Compare PDF files and see every change

Who needs to compare PDFs

Any time someone sends you "the updated version" of a document, you need to know what they changed. Reading both copies line by line takes a long time, and it is easy to miss one word. This tool reads both and marks every difference for you, so you only have to look at what changed.

• Contracts and agreements. Before you sign, check that the final copy says what was agreed. One changed number, a new clause or a missing line can matter a lot.
• Rent agreements. The owner or the broker sends a new draft. See at once whether the rent, the deposit, the notice period or the lock-in period changed.
• Offer letters. Compare the offer you accepted with the letter that comes with your joining papers: the salary, the notice period, any bond and the probation terms.
• Tenders and bids. Find what changed between a tender notice and its corrigendum, or between two revisions of your own bid.
• Policies and terms. Insurance policies, loan papers, bank terms and company rules are updated often. See the new wording without reading every page again.
• Assignments and theses. Students can check that the corrections they were asked for are all in. Teachers and guides can see what changed between two drafts.
• Proofreading. Check that the printer's proof, the final export or the copy from a colleague matches the text you approved.

It works for any two versions of the same document, as long as the PDFs have text in them. For scans and photos of pages, which are pictures, there is Compare the look.

Step by step

1. Add the original. Press Choose PDF under "Original", or drop the file on it. You see its name, its page count and a small picture of its first page.
2. Add the changed version. Do the same under "Changed version". You can also pick or drop both files at once: the older file becomes the original.
3. Wait a moment. Each file is read as soon as you add it, and the comparison starts by itself once both are in. A bar shows how many pages have been read. If you picked the wrong file, press Cancel; your files stay where they are.
4. Read the summary. It says how many changes were found, and how many of them are additions, removals and changes.
5. Go through the changes. Use the up and down arrows in the toolbar (or the N and P keys) to move from one change to the next, or switch to List to read them all.
6. Keep a record. Copy list or Save report gives you every change as plain text with its page numbers, to paste into an email or keep with your papers.

If the files went in the wrong way round, press the swap button between them. The changes are found again the other way, so what was added now shows as removed. Change and Remove under each file let you replace or take out one file without starting over.

Want to see it first? Press "Try with sample PDFs". It makes two versions of a short rent agreement right in your browser, with five changes, and compares them.

How the changes are found

A PDF saved from a document keeps its text, and the place of every piece of text on the page. The tool reads the text of every page of both files with pdf.js, the PDF reader made by Mozilla, and splits it into words. Each word keeps its page and its place on that page, so a change can be marked exactly where it sits.

Then it puts all the words of each file in one long line, from the first page to the last, and compares the two lines. It looks for the longest runs of words that are the same in both, in the same order. What is left between those runs is a change: words only in the original were removed, words only in the changed version were added, and where both have different words in the same spot, they were changed.

Because the words are compared as one long line, pages and lines don't matter. If a new sentence on page 1 pushes everything after it down, the words after it are still the same words in the same order, so only the new sentence is reported. The same goes for a document that was saved again on other paper, with other margins or a different font size.

Headers and footers are handled on their own. A line near the top or bottom of the page that repeats on at least three pages (two in a PDF of up to four pages), like "Confidential" or "Page 3 of 20", is taken out of the main comparison. If it weren't, every page number would look changed wherever text moved to another page. The headers and footers are then compared with each other. Page numbers that count up with the pages, and the total number of pages, are not differences. A header that really changed, say from "Version 1" to "Version 2", is reported once, as one change in a header or footer, not once for every page.

A few more details help it treat as the same what you would read as the same:

• Numbers like 25,000 or 10:30, and words like e-mail or don't, count as one word. A mark such as a full stop or a bracket counts as a word of its own, so a changed comma is found.
• Letters that some programs join into one sign, like "fi", count as their plain letters.
• Chinese and Japanese characters are compared one by one, because they are written without spaces. Hindi and other scripts are compared word by word, with their vowel signs.
• For long documents, the tool first lines up the words that appear exactly once in each version, then compares the short pieces between them closely. That keeps a 300-page pair quick, and stops one change from pulling the words around it out of place.
• The comparing runs in the background, so the page keeps responding even with long files, and both files are read at the same time.

Reading the results

The summary gives you the answer first: the number of changes, or "No differences in the text". Coloured labels split the changes into added, removed and changed. The small line under them says how long it took and how many pages each file has. If you chose the very same file twice, the summary says that too: "They are also exactly the same file, byte for byte."

The Pages view

The Pages view shows both versions side by side, the original on the left and the changed version on the right. The changes are marked on the real pages, and each kind has its own shape as well as its own colour:

• Added: green, with a line under it. These words are only in the changed version.
• Removed: red, with a line through it. These words are only in the original.
• Changed: yellow, with a dashed border. The old words on the left became the new words on the right.

