Image to Excel Converter

Turn a photo, screenshot or scanned PDF of a table into an editable spreadsheet. Check the cells next to the picture, fix anything the OCR got wrong, then download Excel or CSV — or paste straight into Excel or Google Sheets. It all runs in your browser, so the image never leaves your device.

Drop a picture of a table here

Photo, screenshot (PNG, JPG, WebP) or scanned PDF — or paste a screenshot with Ctrl+V. Each image or PDF page becomes its own table.

🔒 Processed on your device — nothing is uploaded.

Turn a Picture of a Table into a Real Spreadsheet

Retyping figures from a screenshot, a scanned statement or a phone photo of a printed report is slow and error-prone. This free image to Excel converter reads the table with OCR (optical character recognition), works out the rows and columns, and gives you an editable grid you can check against the picture before you export it as an .xlsx workbook or a CSV file — or paste it straight into Excel or Google Sheets.

Excel has a similar Data from Picture feature, but it needs a Microsoft 365 subscription and sends the picture to an online service to be read. This tool needs no account or Office licence, works in any modern browser — Edge, Chrome, Firefox or Safari — and does the recognition on your own computer. Once the page and the language file have loaded it even works offline, which makes it practical for confidential figures such as budgets, payroll exports and bank statements.

How to convert an image to Excel

  1. Add the picture — drop a PNG, JPG or WebP file or a scanned PDF (every page becomes its own table), or take a screenshot with Win+Shift+S and press Ctrl+V on this page.
  2. Check the grid — cells the OCR engine was unsure about are shaded amber. Click a cell to correct it, and add or delete rows and columns as needed.
  3. Limit it to the table — if the picture also contains a heading, notes or a logo, drag a box around the table on the image. Only that area is read again.
  4. Export — download Excel (one sheet per table) or CSV, or copy the cells and paste them into Excel or Google Sheets. Markdown is there for wikis and README files.

Numbers stay numbers

With Convert numbers on, values such as 1,204.50, $45.00, ₹1,23,456, 1.234,56, (125.00) or 12% become real numeric cells, so SUM, sorting and pivot tables work straight away, and the original look — decimals, thousands separators, currency symbol, accounting brackets, percentages — is kept as the Excel number format. Codes with leading zeros such as item numbers, PIN and ZIP codes or account numbers stay text, so the zeros don't vanish. Mark the first row as a header to keep it as text and get an AutoFilter in Excel.

Tips for accurate table OCR

Limitations

Merged cells, cells whose text wraps onto several lines, rotated headings and handwriting are not rebuilt reliably, so check those areas in the grid. Tiny or blurry text lowers accuracy; a larger screenshot usually fixes it. Up to 20 PDF pages are read at a time — split longer files first. If your PDF already has selectable text, Invoice to Excel reads its tables exactly, without OCR.

Private by design

The image, the recognised text and your edits stay in your browser tab. Nothing is uploaded to SarvKit or anyone else — the only downloads are the OCR engine and language data, which are cached for next time. Need the words rather than a table? Use Image to Text. Collecting receipts for an expense claim? The Receipt Scanner builds the expense sheet for you, the Business Card Scanner turns cards into contacts, and OCR PDF makes a whole scanned document searchable.

Frequently Asked Questions

Yes. There is no sign-up, watermark, page limit or paid tier. The OCR runs on your own device, so there is no per-image server cost to pass on.
No. The Tesseract OCR engine runs inside your browser using WebAssembly. Your picture, the extracted table and your edits never leave your computer, which you can confirm in your browser's network panel.
Both turn a picture into cells you can review. Data from Picture is built into Microsoft 365 and reads the image with an online service. This tool needs no Microsoft 365 subscription, reads the image locally (so it also works offline), handles scanned PDFs page by page and can export CSV, tab-separated text or Markdown as well as Excel.
Yes. Drop a scanned PDF and each page (up to 20 at a time) becomes its own table and its own sheet in the Excel file. For PDFs that already contain selectable text, Invoice to Excel reads the table directly and exactly.
Columns are found from the gaps between words. Tightly spaced tables can merge neighbouring columns and very wide cells can split. Move the Column sensitivity slider — the grid is rebuilt instantly from the same OCR result. Cropping to just the table also helps.
Yes, with Convert numbers ticked. Amounts with thousands separators, currency symbols, brackets for negatives or a percent sign are stored as numbers with a matching number format, while IDs with leading zeros stay text. Untick it to export every cell exactly as written.
Not reliably. The engine is trained on printed text. Neat block capitals sometimes work, but for handwritten sheets expect to correct many cells by hand.
21 languages, including English, Hindi, Marathi, Bengali, Tamil, Telugu, Gujarati, Spanish, French, German, Portuguese, Arabic, Russian, Chinese, Japanese and Korean. Each language file downloads once and is then cached by your browser.