fileformer
search⌘KtranslateENexpand_moreupload_fileConvert
Converters8 file families · 44 direct conversions
What changes in a conversionarrow_forward
home/pdf/pdf-to-txt

Convert PDF to TXT Online Free

A PDF knows where each letter goes, not what order to read them.

Inside a PDF, text is scattered pieces placed at coordinates. Pulling it out naively mixes two columns line by line, repeats the header on every page and splits words at hyphens. Toggle each cleanup step and watch the plain text change.

annual-report.pdf · page 4 · numbers show the reading order
-
1
2
4
-
3
5
-
Follow the columnsread each column top to bottomDrop headers & footersrepeated on every pageRejoin hyphenated wordsre-covery → recoveryKeep figure captionsas [bracketed] lines
annual-report.txt9 lines · UTF-8
Growth in a difficult yearRevenue rose 12% to £48m, driven by new contracts in the north and a recovery in export orders.The board proposes a final dividend of 4.2p per share.Costs were held flat despite inflation, thanks to the new plant.Staff numbers grew to 312, with apprentices for the first time.
Clean, readable text in the order a person would read it.notesExtract text
1 · your file
folder_openOpen filedescriptionSample .docx
The converter engine (doc-engine.js) failed to load: Failed to fetch dynamically imported module: http://127.0.0.1:9777/js/main/general/doc-engine.jsDrop any of 17 readable formats here: Word, ODT, RTF, AbiWord, KWord, StarOffice, EPUB, DocBook, XPS, HTML, Markdown, CSV or text. The format is read from the bytes, not the name.
2 · convert to
.pdfshare, print.docxedit in Word.odtLibreOffice.epube-readers.mdREADMEs.txtwords only
searchAll 28 formats · 14 written hereexpand_more
search
.
No format matches.
A4Letter
PDF 1.7 with the standard 14 fonts (nothing embedded), lines wrapped with Helvetica metrics, link annotations.
swap_horizConvert to .pdfdownload ·
What survives waiting for a file
partin sourceread backstatus
text_fieldsWords---
titleHeadings---
format_list_bulletedList items---
tableTables · cells---
format_boldBold runs---
format_italicItalic runs---
linkLinks---
imageImages---
rate_reviewComments---
translateCharacters---
Open a document or the sample. The table fills in from the file itself; after converting, the output is read back and counted again.
~/one-document-ten-outputs

What each format does to your document.

We converted one source document into every major output format and measured the file that came out. The source was a 37.4 KB DOCX with headings, bold and italic text, a table, numbered and bulleted lists, and one embedded image.

289%PDF, the largest output
166%RTF, bloated but portable
3%Markdown and TXT, image dropped
formattypeoutput size · vs 37.4 KB sourcelayoutkeepsverdict
.docxEditable37.4 KB · 100%ReflowsEverythingThe editing default
.docEditable51 KB · 136%ReflowsEverythingLegacy, larger than DOCX
.odtEditable33.5 KB · 90%ReflowsEverythingOpen standard, smallest office format
.rtfEditable62.1 KB · 166%ReflowsText, styling, tablesVery portable but bloated
.pdfFixed108 KB · 289%ExactExact appearanceFor sharing and printing, not editing
.epubReflowable31.3 KB · 84%Adapts to screenText, styling, imagesFor e-readers
.htmlMarkup1.5 KB · 4%ReflowsText, styling, tablesOpens in any browser
.mdPlain text1.2 KB · 3%ReflowsText, basic stylingSmallest, great for docs and notes
.txtPlain text1.1 KB · 3%NoneText onlyUniversal, but drops all formatting
.dbkMarkup3.5 KB · 9%StructuredText, structureTechnical documentation only
Plain text is tiny because the image goesMarkdown and TXT both came out at 3% of the source, against 100% for DOCX: the embedded image is the bulk of the file, and neither format can hold it.
PDF pays for a fixed pageAt 289% of the source, PDF was the largest output. That is the cost of fixing the layout, fonts and image into a page that looks identical everywhere.
RTF is the surpriseIt keeps formatting but grew to 166% of the DOCX, so it is only worth using when the receiving program cannot read anything better.