Where words were added or removed, the other side gets a small marker at the spot where they would be, so you can see where the gap is. The legend above the pages repeats what each mark means.

When you scroll one side, the other follows to the same place in the text, even when the pages no longer match. A page is drawn only when you get close to it, and is let go again when you are far away, so a 300-page pair doesn't fill your device's memory.

The arrows in the toolbar move to the previous or the next change. The current change gets a blue outline, both sides scroll to it, and the line above the pages says what it is in words, for example: Change 1 of 5: changed on page 1: "1" → "15". Screen readers read that line out. Zoom goes up to 300% for small print, and the full screen button gives the pages the whole screen. Press Esc to leave full screen.

The List view

List shows every change as a card: its number, its kind, its page, and the words before and after, with a few words around them so you know where it is. If the text moved to another page, the card shows both page numbers, like "Page 7 → 8". Tap a card to see that change on the pages.

A very long list shows the first 1,000 changes. Save report always has all of them.

Notes in the summary

Two notes may appear. If fewer than 35% of the words match, the summary says the PDFs have little text in common: check that you picked two versions of the same document. If some pages have no text at all, such as a scanned page, a picture or a blank page, it says how many, because changes on those pages can't be listed. Compare the look checks them.

The two options

Ignore extra spaces and line breaks

This one is on at first, and most of the time you should leave it on. The tool then compares the words only. A line that breaks in a new place, or a word split by a hyphen at the end of a line, is not a change. Neither is a space that only splits or joins words, so "Rs.25,000" and "Rs. 25,000" count as the same.

Turn it off when the lines themselves matter, such as a poem, an address block or a list typed line by line. Then every line break in a new place is reported, and so is a space that splits or joins words. Extra spaces between words are never counted, because a PDF often doesn't store the spaces at all, only where each word sits.

A real example: we laid out the same 21 pages of text again on Letter paper with other margins, so it came out as 20 pages with the lines breaking in new places. With this option on, the result was "No differences in the text". With it off, the same pair gave 830 changes, every one of them a line break in a new place.

Ignore capitals (case)

This one is off at first. Turn it on when you only care about the words, not how they are written: "Agreement" and "agreement", or "RENT" and "Rent", then count as the same. Leave it off when capitals matter, as in names, codes or headings.

You can change either option after comparing. The files are compared again straight away with the new setting, and Save report notes which options were used.

On a phone

A phone is too narrow for two pages side by side, so on a phone you see one version at a time. The Original and Changed switch in the toolbar flips between them, and the version you flip to opens at the same place in the text.

When you move to the next change, the tool shows the side where you can see it: a removed word on the original, an added word on the changed version. A changed word is on both, so you stay where you are and can flip to see the other.

The buttons are big enough to tap, and the page fits the screen without sideways scrolling, down to screens 320 pixels wide. When the comparison is done, the summary comes into view by itself. On a small screen the List view is often the quickest way to check: each card shows the words before and after.

Phones have less memory and slower processors than most computers, so long files take longer to read. Pages are drawn only as you reach them, which keeps memory use low.

Real results

We tested the tool with PDFs in which we knew every change, and with real PDFs made by many programs. The tests ran in Chromium 141, the open-source browser Google Chrome is built on, on a Linux computer with 4 processor cores and 16 GB of memory. Your device may be faster or slower.

• 6 of 6known changes found in two 298-page PDFs
• 4.6 sto compare those 298-page PDFs
• 0changes reported for text that only moved
• 0 bytesof your files sent anywhere

What it found

How long it took

These are the times the page showed, from choosing the files to the result, over five runs on a freshly opened page (three runs for the 42-page scan). For Compare the look, the time starts when you press the button.

Most of the time goes into reading the text of every page, and that depends on how a PDF was made as much as on its length. The 117-page report has fewer words than the 148-page document, about 30,000 against 71,000, yet it took longer to read. Finding the changes once the words are read is quick: in separate runs on the 117-page and 298-page pairs, it took 0.03 to 0.13 seconds, including showing the result.

Scanned PDFs

A scan is a picture of a page. It has no text inside, so there are no words to compare. When one or both files are scans, the summary says so plainly, "These PDFs have no text to compare", and offers Compare the look.

Compare the look draws each page of both files as a picture, 800 dots across. It lines the two pictures up, because a scan can sit a few dots off, and compares them colour by colour. A dot only counts as different when no dot right next to it on the other page matches, so the soft edges of letters and a slight shift don't count. Areas that differ are grouped into boxes and marked on both versions, page by page. A page that is only in one version, or a page of a different shape, is reported too.

