Image to Excel

How to convert an image of a table to Excel

Choose or drop a photo or screenshot of a printed table — a price list, a timesheet, a statement or a report. The words are read in your browser and arranged into rows and columns. Check the preview, then download an Excel (.xlsx) or CSV file.

• Step 1 — choose the image (PNG, JPG or WebP, up to 20 MB).
• Step 2 — wait while the words are read and placed into rows and columns.
• Step 3 — check the table preview, then download Excel (.xlsx) or CSV.

How the columns are found

Rows come from words on the same line. Columns come from the vertical gaps between words: where no word crosses a gap on any line, a new column starts. This works well for tables with clear space between columns. It can merge two columns that sit very close together, or split a long cell that has wide spaces in it — check the preview, and fix anything in Excel after you download. Every cell is saved as text, so numbers keep exactly the characters that were read.

Private by design

The file never leaves your device. The reading runs inside your browser with an open-source OCR engine (Tesseract), so nothing is uploaded, nothing is stored and no AI service sees it. You can even disconnect from the internet once the page and the reader have loaded.

Getting a good result

Use a straight, sharp image where the table fills most of the frame. A screenshot of a table is usually better than a photo of a printed page. Crop away text above and below the table before you start; it would otherwise become extra rows.

The first time you use a browser tool, your browser downloads the reader — a few megabytes — from mahadocr.com and keeps it. Later runs start much faster. On an older phone a full page can take a little while; keep the page open until the result appears.

When you need more than a table

These free tools give you plain text or a simple table. If you need named fields — an invoice number, a passport’s date of birth, a CV’s work history — checked and marked ✓ confirmed or ⚠ check it, use the Mahad OCR API. It reads documents on our servers, returns JSON, and has a free Sandbox plan to try it.

Frequently asked questions