Text blur

Blur Text in an Image or Screenshot

Smart blur finds text lines and blurs them for you. Anything it misses can be painted over by hand, all in the browser with no upload.

Blur text now

Redact text before you share a screenshot

Screenshots carry more than people expect: names, email addresses, order numbers, chat messages, and internal labels. Blurring the text is faster than cropping it out and keeps the surrounding context visible.

Smart blur runs text detection as soon as the image loads, draws a box around each detected line, and blurs those boxes. From there you can expand the boxes, change the blur radius, re-run detection, or switch to partial blur and paint over anything the detector missed.

Automatic detection

Detected text lines are blurred individually, with a radius from 4px to 32px.

Adjustable coverage

Padding from 0px to 32px expands each box so characters near the edge are covered.

Manual fallback

Partial blur handles handwriting, logos, and low-resolution text the detector cannot read.

How to blur text in an image

Detection runs on its own; you only tune the result.

  1. 1

    Upload a screenshot or photo

    JPG, PNG, WebP, and GIF files are supported.

    Detection starts automatically once the file loads, so there is no separate scan button to press.

  2. 2

    Switch to smart blur

    Smart blur reads the text in a downscaled copy of the image and maps the detected lines back onto the original.

    The panel reports how many lines were found and how long detection took.

  3. 3

    Tune radius and padding

    Raise the blur radius for heavier redaction and the padding to expand each box.

    Radius runs from 4px to 32px and padding from 0px to 32px, and the preview refreshes after each change.

  4. 4

    Re-run or paint by hand

    Use Re-run detection after switching images, or switch to partial blur and brush over text the detector missed.

    Download the result as a PNG once every sensitive line is covered.

What you control

Line-level detection

Text is returned as lines with bounding boxes, and each box is blurred on its own instead of blurring the whole image.

Adjustable coverage

Padding expands the box so characters near the edge are covered; radius controls how unreadable the text becomes.

Manual fallback

The brush and eraser in partial blur handle handwriting, logos, and text in images the detector cannot read.

Nothing leaves the tab

Detection and blurring both run locally, which matters for screenshots of private messages or documents.

사용 방법

1

왼쪽 업로드 영역에 이미지를 업로드합니다.

2

흐림 모드를 선택하세요. 전체 흐림으로 전체를 수정하거나, 스마트 흐림으로 텍스트를 자동 감지하거나, 부분 흐림으로 직접 그립니다.

3

흐림 설정 또는 브러시 영역을 조정한 뒤 결과를 미리 보고 다운로드하세요.

Text blur questions

Which languages does text detection support?+

The detector is configured for Simplified Chinese and English. Text in other languages, or heavily stylised lettering, is better handled with partial blur and the brush.

Do I have to start detection manually?+

No. Smart blur runs automatically when an image is loaded and re-runs when you change the radius or padding. The Re-run detection button forces a fresh scan.

Why did it miss some text?+

Detection needs reasonably clear, high-contrast text. Small, blurred, rotated, or decorative text may be skipped, and images wider than 1600px are downscaled before detection. Paint over anything missing with partial blur.

Is the blurred text really unreadable?+

A high radius plus padding covers the characters completely. Check the preview at full size before sharing, especially for short strings such as numbers.

Are my screenshots uploaded for detection?+

No. The text recognition model runs in the browser and the image never leaves the device.

What file do I get back?+

A PNG export of the processed image, named after the original with a _blurred suffix.

Hide the text before you hit share

Upload a screenshot, let smart blur find the text, and download a redacted PNG.

Open the blur tool