]
  }[]
}[]

Search

post

search.post Search the web with ranked, highlightable results Risk: read

await corsair.octen.api.search.post({});

Input

Name	Type	Required	Description
`count`	`number`	No	—
`include_domains`	`string[]`	No	—
`exclude_domains`	`string[]`	No	—
`include_text`	`string[]`	No	—
`exclude_text`	`string[]`	No	—
`time_basis`	`auto \| published \| crawled`	No	—
`time_range`	`day \| week \| month \| year \| d \| w \| m \| y`	No	—
`start_time`	`string`	No	—
`end_time`	`string`	No	—
`language`	`ar \| de \| en \| es \| fr \| hi \| id \| it \| ja \| ko \| nl \| pl \| pt \| ru \| th \| tr \| vi \| zh[]`	No	—
`highlight`	`object`	No	—
`full_content`	`object`	No	—
`format`	`markdown \| text`	No	—
`safesearch`	`off \| strict`	No	—
`include_images`	`boolean`	No	—
`query`	`string`	Yes	—

highlight full type

{
  enable?: boolean,
  max_tokens?: number
}

full_content full type

{
  enable?: boolean,
  max_tokens?: number
}

Output

Name	Type	Required	Description
`query`	`string`	No	—
`results`	`object[]`	No	—

results full type

{
  title?: string,
  url?: string,
  highlight?: string,
  full_content?: string,
  authors?: string,
  time_published?: string,
  time_last_crawled?: string,
  favicon?: string,
  cover_image?: {
    url?: string,
    description?: string
  },
  images?: {
    url?: string,
    description?: string
  }[]
}[]

Video Generation

create

videoGeneration.create Queue a text-to-video generation task Risk: write

await corsair.octen.api.videoGeneration.create({});

Input

Name	Type	Required	Description
`model`	`minimax/minimax-h3 \| bytedance/seedance-2.5 \| google/veo-3.1 \| google/veo-3.1-fast \| google/veo-3.1-lite`	Yes	—
`prompt`	`string`	Yes	—
`duration`	`number`	No	—
`resolution`	`480p \| 720p \| 768p \| 1080p \| 2K`	No	—
`aspect_ratio`	`16:9 \| 4:3 \| 1:1 \| 3:4 \| 9:16 \| 21:9 \| adaptive`	No	—
`frame_images`	`object[]`	No	—
`input_references`	`object[]`	No	—
`generate_audio`	`boolean`	No	—
`seed`	`number`	No	—
`extra_params`	`object`	No	—

frame_images full type

{
  image: string,
  frame_type: first_frame | last_frame
}[]

input_references full type

{
  image: string
}[]

extra_params full type

{
}

Output

Name	Type	Required	Description
`id`	`string`	Yes	—
`model`	`string`	Yes	—
`status`	`queued`	Yes	—
`created`	`number`	Yes	—

get

videoGeneration.get Get a video generation task and its download URL Risk: read

await corsair.octen.api.videoGeneration.get({});

Input

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

Output

Name	Type	Required	Description
`id`	`string`	Yes	—
`model`	`string`	Yes	—
`status`	`queued \| in_progress \| completed \| failed`	Yes	—
`created`	`number`	Yes	—
`video_url`	`string`	No	—
`expires_at`	`number`	No	—
`duration`	`number`	No	—
`resolution`	`string`	No	—
`usage`	`object`	No	—
`error`	`string`	No	—
`code`	`string`	No	—

usage full type

{
  video_seconds?: number,
  video_tokens?: number,
  input_image_count?: number
}

Video Search

post

videoSearch.post Search videos by a text query Risk: read

await corsair.octen.api.videoSearch.post({});

Input

Name	Type	Required	Description
`inputs`	`object[]`	Yes	—
`count`	`number`	No	—
`time_range`	`day \| week \| month \| year \| d \| w \| m \| y`	No	—
`start_time`	`string`	No	—
`end_time`	`string`	No	—
`safesearch`	`off \| strict`	No	—

inputs full type

{
  type: text,
  data?: string
}[]

Output

Name	Type	Required	Description
`results`	`object[]`	No	—

results full type

{
  title?: string,
  url?: string,
  source_page?: string,
  description?: string,
  cover_url?: string,
  duration_seconds?: number,
  authors?: string,
  time_published?: string,
  time_last_crawled?: string
}[]

Vl Embedding

create

vlEmbedding.create Create multimodal embeddings from text, images and video Risk: read

await corsair.octen.api.vlEmbedding.create({});

Input

Name	Type	Required	Description
`model`	`octen-vl-embedding \| octen-vl-embedding-large`	Yes	—
`input`	`object`	Yes	—
`enable_fusion`	`boolean`	No	—
`dimension`	`number`	No	—
`fps`	`number`	No	—
`instruct`	`string`	No	—

input full type

{
  contents: {
    text?: string,
    image?: string,
    video?: string
  }[]
}

Output

Name	Type	Required	Description
`results`	`object[]`	No	—
`model`	`string`	No	—

results full type

{
  index?: number,
  embedding?: number[],
  type?: vl | fusion
}[]

Overview

Plugins

Answer

create

Broad Search

post

Business Search

post

Embedding

create

Extract

post

Grounded Generation

create

get

Image Search

post

News Search

post

Search

post

Video Generation

create

get

Video Search

post

Vl Embedding

create

​Answer

​create

​Broad Search

​post

​Business Search

​post

​Embedding

​create

​Extract

​post

​Grounded Generation

​create

​get

​Image Search

​post

​News Search

​post

​Search

​post

​Video Generation

​create

​get

​Video Search

​post

​Vl Embedding

​create

Answer

create

Broad Search

post

Business Search

post

Embedding

create

Extract

post

Grounded Generation

create

get

Image Search

post

News Search

post

Search

post

Video Generation

create

get

Video Search

post

Vl Embedding

create