Pull specific data from documents into structured JSON.

POST JSON body with input and instructions.schema. Extract runs Parse internally, then returns values matching the requested schema.

{
  "body": {
    "input": "string (required: file_id, jobid:// reference, or public URL)",
    "instructions": {
      "schema": "object (required JSON schema of fields to extract)",
      "system_prompt": "string (optional context about document structure)"
    },
    "settings": {
      "array_extract": "boolean (optional for long repeating arrays)",
      "citations": {
        "enabled": "boolean (optional return source locations)"
      }
    }
  }
}

Parse a document by URL and extract structured content. Supports PDFs, images, spreadsheets, and other document formats.

POST JSON body with document_url pointing to a publicly accessible document. Returns extracted text, tables, and structured content.

{
  "body": {
    "document_url": "string (publicly accessible URL to document)",
    "advanced_options": {
      "ocr_system": "string (optional: 'highres' or 'standard')",
      "table_output_format": "string (optional: 'markdown' or 'html')"
    }
  }
}