Free Online PDF Editing: Which Features Actually Work

You get a contract as a PDF. A clause needs changing, you need to add your signature, and there's a phone number to redact before forwarding.

Sounds simple. PDF was designed in 1993 to make documents look identical on every screen and printer. That goal conflicts with being editable, and knowing why saves you from picking the wrong tool.

Short answer: Free online PDF editors handle annotations, signatures, form filling, and page reordering reliably, because none of that requires touching the document's existing content. True text editing, OCR on scanned pages, and real redaction are where free tools fall apart, because those require actually rewriting or stripping what's already there. The rest of this article covers why, and which specific tool to reach for.

Why Editing a PDF Is Hard

A Word document stores structure: paragraphs, headings, tables, columns. Click a paragraph and Word knows where it starts, where it ends, what font it uses.

A PDF stores drawing instructions. Place this glyph at coordinate (72, 650) using font F1 at 12 points. Draw a line from here to there. Render this JPEG at these coordinates. The document has no concept of a "paragraph." It knows where to put ink, nothing more. That's not an implementation shortcut — it's the design goal spelled out in ISO 32000-2, the current PDF specification: a page is a fixed sequence of drawing operators, not a document model, because the format was built to look identical on any printer or screen, long before "editable" was ever a requirement.

What this means in practice:

No reflow. Change one word and nothing else adjusts. Text won't rewrap, images won't shift, page breaks stay fixed.

Missing fonts. PDFs embed only the characters they use. If the document never contained the letter "Q" and you try to type one, the embedded font simply lacks it.

Flattened layers. Most PDFs are a single flat layer. Moving a text block can corrupt the background behind it because the editor has no way to separate them.

Scanned documents are worse still. A scanned PDF is a stack of images, no text at all. You need OCR first, and OCR introduces its own errors.

Six Types of PDF Editing

These range from trivially easy to borderline impossible in free tools.

Annotations

Highlights, comments, sticky notes, freehand drawings layered on top of existing content. Annotations are stored as separate objects that don't touch the original page content. Almost every PDF editor handles this well.

Form Filling

Interactive PDF forms have fields designed to be filled in: checkboxes, text inputs, dropdowns. Works reliably because the form fields were built for it. Not every PDF with blank lines and boxes actually has real form fields, though — a scanned application or a flat contract just draws the blanks visually, with nothing behind them for a viewer to click into. A tool has to tell the two apart: one is a form to fill, the other is a page to annotate with a typed-in text box that only looks like a filled-in field.

Page Manipulation

Reordering, deleting, rotating, extracting pages. Operates at the page level without parsing individual content. Widely supported.

Text Editing

Modifying existing text in a PDF. The hardest type by far. It requires parsing content streams, handling font subsets, recalculating glyph positions. Most free tools can't do this. They overlay new text boxes instead, which looks similar but isn't the same thing — the original characters are still sitting in the content stream underneath, covered rather than deleted, and a paragraph that was justified or kerned to a specific font's metrics rarely lines back up once new characters of a different width are dropped in. Even Adobe Acrobat's own "Edit Text & Images" mode, working against genuinely parsed content, visibly reflows the rest of a paragraph when you lengthen a sentence in a justified column. The hard part isn't unique to free tools — free tools just don't attempt it at all.

Digital Signatures

Two different things share this name. Electronic signatures (drawing or typing your name) are annotations, and their legal standing in the US comes from the ESIGN Act (15 U.S.C. Chapter 96), which says a signature can't be denied legal effect solely because it's in electronic form; the EU's equivalent is the eIDAS Regulation. Cryptographic digital signatures are a different mechanism entirely — a certificate-backed hash that validates document integrity and requires specialized software to create and verify. Free tools handle the first type only.

Redaction

Permanently removing sensitive content. Proper redaction strips data from the content stream. Drawing a black box over text does nothing — the string is still there for anyone who copies the "hidden" text or opens the file in a tool that doesn't render the drawn rectangle. A 2021 study by researchers at Université Grenoble Alpes and Inria examined nearly 40,000 PDFs published by 75 government security agencies and found that even among the small number that had attempted sanitization at all, 65% still contained the sensitive information underneath. Bad redaction creates a false sense of security, which may actually be worse than leaving the text visible.

The Security Problem

Online PDF editors upload your file to a server. If that file contains anything sensitive, you have a problem.

One incident worth knowing: in April 2026, Adobe patched CVE-2026-34621, a prototype-pollution zero-day in Acrobat and Reader that let a maliciously crafted PDF run code on the machine that opened it, simply by opening the file, no macro or plugin required. Exploitation was traced back to around November 2025, roughly four months before the patch shipped. It's a reminder that the risk in this space isn't hypothetical, and it cuts both ways: a PDF you receive can attack the device you open it on, and a PDF you upload can expose its contents to whatever happens on the other end of that upload.