It works best on two scans of the same kind, such as two scans from the same scanner. A scan and a text PDF never line up exactly, so comparing their look marks most of each page.

You can use Compare the look on text PDFs as well. It sees what the word comparison can't: a new stamp, a signature, a changed logo, a photo or a colour. When you are done, Back to the text takes you back to the words.

It can't tell you which words changed, only where the pages look different. If the same file was chosen twice, it says so at once without drawing anything.

Some scans were made with text recognition (OCR). They carry hidden text under the picture, and the word comparison uses it. The result is then only as good as the recognition: a letter read wrongly in one copy shows up as a change.

Why nothing is uploaded

Your files never leave your device. The page reads them, compares them and draws them in your browser. There is no upload step, no server that sees your documents and no account.

Many online compare tools work differently. They upload both files to their server, compare them there and send the result back. Your documents then sit, at least for a while, on a computer you don't control, and you have to trust what happens to them there. For a rent agreement with your address, an offer letter with your salary or a bid before its deadline, that is a real risk.

You can check this yourself. Once this page has loaded, switch off your Wi-Fi and mobile data, then compare two files. It still works for most PDFs, because the tool is already on your device. A few need an extra part of the PDF reader the first time, such as the fonts for Chinese or Japanese text, so those need the connection on. In our tests we also logged every request the page made after it loaded: none carried a file, and the only one fetched a piece of the page's own code.

The tool doesn't keep your files after you leave the page. Page pictures are let go as you scroll away from them, and the small previews are freed when you remove a file or leave. Our privacy policy has the details.

What it can't see

The word comparison looks at the words and their order. That covers most changes that matter in a document, but not all of them:

• Formatting. Bold, italics, fonts, sizes and colours are not compared, only the words. Compare the look shows changes like these when they can be seen on the page.
• Pictures, stamps and signatures. They have no words, so the word comparison doesn't see them. Use Compare the look.
• Tables. A table is read as words, usually row by row. A changed figure in a cell is found and marked. If the table was built again, for example with its columns in a different order, the words come in a different order and you will see many changes.
• Moved text. A paragraph moved to another part of the document shows as removed where it was and added where it went.
• Form fields and comments. Text typed into form boxes and notes added as comments are kept apart from the page text, so the word comparison doesn't see them. Compare the look shows them if they are drawn on the page.
• Two files at a time. If you pick more than two, the first two are used.
• Size. We set no limit on size or pages, but a large file needs memory, and a phone has less of it than a computer.

If something goes wrong

• "This file can't be read as a PDF." The file is damaged, cut short, or not really a PDF, such as a renamed document. Download it again or ask for a new copy, then use Change to pick it.
• "This PDF is locked." Type its password in the box that appears with the file and press Open. The password is used only on your device to open the file. If it says the password didn't work, check capital letters and try again.
• "These PDFs have no text to compare." One or both files are scans. Press Compare the look.
• Far too many changes. Check that both files are versions of the same document, and that Ignore extra spaces and line breaks is on. If one file was made in a different program, its text may come in another order, for example in two columns. The List view then helps you see what really changed.
• A change you expected is missing. It may be formatting, a picture or a stamp, which the word comparison doesn't see: try Compare the look. It may be on a page without text; the summary says so if it is. Or an option may be hiding it, such as Ignore capitals.
• The changes seem the wrong way round. The newer file went in as the original. Press the swap button.
• Strange letters in the list. Some PDFs, often ones typed with older Hindi fonts, store their letters in a code of their own. Use the Pages view, which shows the real pages, or Compare the look.
• It takes long. Long files take longer on a phone. Close other apps and tabs to give it memory. Cancel stops it, and your files stay ready to compare again.
• Copy list does nothing. Some browsers block copying. Use Save report instead: it saves the same list as a text file.
• "This tool couldn't start in this browser." Your browser is too old for it. Try an up-to-date Chrome, Edge, Firefox or Safari.

Tips

• Put the older file first. Then "added" means new in the latest version. If you pick both at once, the older file goes first by itself.
• Ask for the PDF, not a photo. A PDF saved from the document can be compared word by word. A photo or scan can only be compared by its look.
• Check the numbers. Amounts, dates, notice periods and percentages are where small changes matter most. Every changed number is listed as a change of its own.
• Do a last check with Compare the look. Before you sign, it shows a new stamp, a missing signature or a changed logo that the words alone don't.
• Keep the report. Save report gives a short text file listing every change with its page numbers. Keep it with the signed copy, or paste it into your reply to the sender.
• Use List on a phone. It shows each change with the words around it, without zooming.
• Got only photos of the pages? JPG to PDF turns them into one PDF, which you can then compare by its look.
• Sending the final copy? The PDF compressor makes it smaller for email, and Resize PDF fits it under an upload limit.