Chunkr Extract Task API

openapi: 3.0.1
info:
  title: Chunkr API
  description: >-
    The Chunkr Cloud API turns complex documents (PDF, Office, images) into
    RAG- and LLM-ready data. It exposes asynchronous parse and extract tasks
    that run layout analysis, OCR, segmentation, and chunking, plus file
    management and utility endpoints. Tasks are created and then polled until
    they reach a terminal status. Authentication uses an API key passed in the
    Authorization header.
  termsOfService: https://chunkr.ai/terms
  contact:
    name: Chunkr Support
    url: https://chunkr.ai
  license:
    name: AGPL-3.0 (open-source release) / Commercial (Cloud)
    url: https://github.com/lumina-ai-inc/chunkr/blob/main/LICENSE
  version: '1.0'
servers:
  - url: https://api.chunkr.ai
    description: Chunkr Cloud API
  - url: https://localhost:8000
    description: Self-hosted (Docker Compose) deployment
security:
  - apiKey: []
tags:
  - name: Tasks
    description: Create, poll, list, cancel, and delete parse and extract tasks.
  - name: Files
    description: Upload and manage files referenced by tasks.
  - name: Health
    description: Liveness and metadata utilities.
paths:
  /tasks/parse:
    post:
      operationId: createParseTask
      tags:
        - Tasks
      summary: Create a parse task
      description: >-
        Creates a parse task that runs layout analysis, OCR, segmentation, and
        chunking over the supplied document and returns the initial task record.
        Poll GET /tasks/{task_id} until status is Succeeded or Failed.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CreateParseTaskRequest'
      responses:
        '200':
          description: Task created.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskResponse'
        '401':
          description: Missing or invalid API key.
  /tasks/extract:
    post:
      operationId: createExtractTask
      tags:
        - Tasks
      summary: Create an extract task
      description: >-
        Creates an extract task that pulls schema-driven structured data from a
        document and returns JSON output with citations and metrics.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CreateExtractTaskRequest'
      responses:
        '200':
          description: Task created.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskResponse'
        '401':
          description: Missing or invalid API key.
  /tasks:
    get:
      operationId: listTasks
      tags:
        - Tasks
      summary: List tasks
      description: Returns a paginated list of tasks for the authenticated account.
      parameters:
        - name: page
          in: query
          schema:
            type: integer
            default: 1
        - name: limit
          in: query
          schema:
            type: integer
            default: 10
        - name: include_chunks
          in: query
          schema:
            type: boolean
            default: false
      responses:
        '200':
          description: A list of tasks.
          content:
            application/json:
              schema:
                type: array
                items:
                  $ref: '#/components/schemas/TaskResponse'
  /tasks/{task_id}:
    get:
      operationId: getTask
      tags:
        - Tasks
      summary: Get a task
      description: Retrieves a parse or extract task by id, including output when completed.
      parameters:
        - $ref: '#/components/parameters/TaskId'
        - name: base64_urls
          in: query
          description: Return base64-encoded URLs instead of presigned URLs.
          schema:
            type: boolean
            default: false
        - name: include_chunks
          in: query
          description: Include chunks in the response.
          schema:
            type: boolean
            default: true
      responses:
        '200':
          description: The task record.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskResponse'
        '404':
          description: Task not found.
    delete:
      operationId: deleteTask
      tags:
        - Tasks
      summary: Delete a task
      description: Deletes a task and its associated artifacts. Tasks must not be processing.
      parameters:
        - $ref: '#/components/parameters/TaskId'
      responses:
        '200':
          description: Task deleted.
        '404':
          description: Task not found.
  /tasks/{task_id}/parse:
    get:
      operationId: getParseTask
      tags:
        - Tasks
      summary: Get a parse task
      description: Retrieves the parse-specific output for a task.
      parameters:
        - $ref: '#/components/parameters/TaskId'
      responses:
        '200':
          description: The parse task record.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskResponse'
  /tasks/{task_id}/extract:
    get:
      operationId: getExtractTask
      tags:
        - Tasks
      summary: Get an extract task
      description: Retrieves the extract-specific output for a task.
      parameters:
        - $ref: '#/components/parameters/TaskId'
      responses:
        '200':
          description: The extract task record.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskResponse'
  /tasks/{task_id}/cancel:
    post:
      operationId: cancelTask
      tags:
        - Tasks
      summary: Cancel a task
      description: Cancels a task that is queued or processing.
      parameters:
        - $ref: '#/components/parameters/TaskId'
      responses:
        '200':
          description: Task cancelled.
        '404':
          description: Task not found.
  /files:
    post:
      operationId: uploadFile
      tags:
        - Files
      summary: Upload a file
      description: >-
        Uploads a file that can later be referenced from a parse or extract task
        via a ch://files/{file_id} reference.
      requestBody:
        required: true
        content:
          multipart/form-data:
            schema:
              type: object
              properties:
                file:
                  type: string
                  format: binary
      responses:
        '200':
          description: File uploaded.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/FileResponse'
    get:
      operationId: listFiles
      tags:
        - Files
      summary: List files
      description: Returns a paginated list of uploaded files.
      parameters:
        - name: page
          in: query
          schema:
            type: integer
            default: 1
        - name: limit
          in: query
          schema:
            type: integer
            default: 10
      responses:
        '200':
          description: A list of files.
          content:
            application/json:
              schema:
                type: array
                items:
                  $ref: '#/components/schemas/FileResponse'
  /files/{file_id}:
    get:
      operationId: getFile
      tags:
        - Files
      summary: Get a file
      parameters:
        - $ref: '#/components/parameters/FileId'
      responses:
        '200':
          description: The file record.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/FileResponse'
    delete:
      operationId: deleteFile
      tags:
        - Files
      summary: Delete a file
      parameters:
        - $ref: '#/components/parameters/FileId'
      responses:
        '200':
          description: File deleted.
