How to Extract Text From a PDF Without Paying for Adobe

A PDF arrives in your inbox. It is a contract, a report, a paper, a manual. You need to quote a section in an email. You need to copy a paragraph into a document. You need the text in a form you can edit. Adobe Acrobat costs $20/month. The free Acrobat Reader does not let you copy text. What do you do?

You use a browser-based tool. The text is already in the PDF — most modern PDFs have a text layer that search engines, screen readers, and copy-paste can read. A few clicks and you have the text. The whole process takes less time than opening Acrobat, signing in, navigating the menus, and clicking "Export".

The two kinds of PDFs

There are two kinds of PDFs, and the workflow for extracting text depends on which kind you have.

Native-text PDFs. The PDF was created from a digital source (Word, Google Docs, a website, LaTeX) and the text is embedded in the file as actual characters. You can search the PDF (Cmd+F), copy text from it, and screen readers can read it. Most modern PDFs are this kind. The extraction is trivial: the tool reads the embedded text and gives it to you.

Scanned PDFs. The PDF was created by scanning a paper document. There is no text layer — the PDF is a series of images, one per page. You cannot search the PDF, cannot copy text from it, and screen readers see nothing. These are the PDFs from old offices, government archives, and faxes. The extraction requires OCR: an engine that looks at the image and recognizes the characters.

The first step is to figure out which kind you have. Open the PDF in your browser (drag it onto a Chrome tab). Try to select some text. If you can, it is a native-text PDF. If you cannot — the cursor slides over the page as if it were an image — it is a scanned PDF.

How to extract text from a native-text PDF

For a native-text PDF, the workflow is two steps:

1. Open the PDF in your browser. Drag it onto a new tab, or use Cmd+O / Ctrl+O to open it from your filesystem.
2. Select the text you want with the mouse, copy it (Cmd+C / Ctrl+C), paste it into a document.

That is it. The browser has a built-in PDF viewer that lets you select and copy text. No tool needed. The output is usually good but not perfect: line breaks are sometimes in the wrong place, ligatures (fi, fl) are sometimes split into two characters, and some special characters (smart quotes, em dashes) might come through as their HTML entities.

For a cleaner extraction — a single block of text per page, with line breaks only at paragraph boundaries — use a dedicated PDF text extractor. The tool reads the embedded text and reformats it, usually with a checkbox for "preserve layout" or "extract plain text". The plain text output is cleaner; the preserve-layout output is closer to the original.

How to extract text from a scanned PDF

For a scanned PDF, you need OCR. The workflow is three steps:

1. Open the PDF in the PDF OCR tool.
2. The tool converts each page to an image, then runs OCR on each image. The output is a text block for the whole document, or one text block per page.
3. Copy the text, paste it into a document, proofread.

The OCR is not perfect. For a clean scan at 300 DPI, accuracy is 95-99%. For a poor scan, a fax, or a low-resolution image, accuracy drops to 80-90%. You will need to proofread the output and fix any errors. The errors are usually systematic — every "rn" comes out as "m", every "l" comes out as "1" — so once you know the pattern, the proofreading is fast.

Why browser-based, not a desktop app

For native-text PDFs, the browser is enough. The text is already in the file; the browser's built-in PDF viewer lets you select and copy it. There is no need to install anything.

For scanned PDFs, the OCR engine is the question. Adobe's OCR is good but locked behind a subscription. Tesseract.js is the standard open-source alternative, and it runs entirely in the browser. The Uttir PDF OCR tool wraps Tesseract.js with a simple UI: upload the PDF, get the text. The file never leaves your device.

The trade-off: Tesseract.js is slower than Adobe's engine (10-30 seconds per page vs 1-3 seconds) and slightly less accurate on edge cases (handwriting, low-contrast scans, mixed languages). For the common case — a clean scan of a printed document — the accuracy is good enough that the proofread takes a minute, not an hour.

How to handle the 12 special cases

Some PDFs have features that break the simple workflow. Here is how to handle them.

Multi-column layout. A two-column PDF, when extracted as plain text, often comes out with the columns interleaved (line 1 of column A, line 1 of column B, line 2 of column A, line 2 of column B, ...). To fix this, use a tool with a "preserve layout" option, or OCR the PDF with a layout-aware engine.

Headers and footers. Most PDFs repeat the header and footer on every page. When you extract all pages, you get the header 20 times. The fix: use a tool with a "strip headers/footers" option, or paste into a document and find-and-replace the duplicates.

Footnotes and endnotes. Footnotes are sometimes in a smaller font at the bottom of the page, sometimes in a separate section at the end of the document. The simple extraction may merge them into the main text, or separate them in a confusing way. The fix: pick the extraction mode that matches what you want. For most purposes, "extract plain text" with footnotes inline is fine.

Tables. Tables in a PDF are hard. The text extraction loses the structure: cell 1 of row 1 comes out, then cell 1 of row 2, with no separator. The fix: OCR the PDF to an image, then use a table-aware OCR engine that preserves the cell structure. Or manually re-create the table in a document.

Embedded images. An image with text in it (a screenshot, a scanned signature, a chart with labels) is not part of the text layer. The extraction will skip the text in the image. The fix: run OCR on the whole PDF, including the images, or extract the images separately and run OCR on each.

Password-protected PDFs. The extraction tool will prompt for the password. The browser's built-in viewer will also prompt. Once unlocked, the extraction works the same as for an unprotected PDF.

How to keep the file on your device

A browser-based PDF text extractor reads the file in your browser and never sends it to a server. The text output is generated locally and copied to your clipboard. There is no upload, no account, no log of what you extracted.

The check: open the browser's DevTools, go to the Network tab, drag a PDF onto the tool, and confirm no requests are made while the extraction runs. The only network request should be the initial page load; the extraction itself should be entirely local.

This matters for two reasons. The first is privacy: a contract, a medical record, a personal letter — these are documents you may not want to upload to a third-party service. The second is reliability: a tool that runs locally does not depend on a third-party service staying online. The tool can disappear tomorrow and the PDFs you already extracted are still extracted.

When you actually need Acrobat

For extracting text, you do not need Acrobat. The browser-based tool does the job in less time and for less money.

For editing a PDF — changing the text in the PDF itself, not extracting it to another document — you do need a real PDF editor. Acrobat, Foxit, PDFsam, and a few others are the options. Most of them are not free. For the common case of "I need to fix a typo in this PDF", the workaround is: extract the text, edit it in a document, regenerate the PDF from the edited document.

For filling in a form PDF (the kind the IRS sends), Acrobat is the standard. There are free alternatives, but Acrobat's form-filling is the gold standard. If you fill in a lot of form PDFs, the subscription is worth it. If you do it once a year, the browser's built-in form-filling (Chrome and Edge both have it) is enough.