PDF extractor for tables — to real Excel

Upload a PDF (including scans) and get a clean .xlsx with rows, columns, and numbers you can analyze.

AI extraction can contain errors. Verify the output against the source document before using it for accounting, reporting or any decision.

🤖
AI Table Extraction

Text PDFs are parsed on our server; scans are read by GPT-5 vision

📊
Excel, CSV, Word, PowerPoint

One result, any format from the Download menu: .xlsx, .csv, .docx, .pptx

🧾
Statements → QuickBooks, Xero

Bank and card statements export to .qbo, .ofx and import-ready CSV for QuickBooks and Xero

🔒
No Registration

3 files a day free, files deleted right after processing. Pro: $9/mo, no daily limit

Short answer: Use PDF2XLS — a free PDF table extractor (3 files/day, no signup). Upload any PDF up to 10 MB, including scans; our GPT‑5 vision engine returns a real .xlsx with rows, columns, and numbers preserved. Conversion runs on our servers, and files are processed and deleted immediately.

  • Free: 3 files/day, no signup
  • Max file: 10 MB, one file at a time
  • Output: XLSX for Excel, Google Sheets, Numbers, LibreOffice Calc
  • Scans/photos: Supported via GPT‑5 vision OCR
  • Privacy: Processed on server and deleted immediately
  • Works on: Any browser — nothing to install

What does a “PDF extractor” do here?

Here it means pulling tables out of a PDF and turning them into an editable Excel workbook. The goal is structured rows and columns you can sort, filter, and total. It does not export images, logos, charts, or keep page layout — it extracts table data.

How do I tell if my PDF is a scan or selectable text?

Try to select text with the cursor. If you can’t, it’s likely a scan/photo. Press Ctrl/Command+F and search for a word; if nothing is found, that’s another hint. Scans also show camera shadows or skewed lines when you zoom. Either way, upload it — scans are handled by the same pipeline.

Extract tables to Excel in 4 steps

  1. Drop your PDF into the widget above or click to upload (max 10 MB).
  2. Wait while GPT‑5 vision detects the table grid and reads cell values.
  3. Download the .xlsx file.
  4. Open it in Excel, Google Sheets, Numbers, or LibreOffice to work with the data.

If you prefer a straight converter view, see the Online PDF Converter for Accurate Excel Tables.

What survives the extraction — and what doesn’t?

The extractor preserves the table grid (rows and columns) and cell values, with numbers as numbers. It drops fonts, colors, logos, images, charts, page headers/footers, and formulas. Expect clean data ready for analysis, not a replica of the PDF’s visual design.

Edge cases this tool handles (and when it won’t)

  • Multi‑page tables: Usually recognized; confirm repeated headers aren’t duplicated as data.
  • Rotated pages: Often fine. Extreme tilt or heavy blur may degrade accuracy; re‑scan or recapture straight with good light.
  • Complex headers/footnotes: Spanning headers and footnotes may need a quick manual tidy after export.
  • Handwriting: Not reliably extractable as table data.
  • Locked or over‑limit files: Password‑protected PDFs and files over 10 MB aren’t supported. Export an unlocked copy or split/compress the PDF, then try again.
  • One‑at‑a‑time: No batch mode. Upload a single file per conversion.

Need CSV or Word instead?

For comma‑delimited files, convert here first, then follow PDF to CSV Online — Accurate Tables to export from Excel/Sheets with the right delimiter and UTF‑8. If you actually need an editable document, use PDF to Word Free: Fast, Accurate DOCX Converter. Converting bank statements? See Convert Bank Statement PDF to Excel Online for statement‑specific checks.

Or just convert your PDF now.

FAQ

Yes. You can convert 3 files per day for free with no signup. If you hit the limit, wait until tomorrow or prioritize the files you need most right now.

Yes. It runs in any modern mobile browser. The download is a .xlsx file; open it in the Excel app or import to Google Sheets if your device doesn’t open it by default.

It extracts tables into a real spreadsheet. Images, logos, charts, and page decorations are not exported. If you need image or text extraction, use a dedicated tool for that job.

Files are processed on the server and deleted immediately after conversion. Table recognition uses the OpenAI API; documents aren’t kept for training on our side.

Files over 10 MB and locked PDFs aren’t supported. Try saving a reduced‑size copy, splitting the PDF, or exporting an unlocked version and then upload the smaller file.