Extract to Excel

Upload invoices, quotes, or contracts and pull the key fields into a clean table — then download as a spreadsheet (CSV, opens in Excel & Google Sheets). The AI only reports what's actually in each document.

Drag & drop PDFs (or a folder) here, or click to choose

// Benefits

Why pull document data into a spreadsheet

Turn a stack of invoices, quotes, or contracts into one clean table — the AI reads the text of each document and lifts out the fields you need.

Batch in, one table out

Drop in a whole folder of documents and get a single spreadsheet — one row per file, the same columns lined up across every document.

Only what's actually there

The AI reports the values it can read in the text; fields it can't find are left blank rather than guessed, so a missing number stays empty instead of invented.

CSV that opens anywhere

Download as a CSV that opens straight in Excel and Google Sheets — ready to sort, filter, or paste into your own reconciliation.

// Workflow

How extraction fits your document work

For the moment the numbers and terms you need are trapped across many PDFs and you'd rather not retype them by hand.

  1. 1

    Add your invoices, quotes, or contracts and pick the document type.

  2. 2

    The document's text is analyzed by AI, which lifts the key fields into rows.

  3. 3

    Review the table and download the spreadsheet as CSV.

// Recommended reading

More ways to work through documents

Related AI tools for reading, comparing, and questioning your files.

Extract PDF data to a spreadsheet — FAQ

How do I extract data from PDFs into a spreadsheet?

Drop in your invoices, quotes, or contracts (or pick a whole folder to batch them), choose the document type, and click Extract. The AI pulls the key fields — totals, dates, parties, terms — into one table you can download as a CSV that opens in Excel, Google Sheets, or Numbers. It's free.

Are my files uploaded to a server?

The PDF itself never leaves your device — it's read right in your browser. Only the plain text it pulls out is sent to the AI to sort into columns; the original file, with its layout and any images, stays local. If that text-out step is a dealbreaker for sensitive contracts, that's worth knowing up front.

How do I know the numbers are right?

Every value is tagged with the exact sentence it came from in the original document, so you can spot-check it in one glance. If the AI can't clearly find a field, it leaves the cell blank instead of guessing — and we drop any source quote that doesn't actually appear in your file, so nothing is fabricated.

What are the limits?

Up to 8 documents at a time, and the combined text caps out around 60,000 characters — roughly a stack of normal invoices, not a 200-page master agreement. For big batches, run them in a few rounds.

It pulled nothing out — what happened?

Almost always a scanned or photographed PDF. If the text isn't selectable in a normal PDF reader, there's nothing for the browser to read and the AI gets an empty page. Run those through OCR first. Password-protected PDFs also can't be read until you unlock them.

Which documents work best?

Structured paperwork with consistent fields — invoices, quotes, and contracts — where each preset field (vendor, total, due date, payment terms, and so on) is actually printed somewhere in the doc. Free-form letters or unusual layouts will leave more cells blank.