Unstructured Partition API

Author: langgenius
Version: 0.0.1
Type: tool

Partition Input Parameters

The Unstructured Partition Endpoint provides parameters to customize the processing of documents.

The only required parameter is – the file you wish to process.

Main Parameters

Python / POST	JavaScript/TypeScript	Description
(shared.Files)	(File, Blob, shared.Files)	Required. The file to process.
(str)	(string)	Use one of the supported strategies to chunk the returned elements after partitioning. When not specified, no chunking is performed and any other chunking parameters are ignored. Supported: , , , .
(str)	(string)	A hint about the content type to use (such as ), when there are problems processing a specific file. This value is a MIME type in the format .
(bool)	(boolean)	to return bounding box coordinates for each element extracted with OCR. Default: .
(str)	(string)	The encoding method used to decode the text input. Default: .
(List[str])	(string[])	The types of elements to extract as Base64 encoded data in element metadata fields, e.g. . Supported filetypes: image and PDF.
(str)	(string)	If file is gzipped, use this content type after unzipping. Example:
(str)	(string)	The name of the inference model used when strategy is . Options: , . Default: .
(bool)	(boolean)	for the output to include page breaks if the filetype supports it. Default: .
(List[str])	(string[])	The languages present in the document, for use in partitioning and OCR.
(str)	(string)	The format of the response. Supported: , . Default: .
(bool)	(boolean)	Deprecated! Use instead. If and strategy is , any Table elements extracted from a PDF will include an additional metadata field, , with the HTML table.
(List[str])	(string[])	The document types to skip table extraction for. Default: .
(int)	(number)	The page number to assign to the first page in the document. This will be included in elements’ metadata.
(str)	(string)	The strategy to use for partitioning PDF and image files. Options: , , , , . Default: .
(bool)	(boolean)	to assign UUIDs to element IDs (guarantees uniqueness). Otherwise, a SHA-256 of the element’s text is used. Default: .
(str)	(Not yet available)	Applies only when strategy is . The name of the vision language model (VLM) provider to use for partitioning. must also be specified.
(str)	(Not yet available)	Applies only when strategy is . The name of the vision language model (VLM) to use for partitioning. must also be specified.
(bool)	(boolean)	to retain the XML tags in the output. Otherwise, only the text within the tags is extracted. Only applies to XML documents.

Chunking Parameters

The following parameters only apply when a chunking strategy is specified. Otherwise, they are ignored.

Python / POST	JavaScript/TypeScript	Description
(int)	(number)	Applies only when the chunking strategy is set to . Combines small chunks until the combined chunk reaches a length of n characters. Default: same as .
(bool)	(boolean)	(default) to have the elements used to form a chunk appear in for that chunk.
(int)	(number)	Cut off new sections after reaching a length of n characters. (Hard maximum.) Default: .
(bool)	(boolean)	Applies only when the chunking strategy is set to . Determines if a chunk can include elements from more than one page. Default: .
(int)	(number)	Applies only when the chunking strategy is specified. Cuts off new sections after reaching a length of n characters. (Soft maximum.) Default: .
(int)	(number)	A prefix of this many trailing characters from the prior text-split chunk is applied to second and later chunks formed from oversized elements by text-splitting. Default: none.
(bool)	(boolean)	to have an overlap also applied to “normal” chunks formed by combining whole elements. Use with caution, as this can introduce noise into otherwise clean semantic units.
(float)	(number)	Applies only when the chunking strategy is set to . The minimum similarity text in consecutive elements must have to be included in the same chunk. Range: 0.01–0.99. Default: .

Client-Specific Parameters (Not sent to server)

Python / POST	JavaScript/TypeScript	Description
(bool)	(boolean)	to split the PDF file client-side.
(bool)	(boolean)	When , a failed split request will not stop the processing of the rest of the document. The affected page range will be ignored in the results. Default: .
(int)	(number)	The number of split files to be sent concurrently. Default: . Maximum: .
(List[int])	(number[])	A list of 2 integers within the range . When PDF splitting is enabled, this will send only the specified page range to the API.

Notes

For more details on each parameter, refer to the official documentation.
Some parameters are only available in specific strategies or file types.
Default values are shown where applicable.

API Response Structure

Top-Level Fields

Field	Type	Description
	string	The full parsed content in Markdown format, including images, sections, etc.
	array	List of attached files (if any).
	array	Structured JSON data (if any).
	array of object	List of image objects extracted from the content.
	array of object	List of structured content blocks (sections, paragraphs, images, etc).

Field Details

text

Type: string
Description:
The entire content, formatted in Markdown. This may include images, headings, bullet points, and other formatting for direct rendering.

files

Type: array
Description:
List of attached files. Typically empty in the current output.

json

Type: array
Description:
Structured JSON data for advanced use cases. Typically empty in the current output.

images

Type: array of objects
Description:
List of images found in the content. Each image object contains the following fields:

Field	Type	Description
	string	File extension (e.g., )
	string	Unique image ID
	string	MIME type of the image
	string	Image file name
	string	URL for image preview
	integer	Image file size in bytes
	string	Always

Example:

elements

Type: array of objects
Description:
List of structured content blocks. Each object represents a section, paragraph, image, or other content element.

Field	Type	Description
	string	Unique identifier for the element
	object	Metadata for the element (see below for details)
	string	The text content of the element
	string	The type of element (e.g., , , , )

metadata (object)

The field provides detailed information about the element's origin, layout, and context.
Possible fields include (not all fields are present in every element):

Field	Type	Description
	string	Name of the source file (e.g., )
	string	MIME type or file type (e.g., , )
	array	List of detected languages (e.g., )
	integer	Page number in the source file (if applicable)
	object	Layout information (see below)
	float	Confidence score for element detection (0-1)
	string	MIME type for images (e.g., )
	string	Unique file ID for images
	string	Preview URL for images
	array	List of original sub-elements (for composite elements)

coordinates (object)

Describes the position and size of the element in the source file (if available):

Field	Type	Description
	integer	Height of the layout (e.g., page height in pixels)
	integer	Width of the layout (e.g., page width in pixels)
	array	List of four [x, y] coordinate pairs (bounding box)
	string	Coordinate system used (e.g., )

Example Response

Notes

The field is suitable for direct display in web or app frontends.
The field is useful for structured processing, highlighting, or further analysis.
The field provides all image resources for preview or download.
The and fields are reserved for future use or advanced scenarios.
The object in each element may contain additional fields depending on the extraction process and file type.