Drop one photo or fifty into the box and the tool re-draws each of them at the size and format you choose, then shows the result next to the original: new dimensions, file size before and after, and a small preview. You can compress for the web, shrink to fit an email limit, convert between JPG, PNG and WebP, or just remove what a photo says about where and how it was taken. Every file is processed inside your browser, in the background, so the page stays responsive even with a long list.
Because the work happens on your device, the pictures are never sent anywhere. That matters for photos of documents, children, a home or a workplace, which is exactly what people tend to run through online compressors.
How to use it
- Drop images on the box, choose them with the button, or paste a screenshot with Ctrl+V or Cmd+V.
- Pick the output format. "Keep original format" writes JPG as JPG, WebP as WebP and PNG as PNG. GIF and BMP files become PNG, because browsers cannot write those two formats.
- For JPG and WebP, set the quality. Around 75 to 85 is a good starting point for photos; lower numbers give smaller files and more visible blocking.
- Optionally resize: set a width, a height, a percentage, or a box the picture must fit inside. The aspect ratio is always kept, and a picture is never enlarged.
- Check the list. Each row shows the new size, how much smaller it got, and what was removed. Download a single file with its button, or take everything as one ZIP.
Changing a setting re-processes the whole list, so you can move the slider and watch the numbers change.
What happens to the hidden data in your photos
A photo is more than pixels. Cameras and phones write a block of metadata called EXIF into the file: the make and model of the camera, the date and time, the exposure settings, often an embedded thumbnail, and, if location was switched on, the GPS coordinates of where you stood. Editing software can add XMP and IPTC blocks with captions and author names, and a colour profile.
This tool reads those blocks itself and tells you what each file contained. Then it draws the pixels onto a clean canvas and encodes a brand-new file, so none of that is copied across. If a photo carried GPS coordinates, the row shows them in a highlighted label together with the word removed, so you can see what would have been shared. The colour profile is converted to standard sRGB while drawing, which keeps colours correct on ordinary screens.
Phones also store whether a picture was taken sideways in an orientation flag. The tool applies that rotation to the pixels, so the result looks the same everywhere and no flag is needed.
- JPG is the right choice for photographs. It is lossy: each re-save throws away a little detail, so keep your originals and avoid compressing an already compressed JPG several times.
- WebP usually makes a smaller file than JPG at similar quality and also supports transparency. All current browsers display it, but some older programs and upload forms do not accept it.
- PNG keeps every pixel exactly, which suits screenshots, logos and diagrams with sharp edges and flat colour. For photographs it produces very large files, and it has no quality setting.
When a transparent PNG is converted to JPG, the transparent areas are filled with white, because JPG cannot store transparency.
A worked example
A phone photo is 4000 by 3000 pixels, which is 12 million pixels. For a web page that shows it at most 1600 pixels wide, set "Set the width" to 1600. The result is 1600 by 1200, or 1.92 million pixels, which is 16 percent of the original pixel count. With JPG at quality 80 the file typically drops to a small fraction of its old size, and the row shows the exact figure. The same photo requested at 5000 pixels wide stays at 4000 pixels, with a note, because adding pixels cannot add detail.
Limits and what is not supported
You can add up to 50 images at a time, each up to 40 MB and about 50 megapixels. HEIC and HEIF photos from some phones are not supported: switch the phone camera to its most compatible format, or convert them to JPG first. A GIF is read from its first frame only, so animation is lost. Transparency and the sRGB conversion behave as described above.
