All articles

May 2026: a new OCR engine, merging, and accounting exports

ScanToExcel

The May release rebuilt the extraction engine and introduced flat and non-flat layouts, document merging, and accounting CSV exports. The converter was also updated to make those options easier to find.

Editorial update . Edited for clarity while preserving the announcement’s historical context.

A rebuilt extraction engine

The new pipeline focused on extracting structured data from receipts, invoices, and multi-page statements, including difficult scans and page boundaries. The release linked to the public performance page for its evaluation results.

Choosing the output layout

Flat output places line items in a table with document information repeated where needed. Non-flat output separates document details, line items, and totals. The choice supports different ways of reviewing and working with the extracted data; it does not reproduce every detail of the source page.

Merging documents

The release added a way to combine receipts, invoices, bank statements, or credit-card statements into one spreadsheet. Select flat output and merging, then use the same document type and a compatible output format throughout the batch. This brings related rows together for sorting and analysis.

Xero and QuickBooks exports

Xero CSV exports covered receipts, invoices, bank statements, and credit-card statements. QuickBooks CSV exports covered bank and credit-card statements. These exports target accounting import workflows; the destination’s required accounts, tax mappings, and file checks still matter.

A clearer converter

The updated interface brought output layouts, accounting CSVs, merging, ZIP downloads, and email delivery into the main workflow. The aim was to make the new options easier to choose while preparing a conversion.

May 2026: a new OCR engine, merging, and accounting exports | ScanToExcel