/TOOL · BLUR TEXT
Type something, pick what should happen to it, and watch the before and after change as you go. Blur it into an image you can download, hide it behind a spoiler your reader taps to open, cover it with a censor character, or shuffle the letters. It all happens inside this page, so nothing you type is uploaded anywhere.
Choose what happens to your text
Five different jobs. Pick the one that matches why you are hiding this.
Blur and pixelate can be reversed. Both of them average the pixels that were underneath rather than removing them, and text has been recovered from both in practice. The solid block is the only cover here that genuinely deletes what it sits on, which is why it is the default. Use blur or pixelate when the point is that a field exists rather than what it says. For a password, a card number, an address, a token, a face or anything else that would hurt to leak, use the solid block.
Type your text in, then take the result
Before is the box you type in. After is exactly what other people will get, and it redraws on every keystroke.
After what other people get
Drop a screenshot here
Or paste one straight from the clipboard with Cmd+V on a Mac, Ctrl+V on Windows. That is usually the fastest route: take the screenshot, click this page, paste.
The file is opened by this page and stays in your browser. It is not uploaded, and no server sees it.
Drag across anything that should not be seen. Click a box to select it, drag its middle to move it, drag a corner handle to resize it, and use the cross above it or the Delete key to remove it.
Click the covered block to reveal it.
/WHICH MODE
People search for blurred text meaning five completely different things, and picking the wrong one is how you end up with a result that either leaks the content or destroys something you needed to keep. Here is what each mode actually does to your characters.
Takes a picture you already have, rather than text you type, and covers parts of it. Drop the file in, drag it in, or paste it out of the clipboard, then drag a box over each thing that should not be seen. The box can be a solid block, a blur or a pixelation, and you can move it, resize it by its corners or delete it after the fact. This is the one for the job most people actually arrive with: a screenshot of an inbox, a dashboard, a bank statement or a chat, which has to be shared with one detail taken out of it. Use the solid block unless you have a specific reason not to. It is the only cover of the three that deletes what was underneath, and the exported PNG carries pixels only, so there is no layer to switch off and no text left to select.
Draws the real text onto a canvas and then smears the pixels, either with a gaussian blur or with pixelation. The output is an image, not text, so there is nothing to select and nothing to copy out of it. This is the mode behind the phrase blur text image, and it is the right choice when you want a screenshot, a mockup, a thumbnail or a slide where some text is present but deliberately unreadable. It is the wrong choice when the underlying content is a secret, because the characters were drawn before the blur landed on them.
Keeps the text intact but puts it behind a deliberate act of revealing. In plain text that means block characters standing in for the letters. On Discord and Reddit it means the platform's own spoiler markup, which renders as a covered block the reader taps to open. Reach for this when you are protecting someone's experience rather than their data: a plot twist, a puzzle answer, a result nobody asked to see yet. Anyone who wants the content will get it, and that is the point.
Replaces characters with a mask character before anything is rendered, so the original never exists in the output at all. This is the only mode here that removes information rather than disguising it. The three scopes cover the real jobs: mask everything, mask digits only for card numbers, phone numbers and dates of birth, or keep the first and last letter of each word so a colleague can still follow the shape of a sentence while the specifics are gone. If you are preparing a screenshot for a bug report or a support ticket, censor first and screenshot second.
Reorders the characters and keeps every one of them, so the length, the punctuation and the general texture of the text survive while the words do not. That makes it the mode for filler: placeholder copy that has to occupy the same amount of room as the real thing, a demo dataset that needs to look plausible, a design review where the layout matters and the content does not. It is obfuscation only. The original letters are all still there, and a short scrambled word is an anagram anyone can solve.
/SCREENSHOTS
Most people looking for a way to blur text are not looking for a picture of blurred words. They have a screenshot with one line in it that cannot be shared, and they need the rest of the picture to survive. That is what the Screenshot mode above does, and there are four things worth knowing before you use it on something real.
Paste is the fast path. Take the screenshot the way you normally do, click anywhere on this page, and press Cmd+V on a Mac or Ctrl+V on Windows. The image goes straight from the clipboard onto the canvas without ever becoming a file on disk. Dragging a file in and picking one with the button both work as well, and all three routes end up in the same place: your own browser, holding your own picture.
Cover more than you think you need to. The preview is scaled down to fit the panel, so a box that looks like it swallows a whole line at preview size can leave the top of the capital letters or the descenders of a g and a y poking out at full resolution. Overshoot the box on every side, then open the downloaded PNG at 100 percent and look along the edges. Partial characters are worth more to somebody guessing than most people assume, because a redacted string with two visible letters and a known length is a much smaller puzzle than a redacted string alone.
The solid block is the only one that removes anything. Blur and pixelate are averages of the pixels that were there, which is a different thing from a deletion, and both have been reversed for text. This page offers them because they have honest uses: showing that a column of customer names exists without showing the names, keeping a screenshot readable as a layout, softening a background face in a product shot. If the covered thing is a password, an API key, a card number, an address, a licence plate or a real person's contact details, use the black box. It costs you nothing and it cannot be undone.
