Overview
Extract the raw text from any PDF directly in your browser using PDF.js, Mozilla's PDF renderer. Upload a file, and the tool walks it page by page and returns the text with page markers - useful for pulling quotes, searching long documents, or feeding content to another tool. No upload to any server: the PDF never leaves your device.
How It Works
Select a PDF file with the upload button. The page shows Extracting Text while PDF.js parses the document, then displays the full text in the output area, with each page's content preceded by a --- Page N --- marker so you can see where pages break. Extraction errors, such as a corrupt or password-protected file, are reported plainly in the output area rather than failing silently. Copy the result or work with it in place; processing happens entirely locally.
Step-by-Step Usage Guide
- Select the PDF file to extract from.
- Wait for the Extracting Text indicator to finish.
- Read the output, using the --- Page N --- markers to navigate.
- Copy the text into your editor or notes.
Technical Specifications & Standards
The engine is PDF.js loaded on demand, driven through getDocument and, for each page, getTextContent - the API that returns the text runs (strings with their positions) rather than a rasterised image. Text is concatenated per page in the order the PDF's content streams provide, which matches the visual order for well-made PDFs but can scramble text in documents with complex columns, tables or right-to-left scripts, because PDF is a positioning format, not a flowing-text format. That is the fundamental limit to understand: PDFs store glyphs at coordinates, and any text extractor reconstructs reading order heuristically. The corollary is the scanned-document case: a PDF of photographed pages has no text runs at all, so extraction returns empty or near-empty output - there is no OCR in this tool. Page markers make the boundaries auditable either way. Files are read into memory as ArrayBuffers and closed again; nothing is persisted or transmitted.
Targeted Use Cases
- Pulling the text of a research paper or report for quoting.
- Searching a long contract for specific terms without a PDF reader's limitations.
- Recovering content from a PDF for re-use in a document or CMS.
- Checking whether a PDF actually contains selectable text or is just scans.
Notes & Gotchas
- Expect reading-order quirks in multi-column PDFs; verify quotes against the original layout.
- If output is empty, the PDF is probably scanned images - this tool does not OCR.
- Use the page markers to cite locations accurately.
- Process confidential documents freely: the file is parsed in your browser, not uploaded.
Frequently Asked Questions
Does my PDF get uploaded anywhere?
No. The file is read into browser memory and parsed with PDF.js locally. Nothing is transmitted, stored or logged.
Why is the extracted text jumbled for some PDFs?
PDF stores text at coordinates, not as flowing text. Multi-column or table-heavy layouts reconstruct imperfectly because the extractor follows the content stream's order.
Why did my scanned PDF return no text?
Scanned pages are photographs without a text layer, so there is nothing to extract. This tool does not perform OCR; run an OCR tool first to add a text layer.
What do the page markers mean?
Each --- Page N --- line marks the start of that page's text in the output, so you can locate content and cite page numbers precisely.