openapi: 3.1.0
info:
  title: Tensorlake Datasets Parse API
  description: 'Tensorlake Cloud APIs for Document Ingestion, Serverless Workflows, and Sandboxes. The Document Ingestion API parses documents (PDF, images, office formats) into layout-aware Markdown and structured chunks, performs schema-guided structured extraction and classification, and manages reusable files and datasets. Parsing runs as an asynchronous job: a parse request returns a parse id (job id) that is polled for results or delivered via webhook. All requests are authenticated with an API key passed as a Bearer token (Authorization: Bearer tl_apiKey_...).

    GROUNDING NOTE (API Evangelist, 2026-07-12): The path list, HTTP methods, base URL, and Bearer auth are grounded in Tensorlake''s published OpenAPI document (docs.tensorlake.ai/api-reference/openapi.yaml) and API reference. Request and response BODY SCHEMAS below are MODELED from Tensorlake documentation and SDK behavior and are illustrative, not byte-exact - verify field names against the live specification before generating client code.'
  version: 0.1.0
  contact:
    name: Tensorlake
    url: https://www.tensorlake.ai
servers:
- url: https://api.tensorlake.ai
  description: Tensorlake Cloud
security:
- bearerAuth: []
tags:
- name: Parse
  description: Asynchronous document parsing jobs.
paths:
  /documents/v2/parse:
    post:
      operationId: post_parse
      tags:
      - Parse
      summary: Submit a document parse job
      description: Submit a file (by file id), a public URL, or raw text/HTML for parsing. Returns a parse id (job id) immediately; the job runs asynchronously. Options include chunking strategy, table parsing/output mode, OCR, and an optional structured extraction schema. MODELED request body.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ParseRequest'
      responses:
        '200':
          description: The parse job was accepted.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ParseJobCreated'
        '401':
          $ref: '#/components/responses/Unauthorized'
        '422':
          $ref: '#/components/responses/ValidationError'
    get:
      operationId: list_parse
      tags:
      - Parse
      summary: List parse jobs
      description: Lists parse jobs in the current project, most recent first.
      parameters:
      - $ref: '#/components/parameters/Cursor'
      - $ref: '#/components/parameters/Limit'
      responses:
        '200':
          description: A page of parse jobs.
          content:
            application/json:
              schema:
                type: object
                properties:
                  items:
                    type: array
                    items:
                      $ref: '#/components/schemas/ParseJob'
                  next_cursor:
                    type: string
        '401':
          $ref: '#/components/responses/Unauthorized'
  /documents/v2/parse/{parse_id}:
    parameters:
    - $ref: '#/components/parameters/ParseId'
    get:
      operationId: get_parse
      tags:
      - Parse
      summary: Get a parse job and its result
      description: Retrieves a parse job by id. While the job is pending the status reflects progress; once completed the response carries the parsed Markdown, chunks, page fragments, and any structured extraction output.
      responses:
        '200':
          description: The parse job.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ParseJob'
        '401':
          $ref: '#/components/responses/Unauthorized'
        '404':
          $ref: '#/components/responses/NotFound'
    delete:
      operationId: delete_parse
      tags:
      - Parse
      summary: Delete a parse job
      description: Deletes a parse job and its stored result.
      responses:
        '200':
          description: Deletion confirmation.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/DeleteResponse'
        '401':
          $ref: '#/components/responses/Unauthorized'
        '404':
          $ref: '#/components/responses/NotFound'
components:
  parameters:
    Limit:
      name: limit
      in: query
      required: false
      description: Maximum number of items to return.
      schema:
        type: integer
    ParseId:
      name: parse_id
      in: path
      required: true
      description: The id of the parse job.
      schema:
        type: string
    Cursor:
      name: cursor
      in: query
      required: false
      description: Pagination cursor returned by a previous list call.
      schema:
        type: string
  schemas:
    ParseJob:
      type: object
      description: A parse job and, once complete, its result. MODELED.
      properties:
        parse_id:
          type: string
        status:
          type: string
          enum:
          - pending
          - processing
          - successful
          - failure
        created_at:
          type: string
          format: date-time
        markdown:
          type: string
          description: Layout-aware Markdown rendering of the document.
        chunks:
          type: array
          items:
            type: object
            additionalProperties: true
        pages:
          type: array
          items:
            type: object
            additionalProperties: true
        structured_data:
          type: object
          additionalProperties: true
    ParseRequest:
      type: object
      description: Parse request. Supply exactly one document source (file_id, file_url, or raw_text). MODELED - field names are illustrative.
      properties:
        file_id:
          type: string
          description: Id of a previously uploaded file.
        file_url:
          type: string
          format: uri
          description: Public URL of a document to fetch and parse.
        raw_text:
          type: string
          description: Raw text or HTML to parse instead of a file.
        page_range:
          type: string
          description: Pages to parse, e.g. "1-5,8".
        parsing_options:
          $ref: '#/components/schemas/ParsingOptions'
        structured_extraction_options:
          type: array
          items:
            $ref: '#/components/schemas/StructuredExtractionOptions'
        labels:
          type: object
          additionalProperties:
            type: string
    DeleteResponse:
      type: object
      properties:
        id:
          type: string
        deleted:
          type: boolean
    StructuredExtractionOptions:
      type: object
      description: MODELED structured extraction configuration.
      properties:
        schema_name:
          type: string
        json_schema:
          type: object
          additionalProperties: true
          description: A JSON Schema object describing the fields to extract.
    ParseJobCreated:
      type: object
      description: Response returned when an async job is accepted. MODELED.
      properties:
        parse_id:
          type: string
          description: Identifier used to retrieve the job result.
        status:
          type: string
          enum:
          - pending
          - processing
          - successful
          - failure
    ParsingOptions:
      type: object
      description: MODELED parsing configuration.
      properties:
        chunking_strategy:
          type: string
          enum:
          - none
          - page
          - section
          - fragment
          description: How the parsed output is split into chunks.
        table_parsing_strategy:
          type: string
          description: Strategy used to detect and parse tables.
        table_output_mode:
          type: string
          enum:
          - markdown
          - html
          - json
          description: Serialization format for detected tables.
        signature_detection:
          type: boolean
        figure_detection:
          type: boolean
    Error:
      type: object
      properties:
        error:
          type: object
          properties:
            code:
              type: string
            message:
              type: string
            details:
              type: object
              additionalProperties: true
  responses:
    ValidationError:
      description: The request payload failed validation.
      content:
        application/json:
          schema:
            $ref: '#/components/schemas/Error'
    Unauthorized:
      description: Missing or invalid Bearer token.
      content:
        application/json:
          schema:
            $ref: '#/components/schemas/Error'
    NotFound:
      description: The requested resource was not found.
      content:
        application/json:
          schema:
            $ref: '#/components/schemas/Error'
  securitySchemes:
    bearerAuth:
      type: http
      scheme: bearer
      bearerFormat: apiKey
      description: 'Tensorlake API key created in the Tensorlake Cloud dashboard (cloud.tensorlake.ai). Keys are prefixed tl_apiKey_ and are passed as Authorization: Bearer <token>.'