PDF to Text Converter — Scanned PDFs Too (OCR)
Get plain text from any PDF: text pages are read directly, scanned pages with OCR. Runs in your browser, nothing uploaded. Copy it or download TXT or Word.
Loading only the code needed for this tool…
Accuracy we measured
We ran this tool on our own test images (no customer files) and compared its output with the known text. The character error rate counts wrong, missing and extra characters; 0% means every character was right.
| Test image | Language | Errors | Engine defaults | Time | Phone profile |
|---|---|---|---|---|---|
| PDF saved from a word processor (text layer) | English | 0% | — | 1.3 sec | 2.7 sec |
| Scanned two-column page saved as PDF | English | 0% | — | 6.3 sec | 24.9 sec |
| Two-page PDF: one text page, one photo page | English | 0% | — | 3.6 sec | 7.8 sec |
| Same notice, phone photo at an angle | English | 0.1% | 2.4% | 3.7 sec | 6.2 sec |
| Printed Hindi page | Hindi | 0% | 0% | 1.8 sec | 4.3 sec |
| Printed Simplified Chinese page | Chinese (Simplified) | 0% | 0% | 1.8 sec | 4.3 sec |
Engine defaults: the same engine and language data without our clean-up (automatic layout, no straightening, no light correction).
How we measured: the images come from a script that renders our own sentences, photographs them at a 4° angle with uneven light, and saves JPEGs, PNGs and PDFs. Times are from a 4-core cloud computer; the phone profile adds Chrome's 4× CPU slowdown. First runs also download the engine (about 1.1 MB) and the language data.
Tool quality check
Use the output as a verified working copy
Keep the original source when it may be needed again. Open the generated TXT in the application that will receive it, and check that its content and formatting meet your needs.
Input
Output
TXT
Batch
Up to 5 files per run
Up to 5 files per run. Only use one run when files share the same operation.
Move into a related operation instead of re-finding the tool.
Download a destination-ready result and verify it before sharing.
Configured runtime limit applies.
TXT
Check the upload limit below
Review the result before sharing
Smart workflows
Finish the job, not just the file
Use the existing tool graph to continue into useful next steps. Bulk and output-packaging capabilities are surfaced automatically where the underlying tool supports them.
Continue with Image To Text
Bulk-readyFinish the job after Pdf To Text with a relevant MAXScanner tool, without creating a duplicate converter page.
- Pdf To Text
- Image To TextBulk processing · ZIP-ready output · Quality controls
Best when Pdf To Text is only the first step in your workflow.
Continue with Pdf To Docx
Bulk-readyFinish the job after Pdf To Text with a relevant MAXScanner tool, without creating a duplicate converter page.
- Pdf To Text
- Pdf To DocxPage selection · Workflow-ready
Best when Pdf To Text is only the first step in your workflow.
About This Tool
To convert a PDF to text, add the PDF and select Extract text, then copy the result or download it as a .txt or Word file. Pages that already contain text, such as a PDF saved from Word, a bank statement or an e-book, are read straight from the file: exact, and quick. Pages that are only a picture, such as a scan or a phone photo saved as PDF, are read with OCR in the language you choose.
The result marks each page (--- Page 2 ---) and tells you how many pages came from the PDF's own text and how many needed OCR. The PDF stays on your device: PDF.js reads it and Tesseract recognizes scanned pages inside your browser.
In our test, a scanned two-column newsletter page came out with 0% character errors in the right reading order, and a two-page PDF with one text page and one photographed page took 3.6 s on our 4-core test computer (7.8 s with a slowed-down phone profile).
When to use a different tool
- To keep the PDF and make scanned pages searchable and selectable, use OCR PDF.
- To keep headings, fonts and tables in an editable document, use PDF to Word.
- For Markdown with headings and lists from a text PDF, use PDF to Markdown.
- A password-protected PDF needs its password first: unlock a copy with Decrypt PDF.
How to Use
Add the PDF
Drop or choose up to 5 PDFs. Text PDFs and scanned PDFs both work.
Choose the OCR language
The language is only used for pages without their own text (scans and photos). Pick Columns if a scanned page has two columns that come out mixed.
Extract and save
Select Extract text, check the result, then copy it or download TXT or Word.
Use Cases
Quoting from reports
Take a paragraph or a figure out of a long PDF report without fighting the PDF viewer’s text selection.
Scanned letters and contracts
Get searchable text from a scanned PDF so you can find a clause, a date or an amount.
Text for other tools
Paste the text into a translator, a word counter or an AI assistant, or clean it up with the text tools.
Frequently Asked Questions
How do I convert a PDF to a text file?+
Add the PDF on this page and select Extract text. When it finishes, select Download TXT, or copy the text into another app.
Does it work with scanned PDFs?+
Yes. Pages without a text layer are read with OCR in the language you choose. That is slower than reading a text page: our scanned newsletter page took 6.3 s on our 4-core test computer, against 1.3 s for a two-page PDF with a text layer.
Why is some text missing or in a strange order?+
A PDF stores text in the order it was drawn, which is usually, but not always, the reading order. For scanned pages with two columns, choose Columns in Page layout. Text inside pictures of a text PDF is not read; use OCR PDF for those pages.
Can I extract text from only some pages?+
Every page is read. To get text from a few pages only, take those pages out first with Extract pages, then convert the smaller PDF.
Is my PDF uploaded?+
No. PDF.js reads the file and Tesseract recognizes scanned pages in your browser. Only the recognition engine and language data are downloaded once, from the jsDelivr CDN.
Can I keep the formatting?+
Plain text has no formatting. For an editable document that keeps headings and tables, use PDF to Word; for a PDF that stays the same but becomes searchable, use OCR PDF.