---
canonical_url: "https://docs.composio.dev/toolkits/unstructured"
markdown_url: "https://docs.composio.dev/toolkits/unstructured.md"
product: "platform"
version: null
last_updated: null
---

# Unstructured

Unstructured Transform parses documents into Markdown or structured elements, extracts schema-defined fields, and manages reusable uploads and asynchronous jobs.

- **Category:** ai document extraction
- **Auth:** API_KEY
- **Composio-managed OAuth available?** N/A
- **Tools:** 9
- **Triggers:** 0
- **Slug:** `UNSTRUCTURED`
- **Version:** 00000000_00

## Tools

### Cancel Job

**Slug:** `UNSTRUCTURED_CANCEL_JOB`

Request irreversible cancellation of one queued or processing Parse or Extract job and return its immediate status snapshot; terminal jobs are unchanged.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `job_id` | string | Yes | Queued or processing Parse or Extract job ID whose cancellation should be requested. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### Delete Job

**Slug:** `UNSTRUCTURED_DELETE_JOB`

Request asynchronous removal of one terminal Parse or Extract job from job listings; source and output blobs may remain temporarily.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `job_id` | string | Yes | Terminal Parse or Extract job ID to remove from job listings. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### Delete Uploaded File

**Slug:** `UNSTRUCTURED_DELETE_UPLOADED_FILE`

Permanently delete one temporary uploaded file; related jobs remain and a queued job may fail if it has not consumed the file.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `file_id` | string | Yes | Temporary file ID returned by Upload File. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### Download Uploaded File

**Slug:** `UNSTRUCTURED_DOWNLOAD_UPLOADED_FILE`

Download the exact bytes of a previously uploaded scratch file before it expires.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `file_id` | string | Yes | Temporary file ID returned by Upload File. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### Extract Fields

**Slug:** `UNSTRUCTURED_EXTRACT_FIELDS`

Extract schema-defined fields from one uploaded document, stored file, or completed Parse job, either waiting for a result or returning a queued job.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `prompt` | string | No | Optional instructions that clarify how to populate json_schema fields from the document. |
| `file_id` | string | No | ID of a document already stored by Upload File; use exactly one of file_id, document, or parse_id. |
| `profile` | string ("balanced" | "best") | No | Processing profile for direct document or stored-file extraction. |
| `document` | object | No | Document to extract from. Attach one supported document up to 100 MB; use exactly one of document, file_id, or parse_id. |
| `filename` | string | No | Optional filename to associate with a stored file_id. |
| `parse_id` | string | No | ID of a completed Parse job to extract from; use exactly one of parse_id, document, or file_id. |
| `json_schema` | object | Yes | JSON Schema describing the fields and value types to extract. |
| `wait_seconds` | integer | No | Seconds to wait for completion (0-900). Use 0 to return a queued job immediately; use null to omit Prefer and allow the provider's wait of up to 15 minutes. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### Get Job

**Slug:** `UNSTRUCTURED_GET_JOB`

Get the current status, error, configuration, and available result for one Parse or Extract job.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `job_id` | string | Yes | Parse or Extract job ID returned by Parse Document, Extract Fields, or List Jobs. |
| `output` | string ("markdown" | "elements") | No | Representation to request when the job result is available. |
| `include` | array | No | Extra element metadata to include; valid only when output is elements. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### List Jobs

**Slug:** `UNSTRUCTURED_LIST_JOBS`

Return one newest-first page of Parse and Extract jobs, optionally filtered by status or operation, with a cursor for the next page.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `limit` | integer | No | Maximum jobs to return in this page (1-100). |
| `cursor` | string | No | Opaque next_cursor from a previous List Jobs response; omit for the first page. |
| `status` | string ("queued" | "processing" | "failed" | "completed" | "completed_with_warnings" | "cancelled") | No | Return jobs with this public status. completed_with_warnings maps to completed upstream and may also include completed jobs without warnings. |
| `operation` | string ("parse" | "extract") | No | Filter by the recorded operation. Parse requests with json_schema are listed as extract because extraction runs in the same job; a blocking Extract from parse_id may return a result without creating a listable job. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### Parse Document

**Slug:** `UNSTRUCTURED_PARSE_DOCUMENT`

Parse one uploaded or previously stored document into Markdown or structured elements; optionally extract schema-defined fields and either wait for a result or return a queued job.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `output` | string ("markdown" | "elements") | No | Result representation: Markdown text or structured document elements. |
| `prompt` | string | No | Optional extraction instructions used with json_schema, limited by the provider to 1,048,576 UTF-8 bytes. |
| `file_id` | string | No | ID of a document already stored by Upload File; use either file_id or document, not both. |
| `include` | array | No | Extra element metadata to include; valid only when output is elements. |
| `profile` | string ("balanced" | "best") | No | Processing profile balancing speed and quality. |
| `document` | object | No | Document to parse. Attach one BMP, DOCX, HEIC, JPEG, JPG, PDF, PNG, PPTX, or TIFF file up to 100 MB; use either document or file_id, not both. |
| `filename` | string | No | Optional filename to associate with a stored file_id. |
| `json_schema` | object | No | JSON Schema for fields to extract during parsing. The action serializes this object exactly once for the multipart schema field. |
| `wait_seconds` | integer | No | Seconds to wait for completion (0-900). Use 0 to return a queued job immediately; use null to omit Prefer and allow the provider's wait of up to 15 minutes. |
| `advanced_options` | object | No | Advanced parse options, currently including enrichments.image_description; serialized exactly once as a multipart JSON string. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |

### Upload File

**Slug:** `UNSTRUCTURED_UPLOAD_FILE`

Upload one supported document to 24-hour scratch storage for reuse by Parse Document or Extract Fields.

#### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `document` | object | Yes | Document to store. Attach one BMP, DOCX, HEIC, JPEG, JPG, PDF, PNG, PPTX, or TIFF file up to 100 MB. |

#### Output

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `data` | string | Yes | Data from the action execution |
| `error` | string | No | Error if any occurred during the execution of the action |
| `successful` | boolean | Yes | Whether or not the action execution was successful or not |
