What kinds of documents can it read?

Invoices, receipts, purchase orders, bank and card statements, IDs and passports, application forms, contracts — anything with labeled fields. The AI reads each page as both text and image, so scanned and photographed documents work too.

A clean set of fields grouped by section (so it is easy to read and copy), followed by a JSON object with the same data so you can paste it straight into a spreadsheet, database or another app.

Does it make up values?

No. It is instructed to use only what appears in the document and to leave missing fields blank rather than guess. Always double-check critical figures against the source.

What happens to my document?

It is sent over an encrypted connection only to extract the data, and is not stored on our servers afterwards. Unlike the free tools, this feature does process your file on the server — that is how the AI reads it.

How much does it cost?

Each extraction costs 1 credit for a typical document (longer files priced by page). Credits never expire, and a failed attempt is automatically refunded.

AI 1 credit per use · Get credits →

AI Document Data Extraction

Pull structured fields from invoices, receipts, IDs and forms into clean text and JSON — processed securely, not stored.

How it works

Select a PDF (up to 32 MB) — an invoice, receipt, ID, form or statement, text or scanned.
It is sent over a secure connection to the AI and is not stored on our servers afterwards.
Click Extract — it costs 1 credit and is only charged if it succeeds.
Read the structured fields, copy them, or grab the JSON block to feed another system.

Frequently asked questions

What kinds of documents can it read?: Invoices, receipts, purchase orders, bank and card statements, IDs and passports, application forms, contracts — anything with labeled fields. The AI reads each page as both text and image, so scanned and photographed documents work too.
What do I get back?: A clean set of fields grouped by section (so it is easy to read and copy), followed by a JSON object with the same data so you can paste it straight into a spreadsheet, database or another app.
Does it make up values?: No. It is instructed to use only what appears in the document and to leave missing fields blank rather than guess. Always double-check critical figures against the source.
What happens to my document?: It is sent over an encrypted connection only to extract the data, and is not stored on our servers afterwards. Unlike the free tools, this feature does process your file on the server — that is how the AI reads it.
How much does it cost?: Each extraction costs 1 credit for a typical document (longer files priced by page). Credits never expire, and a failed attempt is automatically refunded.