Tensorlake Structured Extraction API

openapi: 3.1.0
info:
  title: Tensorlake Datasets Structured Extraction 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: Structured Extraction
  description: Schema-guided extraction, classification, read/OCR, and edit.
paths:
  /documents/v2/extract:
    post:
      operationId: post_extract
      tags:
      - Structured Extraction
      summary: Extract structured data from a document
      description: Extracts structured JSON from a document guided by a JSON schema. Accepts a file id (or URL) plus a json_schema object describing the fields to pull. Runs asynchronously and returns a job id. MODELED request body.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ExtractRequest'
      responses:
        '200':
          description: The extraction job was accepted.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ParseJobCreated'
        '401':
          $ref: '#/components/responses/Unauthorized'
        '422':
          $ref: '#/components/responses/ValidationError'
  /documents/v2/classify:
    post:
      operationId: post_classify
      tags:
      - Structured Extraction
      summary: Classify a document
      description: Classifies a document into one of a supplied set of categories (for example invoice, contract, receipt). MODELED request body.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ClassifyRequest'
      responses:
        '200':
          description: The classification job was accepted.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ParseJobCreated'
        '401':
          $ref: '#/components/responses/Unauthorized'
        '422':
          $ref: '#/components/responses/ValidationError'
  /documents/v2/read:
    post:
      operationId: post_read
      tags:
      - Structured Extraction
      summary: Read / OCR a document
      description: Runs OCR / read on a document to return raw text content per page. MODELED request body.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ReadRequest'
      responses:
        '200':
          description: The read job was accepted.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ParseJobCreated'
        '401':
          $ref: '#/components/responses/Unauthorized'
        '422':
          $ref: '#/components/responses/ValidationError'
  /documents/v2/edit:
    post:
      operationId: post_edit
      tags:
      - Structured Extraction
      summary: Edit a document
      description: Applies edits to a document (for example redaction or page operations). MODELED request body.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/EditRequest'
      responses:
        '200':
          description: The edit job was accepted.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ParseJobCreated'
        '401':
          $ref: '#/components/responses/Unauthorized'
        '422':
          $ref: '#/components/responses/ValidationError'
components:
  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'
  schemas:
    EditRequest:
      type: object
      description: MODELED edit request.
      properties:
        file_id:
          type: string
        operations:
          type: array
          items:
            type: object
            additionalProperties: true
    ReadRequest:
      type: object
      description: MODELED read / OCR request.
      properties:
        file_id:
          type: string
        file_url:
          type: string
          format: uri
        page_range:
          type: string
    ExtractRequest:
      type: object
      description: MODELED structured extraction request.
      properties:
        file_id:
          type: string
        file_url:
          type: string
          format: uri
        json_schema:
          type: object
          additionalProperties: true
        page_range:
          type: string
    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
    ClassifyRequest:
      type: object
      description: MODELED classification request.
      properties:
        file_id:
          type: string
        file_url:
          type: string
          format: uri
        classes:
          type: array
          description: Candidate categories to classify the document into.
          items:
            type: object
            properties:
              name:
                type: string
              description:
                type: string
    Error:
      type: object
      properties:
        error:
          type: object
          properties:
            code:
              type: string
            message:
              type: string
            details:
              type: object
              additionalProperties: true
  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>.'