IBM's 2025 Cost of a Data Breach Report put the global average cost of a breach at $4.44 million, down 9% from $4.88 million the year before. That figure covers detection, containment, notification, and lost business, not just the direct cost of whatever data leaked, and it's the number every "we take your privacy seriously" banner on a free upload-based tool is quietly betting won't apply to them.

Every major online PDF editor (Smallpdf, iLovePDF, Sejda, PDFescape) uploads your file. They claim files are deleted within 1-2 hours. You can't verify that. CDN caches, backup systems, and server logs can all retain data well past the stated window — and if the document ever becomes relevant to litigation, regulatory discovery, or a breach investigation, "we deleted it" doesn't undo the fact that it left your control and passed through a third party's infrastructure in the first place. For anything covered by HIPAA, attorney-client privilege, or a client confidentiality agreement, that's a decision worth making deliberately rather than by default because a search result happened to rank first.

What Free PDF Editors Can't Do

Here's what you give up compared to paid software:

Free online tools also impose usage limits that shift over time. At last check, Smallpdf's free tier caps how many conversions you get per day, Sejda's free plan caps tasks per hour and pages per document, and PDFescape's online editor is free with no account but reserves genuine text/image editing — versus just filling and annotating — for its paid Premium tier. Vendors adjust these numbers often enough that quoting an exact figure here would likely be stale within a year; check the linked pricing page for whatever the current caps are before you build a workflow around one.

Editing with OxygenPDF

OxygenPDF runs the most common editing tasks in your browser:

1. Open the Edit PDF tool
2. Drop your file in
3. Add text, annotations, highlights, or signatures
4. Rearrange or delete pages
5. Download the result

Everything runs locally. Your PDF stays on your device. No account, no task limits, and no file-size cap beyond your device's own memory — a file over 50MB shows a dismissible prompt about the paid tier, but it doesn't block the upload or the edit.

What You Can Actually Draw and Place

The editor's tool palette covers text boxes, images, freehand drawing, basic shapes (rectangle, circle, line, arrow), highlights, a whiteout rectangle, stamps, sticky notes, callouts, and signatures, alongside the click-to-edit text tool covered below. Annotate PDF runs the same editor with a narrower toolbar: it locks out the text-editing, image-insertion, and signature tools on purpose, so a page meant for markup only can't accidentally end up with its own words changed. If you specifically want to mark up a document with zero chance of touching its content, that's the more deliberate entry point; if you need the full set, use Edit PDF directly.

How Text Editing Actually Works

Click-to-edit text in OxygenPDF works the way the "Text Editing" section above describes for every free tool: it detects the text already on the page, samples the color behind it, draws a matching rectangle over the original characters, and places your replacement text on top at a matched size. It's a well-executed version of the overlay approach — visually seamless for a short replacement in an unjustified line — not a rewrite of the PDF's actual content stream. The original glyphs are still there, just covered, which matters if you're editing a document you'll later need to redact: don't assume a word you've edited over is gone from a text search or a copy-paste of the file.

Redaction: Search or Whole Page, Not a Freehand Box

Redact PDF works differently from the general editor's whiteout rectangle, and this is the important distinction carried over from the section above. Whiteout is a shape annotation — it covers the page visually but leaves the original content stream untouched underneath it. Redact PDF instead searches the document's text for a term you enter, or lets you select whole pages, then renders every affected page to an image and rebuilds the PDF from those images. The result genuinely has no text underneath the black box, because after that step the page isn't text-plus-a-drawn-shape anymore — it's a picture. The trade-off is that it currently offers two modes only: search a term across the document, or blank an entire page. There's no click-and-drag box over an arbitrary region yet. For a scanned document with no text layer to search, or a sensitive detail that appears as part of an image rather than as text, the accurate option today is redacting the whole page rather than just the corner with the detail in it — OCR the page first if the rest of its text is worth keeping searchable afterward.

Password-Protected Files

Edit PDF opens password-protected files directly — enter the password when prompted, and everything from there works the same as it would on an unlocked file. One edge case worth knowing: decrypting a protected PDF for editing renders every page to an image first, the same approach the app uses anywhere it needs guaranteed-clean bytes to hand to its editing engine. Annotating, signing, and reordering pages behave identically over that raster. Text editing doesn't: there's no text layer left on a page that started out encrypted, so the click-to-edit tool has nothing to detect and nothing to click. Redact PDF doesn't run into this — it hands the password straight to its own text search first, finds real matches against the genuine text, and only rasterizes as the final step that makes the redaction permanent.