Redacting the picture is not the same as redacting the file. A PNG exported from here contains flattened pixels and nothing else, which is exactly what you want. Be more careful when you black something out inside a design tool, a PDF or a document, because the shape you drew usually sits on a layer above the original and the original is still in the file for anyone who opens it properly. Check the metadata as well: a photo or a screen capture can carry device, software and sometimes location details that survive any amount of drawing over the picture.
/THE HONEST PART
This is the thing most blur tools will not tell you, and it is worth two minutes of your attention before you post a screenshot of anything.
Blurring is a convolution. Each output pixel is a weighted average of its neighbours, and averaging throws away detail without throwing away all of it. Enough structure survives that the operation can be attacked, and text is the softest target imaginable. There are only so many characters, they are drawn in a font you can usually identify from the rest of the screenshot, at a size you can measure off the same screenshot. An attacker renders candidate strings, blurs them with the same radius, and compares. The match falls out.
Pixelation is worse, not better. It is a smaller amount of averaging over a rigid grid, which means each block is a compact clue about what sat underneath it, and public tooling for reversing pixelated text has existed for years. The intuition that a chunky mosaic looks more destroyed than a soft blur is exactly backwards.
So the rule is simple. If it would be bad for someone to read it, do not blur it. Remove it. Censor the characters before you take the screenshot, or draw a solid opaque rectangle in one flat colour over the area and then flatten and re-export the image, so the covered pixels are genuinely gone rather than sitting on a layer somebody can toggle off. Watch for two related traps: a PDF or a design file that keeps the original text under the black box, and image metadata that carries device, software and sometimes location details that no amount of drawing over the picture will touch.
Blur has a legitimate home, and it is a big one. Product screenshots where real customer names appear and nobody cares what they were. Mockups where the copy is not written yet. Thumbnails where a name would be distracting. Course material and documentation where the point is that a field exists rather than what it contains. For all of those, blur is exactly right, and the one on this page is drawn onto a canvas and exported as pixels, so at least nobody can select the text back out of it the way they can on any page that blurs with a CSS filter.
/SPOILER SYNTAX
Both platforms hide text behind a tap, and both use their own markers. The tool above wraps your text for either one, but it is worth knowing the syntax so you can type it directly.
DISCORD ||your text here||
Two vertical bars on each side, no spaces inside the bars.
Works in messages, in embeds, and inside a code block only
if the bars are outside the fence. Ctrl+Shift+S wraps the
current selection in the message box for you.
REDDIT >!your text here!<
Greater-than plus exclamation mark to open, exclamation
mark plus less-than to close. No space between a marker
and the text, or it renders as literal characters.
PLAIN ████ ████
Block characters, for anywhere with no spoiler support at
all. Nothing is hidden, the text is simply replaced, so
only use it where the content is disposable.
A few things that catch people out. Spoiler markup is a rendering instruction, not encryption, so the hidden text sits in the message payload in plain form and anything that reads the raw message sees it. Notification previews and link embeds have both been known to leak spoilered content. And the syntax does not travel: paste Reddit markup into Discord and readers see the literal markers, which is why the tool has a separate output for each.
/HOW TO
Type or paste into the box, or paste a screenshot straight from the clipboard. Nothing leaves your browser.
Screenshot covers parts of an image, Blur makes an image out of typed text, Spoiler hides text behind a reveal, Censor masks the characters, Scramble reorders them.
Drag the amount slider, set the text size and background, or choose a mask character and scope.
Copy the text, copy the image, or download a PNG or a .txt file. Every button confirms on itself when it has worked.
/COMMON MISTAKES
A light blur leaves word shapes, letter heights and capitalisation intact, and those are enough for a reader to guess short strings and for software to recover longer ones. Half-blurred text is the worst of both worlds: unreadable to your audience, recoverable by anyone motivated.
FixRaise the radius until the line is a smooth band with no visible letter boundaries, then take one more step.
A CSS blur on a live page leaves the characters in the document. They are selectable, copyable, present in the page source and visible to anything that reads the DOM. The same trap applies to a shape drawn over text in a PDF or a design file, where the original stays underneath.
FixExport to flat pixels, as this page does, or delete the text rather than covering it.
Leaving the last four digits of a card is a convention because those four alone identify very little. Leaving the first six is a different story, since they identify the issuer and the card range, and a partly masked date of birth is often enough to finish off from context.
FixUse the Digits only scope to take out every digit, then put back only what the reader genuinely needs.
Spoiler markup is a courtesy to the reader, not a control. The text sits in the message in plain form, moderators and any client that reads the raw payload can see it, and previews sometimes render it anyway.
FixUse spoilers for plot twists and answers. For anything you would not want quoted back, do not post it.
/FAQS
/DO MORE WITH ASCII MAGIC
The studio behind this page turns images and video into ASCII, pixel art, dither, halftone, voxel and glitch. No signup, no upload, export PNG, GIF or MP4.
Open the editor →/RELATED
This page hides text. To pixelate or stylise a whole image, or turn a photo into characters, open the full studio.