Actions4
Async Submit (Upload File)
AI-generatedSummary
Submit a document to Upstage's async document-digitization API for later retrieval.
Inputs
- binaryPropertyName (required) — Name of the input item's binary property containing the file to parse (default: 'data'). Must reference an existing binary property on the incoming item.
- model — Document parse model to use. 'document-parse' (recommended) or 'document-parse-nightly'. Defaults to 'document-parse'.
- ocr — OCR mode. 'auto' applies OCR only to image documents; 'force' always performs OCR inference before layout detection. Defaults to 'auto'.
- base64Categories — Optional list of categories to return as cropped base64 images. Selections: 'figure', 'table', 'equation', 'chart'. Defaults to empty (no base64 images).
- merge_multipage_tables — When true, merges tables that span multiple pages into single table objects. Defaults to false.
- outputFormats — Output formats for layout elements. Options: 'html', 'markdown', 'text'. Defaults to ['html']. Each layout element is formatted according to the selected formats.
- coordinates — When true, returns bounding box coordinates for each layout element. Defaults to true.
- chartRecognition — When true, charts within the document are recognized and converted to tables. Defaults to true.
- tableOutputFormat — Structuring format for tables in the output. Options: '' (HTML default), 'markdown', 'csv'. Defaults to empty (HTML). Only affects table rendering when output formats include HTML or markdown.
Output shape
{ request_id: string, submitted: boolean }
The node submits the document to Upstage's async endpoint and immediately returns a unique request_id along with submitted:true. This request_id can be used with the 'Async Get Result (By Request ID)' operation to poll for the full parsing result. No document content is returned directly in this operation.