A price changed on a menu you only have as a photo. A date on a screenshot is wrong. A sign in your own artwork has a typo and the layered file is long gone. Changing words that are part of a flat image normally means opening a heavy editor and painting over them by hand.
This editor finds the words for you, clears them, and lets you type new ones, all in the browser tab. It works best on text over plain or smoothly shaded backgrounds, and it says so, rather than promising a flawless result on every picture.
Drag & drop an image here
or click to choose a file
Accepts JPG, PNG and WebP up to 30 MB (40 megapixels). The image opens in your browser and is not uploaded.
Select text to edit
Click a dashed outline on the page. Use Add text to place a new text box, or Repair area to wipe a rectangle that has no detected text.
Click on the page where the new text should go.
Drag a rectangle over the area to clean up.
The outlined box is the area that gets cleared. Drag it to move, or drag the corner dot to resize.
The edited image keeps its original pixel size. JPG and WebP files are re-saved once at high quality because the changes are pixel edits.
Your edited file is ready ()
Overview
About Image Text Editor
Open an image and the editor reads it with OCR (optical character recognition) to find the words in it. Each group of words gets a dashed outline. Click one and its text appears in the panel next to the picture: rewrite it, change the size, font style, colour or alignment, or delete it. The preview updates while you work, and Undo and Redo cover every step.
This is not an AI image generator. There is no model that repaints the picture. The editor clears the old letters by filling that area from the pixels around it, then draws your new text on top with a standard font. Because it is a plain calculation you can see and adjust, you decide whether the result is good enough: move or resize the cleared area, wipe a spot by hand with Repair area, or undo and try again.
The edited image is created in your browser at the same pixel size as the original. If you open a PDF instead, use the PDF Text Editor.
Compatibility
Supported Formats
Accepts JPG/JPEG, PNG and WebP up to 30 MB and 40 megapixels. The file type is checked from the file itself, not just its name. HEIC, TIFF, GIF, BMP and SVG are not supported here; convert them first with HEIC to JPG or PNG to JPG.
The output is in the same format as the input and keeps the original width and height. PNG output is lossless and keeps transparency. JPG and WebP output is saved once at high quality (95), because the edit changes pixels. Text recognition reads English. Replacement text can be set in sans-serif, serif or monospace, in regular, bold or italic. You can also download the edited picture wrapped in a one-page PDF.
Guide
How to Use Image Text Editor
- Choose or drag in a JPG, PNG or WebP image. It opens straight away; nothing is uploaded.
- Wait a few seconds while the text is detected, then click a dashed outline. A first run also loads the recognition engine, so it takes a little longer.
- Edit the text in the panel, then press Apply edit. Delete text removes it. If a word was misread, Re-read this region with OCR tries that spot again.
- Press Add text and click the picture to place new text, or Repair area and drag a rectangle to clean up a spot.
- Check the preview and zoom in on the edges of what you changed. Resize the outlined area if the old text shows through.
- Press Apply Changes, then Download. The file is named after yours with "-edited" added.
Highlights
Features
- Finds the text for you OCR runs in a background worker, groups words into editable blocks, shows a confidence figure for each one, and can re-read a single region at a larger size when a word comes out wrong.
- Not a generative tool Old text is cleared by filling from the surrounding pixels and new text is drawn with a regular font. Nothing is invented, and you can see and adjust the result before downloading.
- Manual fallback for hard cases Move or resize the cleared area, wipe any rectangle with Repair area, or place a new text box by hand where recognition found nothing.
- Same size, format and transparency The edited image keeps its original width and height and its format. PNG stays lossless and transparent areas stay transparent. Your file never leaves the browser.
Applications
Use Cases
- Updating a price, date or opening time on a menu, poster or social graphic you made
- Fixing a typo in a screenshot or an exported slide when the original is no longer available
- Replacing placeholder text in a mock-up or sample image
- Removing a name, number or caption from your own photo before you share it
- Adding a label or a short note to an image
- Translating a short line on your own graphic by retyping it
Edit only images you own or have permission to change. An edited image is a modified copy and should not be passed off as the original, especially for receipts, IDs or other documents.
