Skip to main content
These nodes read documents and spreadsheets, and turn your results back into files you can download.

Getting a file in

Every read node takes a file one of two ways. You can upload one directly on the node, or wire in a file from an earlier step, usually a File field on your User Input node so whoever runs the flow can pick their own.

Reading files

  • Read CSV and Read Excel pull a spreadsheet apart into columns. Type the column names you want and each one becomes its own output, so name them exactly as they appear in the file. Read Excel also asks which sheet to use.
  • Read PDF extracts the text from a PDF. It is quick and free, and works whenever the PDF has a real text layer.
  • AI PDF Reader reads the PDF visually instead, which is what you want for scans, tables, and charts. It is slower and uses credits, so reach for it when Read PDF gives you nothing useful.
  • Read Word and Read PowerPoint pull the content out of documents and presentations.
  • Excel to Markdown turns a whole workbook into text in one go, handy when the next step is an AI node rather than column-by-column wiring.

Exporting files

Export CSV, Export Excel, Export PDF, and Export Word turn your results into a file. The CSV and Excel versions take your columns, the PDF and Word versions take text. Wire the file into an Output node to get it after a run, where it appears in the Output pane as a link you can download.