Signatures

Edit PDF's own signature tool draws, types, or imports a signature and saves it in your browser for reuse across sessions — nothing leaves your device to store it. For signing as the main task rather than one step inside a bigger edit, Sign PDF is the dedicated version of the same tool, and filling an interactive form that has real form fields, rather than a flat page you're annotating text boxes onto, is better handled by PDF to Form, which converts genuine fillable fields into a web form instead of asking you to eyeball where a text box belongs.

For page-level work on its own, the dedicated Organize PDF tool handles reordering, deleting, and rotating without opening the full editor.

Choosing the Right Approach

Annotations, signatures, form filling: Almost any tool works. OxygenPDF does this locally. macOS Preview handles it natively.

Text editing: If you need to change existing text in a PDF, you're looking at Adobe Acrobat ($19.99/mo), Foxit ($14.99/mo), or Sejda's limited free tier. No free tool does this reliably — and as covered above, that includes any file that was ever password-protected, where the text layer is usually gone by the time an editor can reach it, regardless of which tool you use.

Redaction: Use Adobe Acrobat or OxygenPDF's Redact PDF tool, and pick whole-page redaction over the search box for anything scanned or image-based. Never use a tool that draws a black box over text and calls it done — that's what the whiteout tool is for, and it isn't redaction.

Scanned PDFs: OCR first, then edit. OxygenPDF's OCR PDF tool runs locally.

Privacy-sensitive documents: If the file contains financial, legal, or medical information, use a client-side tool. Don't upload it to a server you don't control, and remember that "we delete it in an hour" is a policy statement, not something you can verify from outside the company running the server.

Frequently Asked Questions

Can you actually edit the text of a PDF for free?

Sometimes, and only in a limited way. Free tools, OxygenPDF included, detect the existing text, cover it, and place new text on top rather than rewriting the underlying content stream — fine for a short correction in a plain, left-aligned paragraph, unreliable for anything justified, tightly kerned, or set in a font the original document didn't fully embed. If you need the file's actual internal text stream rewritten, the way a word processor rewrites a document, that's a paid-software job today, not a free-tool one.

Is drawing a black box over text in a PDF the same as redacting it?

No. The box is a shape object layered over the page; the characters stay in the content stream underneath, so select-all, copy-paste, or any text extractor pulls them straight back out. Real redaction deletes the text itself — OxygenPDF's Redact PDF does it by rendering the affected page to an image with the sensitive area erased, then rebuilding the file so there's no text left to recover.

Can free online PDF editors handle scanned documents?

For annotating, signing, and reordering, yes — none of that depends on the page having real text underneath it. For anything that depends on reading the page's text — search-based redaction, text editing, copy-paste, screen readers — no, because a scan is a photograph of a page with no text layer at all. Run OCR first if you need the words themselves to exist as text rather than as pixels.

What's the difference between annotating a PDF and editing one?

Annotating adds objects on top of a page — highlights, sticky notes, drawn shapes, stamps — without touching what was already there. Editing changes the page's own content: its text, its images, its structure. OxygenPDF splits these into two entry points that share one engine: Annotate PDF locks the toolbar to markup-only tools, and Edit PDF opens the full set, including text and image editing.

Is an electronic signature on a PDF legally binding?

In the US, generally yes, for most contracts — the ESIGN Act says a signature can't be denied legal effect solely because it's electronic, and the EU's eIDAS Regulation does the equivalent across the EU. That covers a drawn or typed signature image — what this article calls an electronic signature. It doesn't make it a cryptographic digital signature with certificate-backed integrity verification, which some specific document types, including certain court filings and government submissions, require by name.

Can I edit a PDF that has a password on it?

Yes, for annotating, signing, and reordering pages — enter the password when the tool asks for it. Text editing is the exception: decrypting a protected file for editing typically renders every page to an image first, in OxygenPDF and in most browser-based tools, which removes the text layer that click-to-edit text depends on. If you specifically need to edit the words on a page that's ever been password-protected, that's a harder case than an unprotected file, no matter which editor you reach for.

Why does my "edited" PDF look slightly different after I save it?

If you used the click-to-edit text tool, the new text is placed at a matched font size over a sampled background color rather than flowing through the original font's exact metrics — a small difference in letter spacing, or a subtly different shade behind the new characters, is the visible sign of that. It's the trade-off of the overlay approach every free tool uses. It doesn't affect annotations, signatures, or page-level changes, since none of those touch existing content at all.

Where This Leaves You

PDF editing has real limits. The format wasn't designed for it, and no tool fully overcomes that. Annotations, signatures, form filling, and page reordering all work fine in free tools. Rewriting paragraphs, rebuilding layouts, or genuinely stripping sensitive content requires either paid software or a tool built specifically for that one job — and even then, a file that was ever password-protected changes what's possible.

Match the tool to the task, not to the marketing page.

Edit your PDF locally with OxygenPDF.