Good to Know
Limitations
Detection is not perfect. Text recognition can miss small, curved, stylised or low-contrast text and can misread letters. Each region shows an OCR confidence figure, amber when it is low. Regions it misses can be handled with Add text and Repair area.
Complex backgrounds may need manual adjustment. Over a plain colour or a smooth gradient the old text is removed cleanly. Over a detailed photo, a texture or a pattern, the filled patch can look smooth or smudged compared to its surroundings. Look at the preview at full size, adjust the cleared area, or use Repair area. Some pictures will not come out looking untouched.
Fonts do not match exactly. The editor measures the size and colour of the original text and fits the new text to the same width, but it only draws with sans-serif, serif or monospace faces.
Metadata such as EXIF or GPS data and an embedded colour profile are not carried into the edited image, and a photo's orientation flag is applied before you start. Animated images are not supported.
This tool runs entirely in your browser — your file is never uploaded to our servers. See our Privacy Policy for full details.
Support
Frequently Asked Questions
Is my image uploaded anywhere?
No. The image is opened, analysed, edited and exported in your browser. Text recognition runs in a background worker on your device, and its engine and language data are served from this site rather than a third-party CDN. In our tests no request carried the file, its name or the recognised text. Closing the file releases it from memory.
Does this use AI to redraw the text?
No. It uses OCR to find text, a diffusion fill (smoothly continuing the colours around the text) to clear it, and a normal font to draw the new words. That makes the result predictable, and it also means it cannot invent detail: on a busy background the cleared patch will look softer than its surroundings.
Why does the edited area look blurry or smudged?
The background behind the old text has to be filled in from the pixels around it. On flat colours and gradients that works well. On photos, textures and patterns there is no way to know what was underneath, so the fill is a smooth estimate. Zoom in, resize the outlined area to cover only the letters, or use Repair area. Some images will always show the edit.
Does editing reduce image quality?
PNG output is lossless. JPG and WebP files are saved once at quality 95, so there is a small re-compression step on top of whatever the original had. The width and height always stay the same, and transparency in a PNG or WebP is kept.
The text recognition got a word wrong. What can I do?
Just retype it in the text box, the original is replaced either way. For a stubborn line, select it and use "Re-read this region with OCR", which runs recognition on only that area at a larger size. Recognition works on English text and does best on clear, horizontal, high-contrast lettering.
Can I edit text that OCR did not find?
Yes. Use Repair area to wipe a rectangle that has leftover text in it, then Add text to type over it. You can drag the new text into place and set its size, colour and alignment.
Does Google Analytics see my image?
This site uses Google Analytics for ordinary page-view statistics, like every other page here. The editor does not pass it your file, its name or any recognised text, and in our test the analytics requests contained none of them.
Which format should I download?
The download keeps your original format. Keep PNG for screenshots and graphics with sharp text, since it stays lossless. For photographs JPG or WebP is fine. If you need a document, "Download as PDF" wraps the edited picture in a one-page PDF. The <a href="/jpg-vs-png-vs-webp/">JPG vs PNG vs WebP guide</a> explains the trade-offs.
Keep Going
Related Tools
- Image Cropper Keep one rectangle of a JPG, PNG or WebP. Drag a selection or type exact X, Y, width and height in pixels.
- Image Metadata Remover Remove EXIF, GPS, XMP, IPTC and embedded comments from a JPG, PNG, WebP or GIF without re-encoding the picture.
- PDF Text Editor Click a line of text in your PDF, change or delete it, or add new text, then download the result. It all happens in your browser.
- Face & License Plate Blur Find faces in a photo, add boxes by hand for plates and anything else, then pixelate, blur or black them out. Everything stays in your browser.
- Bangla OCR: Image to Bangla Text Read Bangla text from a photo or scan, in your browser. Copy or download the result as a text file, with a confidence estimate and tips for better results.