Image Watermark
Mark a photo with your name or your logo, positioned and faded so it is visible but does not spoil the picture.
1. Choose an image
Both the photo and the watermark are processed in this tab. Neither is uploaded.
Choose the photo to watermark
You can also paste an image with Ctrl or Command and V
JPG, PNG, or WebP
Choose a watermark image
A logo or signature, ideally with a transparent background
PNG with transparency works best
Choose an image to begin. Nothing is uploaded.
2. Preview
The watermarked image will appear here.
A visible watermark is the one that works. A mark set to 5 percent opacity on a busy photo may be invisible to everyone, including whoever you are trying to discourage. Around 30 to 50 percent in a corner is the usual compromise. If you need to guarantee ownership, keep the unwatermarked original yourself.
Placing a watermark
Keep it out of the middle
A corner survives cropping better than a centre mark, and it leaves the subject of the photo visible.
Use a logo with transparency
A PNG with a transparent background sits over the photo properly. A white rectangle around your logo will not.
Check it small
Zoom the preview out to a quarter size. If the mark is still legible there, it will survive being posted at a thumbnail size.
Text or a picture, and which one to use
A text watermark carries a name, a site, or a copyright line and needs nothing but this tool. A picture watermark takes a logo you have already chosen and lays it over the photograph, which looks more finished and costs you a step. Text is the better choice when the point is attribution; a logo is the better choice when the picture is going somewhere your name is not welcome.
Placement is measured from the image, not guessed
The positions are computed as real fractions of the picture rather than fixed pixel offsets, so the same setting puts the watermark in the same relative place on a phone snapshot and on a large camera file. That is what makes the result usable on more than one picture: you set the position once and it does not creep into a corner on the next, differently shaped image.
Tiling resists being cropped away
A single mark in a corner is one crop away from gone. Tiling repeats the watermark across the whole picture, so removing any one of them means either taking a large piece of the image with it or leaving a visible patch. For a picture you are genuinely worried about, tiling is the stronger choice, and at a low opacity it is less obtrusive than it sounds.
Adjusting the sliders is safe to repeat
Every change redraws the image from the original rather than painting onto what is already there, so dragging the opacity up and down does not stack fifty copies of the mark on top of each other. It is also why the size and position controls stay predictable when you go back and forth: you are always adjusting one mark, not a pile.
Opacity is the whole decision
Low enough to be unobtrusive is high enough to be removable. A watermark at full strength over a detailed area is frequently cropped out by a platform that crops to a standard shape. Somewhere around the point where you can read it without it fighting the photograph is usually right, and tiling makes a lower opacity workable because there is more than one.
It is drawn on, not embedded
The watermark is rendered into the pixels, and the result is a new file. It is not a separate layer, and it cannot be switched off by opening the file in an editor. The trade is that it is also part of the picture permanently, so keep the unwatermarked original somewhere you control.
Check it on the actual device
View the result at the size it will be seen rather than full screen. A mark that reads clearly on a monitor is often illegible in a phone thumbnail, which is the only place most shared photographs are actually looked at.