OCR & extraction

Paper becomes data.

Scan or upload a document: Luneo reads it, classifies it and extracts the key fields — supplier, amount, due date — each paired with a confidence score. Zero re-keying.

99%
precision on key fields
12
languages recognised
0
manual entry
app.luneo.team/capture
Luneo OCR capture interface: invoice with extracted fields highlighted and confidence scores
Interactive

Hover over the document, read the extracted data

Move over a detected zone on the invoice: the matching field lights up on the right, with its normalised value and confidence score.

INVOICE
Supplier
GLOBEX LTD.
14 Kingsway, London WC2B 6LH · GB294471882
Invoice no.
INV-7782
Issue date
31 May 2026
Due date
30 Jun 2026
DescriptionAmount
Cloud hosting, enterprise€9,800.00
Premium support · annual€3,750.00
VAT (20 %)€1,000.00
Total due€14,550.00
Billed to: Northwind Group · Accounts Payable · 22 Rue de la Paix, 75002 Paris
Extracted data
Hover a zone or a field.
Supplier
Globex Ltd. 98 %
Invoice no.
INV-7782 99 %
Issue date
31/05/2026 96 %
Due date
30/06/2026 95 %
VAT number
GB294471882 94 %
Total amount
€14,550.00 97 %
The journey

From capture to search, in five steps

1Capture · scan or email2OCR · recognition3Classification4Field extraction5Search & export

Confidence score

Every field is scored; human review is only requested below a threshold.

Multilingual

Recognition across 12 languages, including mixed-language and handwritten documents.

Extraction models

Invoices, contracts, purchase orders, standard forms or your own templates. Data feeds the intelligent DMS.

Tables & line items

Extraction of tables and line items, not just headers.

Sovereign data

OCR runs on infrastructure hosted in France; nothing leaves the EU.

Export & webhook

Push extracted data to your ERP, your accounting system or via the 30+ native integrations.

Frequently asked questions

Everything you need to know about Luneo OCR

Which document formats does Luneo OCR support?

Luneo OCR supports PDF (native and scanned), JPEG, PNG and TIFF images, mobile captures and files received by email. Multi-page documents and hybrid PDFs (partially scanned) are also supported.

Does the OCR work with handwritten documents?

Luneo recognises printed characters with very high accuracy. For handwritten documents, recognition depends on legibility. Forms with checkboxes are fully supported and extracted automatically.

How does Luneo extract structured data from a document?

Luneo combines OCR with AI extraction models: after digitisation, it identifies key zones in the document (header, table, signature) and extracts the configured fields (amount, date, supplier reference…). Each extracted value comes with a confidence score.

Can you configure your own extraction fields?

Yes. Extraction models are configurable without code: you define the fields to extract (name, type, expected format) and Luneo applies them automatically to every new document of the same type, without any technical intervention.

Test extraction on your documents

Bring a few invoices or contracts: we show you the extracted data live. No commitment.