API

Every aryn.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.

Async Tasks

list

asyncTasks.list List async partitioning tasks Risk: read

await corsair.aryn.api.asyncTasks.list({});

Input

Name	Type	Required	Description
`path_filter`	`string`	No	—

Output

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

tasks full type

{
}

Docset

create

docset.create Create a new DocSet Risk: write

await corsair.aryn.api.docset.create({});

Input

Name	Type	Required	Description
`name`	`string`	Yes	—
`schema`	`any`	No	—
`properties`	`object`	No	—
`prompts`	`object`	No	—

properties full type

{
}

prompts full type

{
}

Output

Name	Type	Required	Description
`docset_id`	`string`	Yes	—
`name`	`string`	Yes	—
`readonly`	`boolean`	Yes	—
`properties`	`object`	No	—
`schema`	`any`	No	—
`size`	`number`	No	—

properties full type

{
}

delete

docset.delete Delete a DocSet Risk: destructive

await corsair.aryn.api.docset.delete({});

Input

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

Output

Name	Type	Required	Description
`docset_id`	`string`	Yes	—
`name`	`string`	Yes	—
`readonly`	`boolean`	Yes	—
`properties`	`object`	No	—
`schema`	`any`	No	—
`size`	`number`	No	—

properties full type

{
}

get

docset.get Get DocSet metadata Risk: read

await corsair.aryn.api.docset.get({});

Input

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

Output

Name	Type	Required	Description
`docset_id`	`string`	Yes	—
`name`	`string`	Yes	—
`readonly`	`boolean`	Yes	—
`properties`	`object`	No	—
`schema`	`any`	No	—
`size`	`number`	No	—

properties full type

{
}

Document

get

document.get Get a document by ID Risk: read

await corsair.aryn.api.document.get({});

Input

Name	Type	Required	Description
`docset_id`	`string`	Yes	—
`doc_id`	`string`	Yes	—
`include_elements`	`boolean`	No	—
`include_binary`	`boolean`	No	—
`include_original_elements`	`boolean`	No	—

Output

Name	Type	Required	Description
`id`	`string`	Yes	—
`elements`	`object[]`	No	—
`properties`	`object`	No	—
`binary_data`	`string`	No	—

elements full type

{
  type?: string,
  bbox?: number[],
  properties?: {
    score?: number,
    page_number?: number
  } | null,
  text_representation?: string | null,
  binary_representation?: any
}[]

properties full type

{
}

getBinary

document.getBinary Download document binary content Risk: read

await corsair.aryn.api.document.getBinary({});

Input

Name	Type	Required	Description
`docset_id`	`string`	Yes	—
`doc_id`	`string`	Yes	—

Output

Name	Type	Required	Description
`docset_id`	`string`	Yes	—
`doc_id`	`string`	Yes	—
`contentBase64`	`string`	Yes	—

partition

document.partition Partition document using Aryn DocParse Risk: write

await corsair.aryn.api.document.partition({});

Input

Name	Type	Required	Description
`file`	`custom`	No	—
`file_url`	`string`	No	—
`options`	`object`	No	—

options full type

{
  threshold?: number | auto,
  text_mode?: string,
  table_mode?: string,
  extract_images?: boolean,
  extract_image_format?: string,
  selected_pages?: (
    number | number[]
  )[],
  strategy?: string,
  max_tokens?: number,
  tokenizer?: string,
  merge_across_pages?: boolean,
  promote_title?: boolean,
  title_candidate_elements?: string[],
  orientation_correction?: boolean,
  include_pagenum?: boolean,
  include_headers?: boolean,
  include_footers?: boolean,
  pipeline?: string,
  add_to_docset_id?: string
}

Output

Name	Type	Required	Description
`status`	`string[]`	Yes	—
`status_code`	`number`	No	—
`error`	`string`	No	—
`elements`	`object[]`	No	—
`markdown`	`string`	No	—

elements full type

{
  type?: string,
  bbox?: number[],
  properties?: {
    score?: number,
    page_number?: number
  } | null,
  text_representation?: string | null,
  binary_representation?: any
}[]

submitAsyncAdd

document.submitAsyncAdd Submit document asynchronously to a DocSet Risk: write

await corsair.aryn.api.document.submitAsyncAdd({});

Input

Name	Type	Required	Description
`docset_id`	`string`	Yes	—
`file`	`custom`	No	—
`file_url`	`string`	No	—
`options`	`object`	No	—

options full type

{
  threshold?: number | auto,
  text_mode?: string,
  table_mode?: string,
  extract_images?: boolean,
  extract_image_format?: string,
  selected_pages?: (
    number | number[]
  )[],
  strategy?: string,
  max_tokens?: number,
  tokenizer?: string,
  merge_across_pages?: boolean,
  promote_title?: boolean,
  title_candidate_elements?: string[],
  orientation_correction?: boolean,
  include_pagenum?: boolean,
  include_headers?: boolean,
  include_footers?: boolean,
  pipeline?: string,
  add_to_docset_id?: string
}

Output

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

Query

generatePlan

query.generatePlan Generate query plan Risk: read

await corsair.aryn.api.query.generatePlan({});

Input

Name	Type	Required	Description
`query`	`string`	Yes	—
`docset_id`	`string`	No	—
`summarize_result`	`boolean`	No	—
`stream`	`boolean`	No	—

Output

Name	Type	Required	Description
`query`	`string`	Yes	—
`nodes`	`object`	Yes	—
`result_node`	`number`	Yes	—
`llm_prompt`	`any`	No	—
`llm_plan`	`string`	No	—

nodes full type

{
}

Overview

Plugins

Async Tasks

list

Docset

create

delete

get

Document

get

getBinary

partition

submitAsyncAdd

Query

generatePlan

​Async Tasks

​list

​Docset

​create

​delete

​get

​Document

​get

​getBinary

​partition

​submitAsyncAdd

​Query

​generatePlan

Async Tasks

list

Docset

create

delete

get

Document

get

getBinary

partition

submitAsyncAdd

Query

generatePlan