PDF to Excel Converter

What the detector can and cannot recover

PDF pages store positioned text, not spreadsheet cells. This converter clusters nearby text into rows, looks for repeated column anchors, pads missing cells, and rejects blocks that do not resemble a regular table. It also accounts for page rotation and the visible page box when reading positions.

The detector is deliberately conservative. It is better to report no table than to turn unrelated page text into a convincing but incorrect spreadsheet. Always compare important values with the source PDF before calculating, importing or reconciling data.

Export formats

XLSX creates one worksheet per selected table, up to 200 worksheets in one download. Every extracted value is written as a text cell. The converter does not infer numbers, dates, formulas, currencies or accounting formats, and it does not recreate borders, fonts, merged cells or the original page design.

CSV and TSV download each selected table separately. Comma, semicolon and tab delimiters are available. CSV-style fields are escaped with double quotes when they contain a delimiter, quote or line break, and records use CRLF line endings following the common rules documented in RFC 4180. A UTF-8 BOM is enabled by default to improve compatibility with spreadsheet programs, but the correct delimiter still depends on the application and regional settings used for import.

Spreadsheet formula safety

A PDF can contain text that begins with =, +, - or @. Some spreadsheet programs may interpret that text as a formula when a CSV file opens. Formula protection is enabled by default and prefixes formula-like CSV cells so they are more likely to remain text, including cells with leading whitespace or full-width variants of those characters.

This is a risk reduction, not a universal security guarantee. OWASP’s CSV Injection guidance notes that spreadsheet behavior varies and that escaping can change after a CSV is saved and opened again. Leave protection enabled for untrusted PDFs, inspect the exported cells, and do not enable macros or dismiss spreadsheet security warnings. XLSX output writes extracted values as explicit string cells rather than formulas.

Limits and private browser storage

• One PDF at a time, up to 20 MiB and 150 pages.
• PDFs with excessive text fragments, text volume or table data are rejected before they can overwhelm the browser.
• The on-page preview shows the first 100 rows of each detected table; the export includes all detected rows within the safety limits.
• A workbook can contain up to 200 selected worksheets.
• Password-protected and encrypted PDFs are not supported. Unlock a permitted copy before using the converter.
• During the three-step flow, the source PDF, detected tables and settings are stored privately in this browser for 30 minutes. At most three jobs and 80 MiB of local job data are retained.
• Only a random job identifier appears in the funnel URL. The PDF, extracted values and downloaded files are not sent to our servers or uploaded to the results server.
• Choosing Start over removes the current local job. Expired or invalid records are removed when the tool next checks browser storage.

The PDF is loaded from its local bytes through PDF.js. The tool downloads its application libraries as normal web assets, but it does not send the selected PDF with those requests.

Accessible review

The upload control works with a keyboard, progress and errors are announced to assistive technology, and detected tables use captions and column headers. The preview supports horizontal scrolling on small screens and preserves the detected direction for right-to-left tables. Accessibility support helps you review the extraction, but it does not replace checking the source when accuracy matters.