Sizes are from one representative document converted with Pandoc and LibreOffice. Your own files will differ, especially with more images, but the relative ordering holds.

~/is-your-pdf-really-text

Why PDF to Word is the hard direction.

A PDF has no paragraphs. It stores “draw these glyphs at x, y in this font”, often one word or one line at a time, and a scanned page stores only a picture. Drop any PDF to see which kind you have and what its page instructions look like.

picture_as_pdfInspect a PDFText PDF sampleScanned PDF sample
Nothing inspected yet. The two samples are written here: one with real text, one that is only a JPEG of a page.
EXTRACTED TEXT (first 600 characters)
first page instructions, decoded · operators explained
~/no-text-layer

A scanned page is a photo. There's no text to extract until something reads it.

If you can't select words in the PDF, it has no text layer and the TXT would be empty. We detect this per page and run text recognition (OCR) on those pages only. Recognition makes mistakes on poor scans; we mark words it's unsure about.

Clean scanPhone photoOld fax
the scan
Tenancy AgreementThis agreement is made on the 4th of March between the landlord and the tenant for the property at 12 Canal Street.The rent of £1,150 is payable monthly in advance.
recognised text
TenancyAgreementThisagreementismadeonthe4tbofMarchbetweenthelandlordandthetenantforthepropertyat12Cana1Street.Therentof£1,15Oispayablemonthlyinadvance.
3 uncertain words highlighted · 91% confident
~/what-plain-means

"Plain text" still has settings. The wrong ones garble names.

A TXT file stores characters as numbers using an encoding, and marks line ends differently on Windows and Mac. Pick what will open the file; the preview shows how the same sentence arrives.

desktop_windowsWindows Notepad (old)Windows 7-10UTF-8 with BOM · CRLFlaptop_macMac TextEditmacOSUTF-8 · LFcodeScript or search indexPython, grep, ElasticUTF-8 · LFhistoryLegacy systemolder databasesWindows-1252 · CRLF
what you getcorrectCafé “Zoë”: résumé £48m
same file, opened with the wrong encodinggarbledCafé “Zoë� – résumé £48m
line endingsCRLF (\r\n)Windows style: two characters. Old Notepad shows one long line without them.
Older Notepad guesses the encoding; a byte-order mark at the start tells it the file is UTF-8, so accents and curly quotes come through.
~/why-plain-text

Plain text is for machines and searching.

If you need to edit the document with its layout, PDF to Word is the better route. TXT wins when the words matter and the looks don't.

searchSearch a big archivebestText files index in seconds; find a phrase across thousands of reports.
smart_toyFeed an AI or scriptbestLanguage models and scripts read clean text far more reliably than PDFs.
record_voice_overScreen readersgoodA linear text is easier to listen to than a PDF with a broken reading order.
translateTranslationgoodPaste into any translation tool without layout debris.
edit_documentEdit the documentuse WordTXT loses headings, tables and images. PDF to Word keeps them.
table_chartPull out tablesuse ExcelTables become ragged lines in TXT. PDF to Excel keeps the cells.

Frequently Asked Questions

Is converting PDF to TXT free?

Yes, PDF to TXT conversion is completely free with no signup, no watermark, and no usage limit. Convert as many files as you need.

Are my files uploaded to a server?

Your PDF to TXT file is sent over an encrypted connection to our servers, converted, and returned to you. Files are processed only to perform the conversion and are not stored or retained afterward, so they are removed automatically once your download is ready.

How long does PDF to TXT conversion take?

Most PDF to TXT conversions finish in under a minute. Larger files take proportionally longer, but everything happens locally so there is no sign-up wait.

Will the conversion reduce file quality?

That comes down to the formats at each end. A lossless conversion changes nothing you can measure. A lossy one shaves the file down and takes a small amount of quality with it.

Does this work on mobile devices?

Yes. The PDF to TXT converter works on phones, tablets, and desktops in any modern browser including Chrome, Safari, Firefox, and Edge.