  /files/{file_id}/download:
    get:
      operationId: downloadFileContent
      tags:
        - Files
      summary: Download file content
      parameters:
        - $ref: '#/components/parameters/FileId'
      responses:
        '200':
          description: The file content.
          content:
            application/octet-stream:
              schema:
                type: string
                format: binary
  /files/{file_id}/url:
    get:
      operationId: getFileUrl
      tags:
        - Files
      summary: Get a file URL
      description: Returns a presigned URL for the stored file.
      parameters:
        - $ref: '#/components/parameters/FileId'
      responses:
        '200':
          description: A presigned URL.
          content:
            application/json:
              schema:
                type: object
                properties:
                  url:
                    type: string
  /health:
    get:
      operationId: healthCheck
      tags:
        - Health
      summary: Health check
      description: Liveness check for the Chunkr API.
      security: []
      responses:
        '200':
          description: Service is healthy.
          content:
            text/plain:
              schema:
                type: string
                example: ok
  /extras/supported-file-types:
    get:
      operationId: getSupportedFileTypes
      tags:
        - Health
      summary: Get all supported file types
      description: Lists the file types accepted by the parsing and extraction pipelines.
      responses:
        '200':
          description: Supported file types.
          content:
            application/json:
              schema:
                type: array
                items:
                  type: string
components:
  securitySchemes:
    apiKey:
      type: apiKey
      in: header
      name: Authorization
      description: >-
        API key issued from the Chunkr dashboard, sent as the raw value of the
        Authorization header (e.g. "Authorization: lu_...").
  parameters:
    TaskId:
      name: task_id
      in: path
      required: true
      description: The unique identifier of the task.
      schema:
        type: string
    FileId:
      name: file_id
      in: path
      required: true
      description: The unique identifier of the file.
      schema:
        type: string
  schemas:
    CreateParseTaskRequest:
      type: object
      required:
        - file
      properties:
        file:
          type: string
          description: >-
            Document source - a ch://files/{file_id} reference, an http(s) URL,
            a data:*;base64,... URI, or a raw base64 string.
        file_name:
          type: string
          nullable: true
          description: Optional custom file name.
        expires_in:
          type: integer
          nullable: true
          description: Seconds until the task and its artifacts are deleted.
        chunk_processing:
          $ref: '#/components/schemas/ChunkProcessing'
        segment_processing:
          type: object
          nullable: true
          description: Per-segment-type configuration (Text, Table, Picture, etc.).
        ocr_strategy:
          type: string
          enum:
            - All
            - Auto
          default: All
          description: Whether to OCR all pages or only when needed.
        segmentation_strategy:
          type: string
          enum:
            - LayoutAnalysis
            - Page
          default: LayoutAnalysis
        error_handling:
          type: string
          enum:
            - Fail
            - Continue
          default: Fail
    ChunkProcessing:
      type: object
      properties:
        target_length:
          type: integer
          default: 4096
          description: Target number of words per chunk.
        tokenizer:
          type: string
          default: Word
          description: Tokenization method used when measuring chunk length.
    CreateExtractTaskRequest:
      type: object
      required:
        - file
        - schema
      properties:
        file:
          type: string
          description: Document source (file reference, URL, base64, or task id).
        schema:
          type: object
          description: The schema describing the structured data to extract.
        system_prompt:
          type: string
          nullable: true
        file_name:
          type: string
          nullable: true
        expires_in:
          type: integer
          nullable: true
        parse_configuration:
          type: object
          nullable: true
    TaskResponse:
      type: object
      required:
        - task_id
        - status
        - task_type
        - created_at
        - completed
      properties:
        task_id:
          type: string
        status:
          type: string
          enum:
            - Starting
            - Processing
            - Succeeded
            - Failed
            - Cancelled
        task_type:
          type: string
          enum:
            - Parse
            - Extract
        created_at:
          type: string
          format: date-time
        started_at:
          type: string
          format: date-time
          nullable: true
        finished_at:
          type: string
          format: date-time
          nullable: true
        expires_at:
          type: string
          format: date-time
          nullable: true
        completed:
          type: boolean
        message:
          type: string
        configuration:
          type: object
        file_info:
          type: object
          properties:
            name:
              type: string
            mime_type:
              type: string
            page_count:
              type: integer
            url:
              type: string
        output:
          type: object
          nullable: true
          description: ParseOutputResponse or ExtractOutputResponse with chunks, pages, and pdf_url.
        task_url:
          type: string
          nullable: true
        version_info:
          type: object
          properties:
            server_version:
              type: string
            client_version:
              type: string
    FileResponse:
      type: object
      properties:
        file_id:
          type: string
        file_name:
          type: string
        mime_type:
          type: string
        size:
          type: integer
        created_at:
          type: string
          format: date-time