Add a watermark to an image
Put your name, a credit line or a PROOF stamp across a photo — in a corner or tiled over the whole frame. Works on a whole batch at once, and the images never leave your device.
The preview is scaled to fit; the file you download is marked at full resolution.
The same settings are applied to every image you dropped in.
Sized against the short edge, not the width
The size slider is a percentage of the image's shorter side. That sounds like a detail and it is the reason this looks right on a mixed batch. If a watermark is sized as a percentage of width, the same setting produces a mark that is visually twice as heavy on a portrait photo as on a landscape one, and a batch of holiday photos comes back inconsistent. Measuring against the short edge keeps the apparent weight the same whichever way round the picture is.
The margin works the same way, at 3.5% of the short edge, so the mark sits the same distance from the corner on a phone snap and on a 6000-pixel raw export. Nothing is measured in fixed pixels, because a fixed pixel size that looks right on a 1000-pixel image is invisible on a 6000-pixel one.
A corner mark or a tiled one
A single mark in a corner is the polite option: it credits the photo without getting in the way of looking at it, and it is what you want on a portfolio or a listing. It is also trivially removed by cropping, which is fine if crediting rather than protecting is the point.
Tiling repeats the text diagonally across the whole frame. It is what proof photos and document scans use, because there is no crop that leaves a clean image behind. The trade-off is obvious: the picture becomes much harder to actually look at, so it suits a proof you are sending for approval rather than something you want admired. When tiling, a lower opacity and a smaller size read better than you would expect — the repetition does the work, not the weight of any single mark.
The angle slider applies to both. For corner marks, keep it at zero unless you have a reason; for tiling, a slight diagonal is the convention, and is what you get by default.
The shadow is not decoration
White text vanishes on a bright sky and black text vanishes on a dark suit. The soft shadow option draws a blurred dark copy behind the text, which means a white watermark stays legible over a pale background without having to raise the opacity to the point where it dominates the image. If you are watermarking a batch of photos you have not individually checked, leave it on — it is the difference between a mark that survives every image and one that disappears on a third of them.
Turn it off when the watermark sits on a flat, controlled background, such as a product shot on white, where a shadow would look like a mistake.
What a watermark does and does not do
It is worth being blunt about this, because a lot of pages selling watermarking software are not. A visible watermark deters casual reuse and establishes a credit. Somebody scrolling past will see whose photo it is; somebody about to drop it into a blog post will think twice; and if the image turns up somewhere it should not, the mark makes the provenance obvious.
What it does not do is prevent copying. A corner mark can be cropped in seconds. A tiled mark can be attacked with content-aware fill or, increasingly, with a generative model that will do a convincing job in one pass. Nothing you can draw on top of an image survives a determined person with modern tools. If the image genuinely must not circulate, the answer is not to publish it at full resolution — publish a smaller version, and keep the original.
A watermark is also not the same as copyright. In most countries you hold copyright in a photograph from the moment you take it, with or without a mark on it, and adding © does not create a right you did not have. What the mark does is make the claim visible and make “I did not know who took it” harder to say. In the United States, removing or altering copyright management information — which a visible credit can qualify as — is separately actionable under section 1202 of the DMCA, which is a real if rarely used lever. None of this is legal advice.
Formats, transparency and quality
Leaving the format on Same as the original gives you a JPG back from a JPG and a PNG back from a PNG, which is usually what you want. Two things are worth knowing about the conversion. JPEG has no transparency, so a PNG with a transparent background saved as JPG gets a white background rather than a black one — that is a deliberate choice here, because black is what you get by accident and it looks broken. And re-saving a JPEG always re-compresses it: the output is encoded at quality 0.92, which is visually clean, but it is a second generation. Watermark from your original file, not from a copy that has already been through this once.
PNG output is lossless and keeps transparency, which matters if you are marking a logo or a cut-out product shot. It is also several times larger for a photograph, so do not use it as a default.
HEIC files from an iPhone can be opened here on browsers that support them, which in practice means recent Safari and recent Chrome on some platforms. If yours cannot, convert first — the output of that conversion is an ordinary JPG or PNG that this page will happily mark.
Doing a whole batch
Drop in as many images as you like; the preview shows the first one, and the settings you land on are applied to all of them. Everything is processed in this page, one after another, and when there is more than one image the results come back as a ZIP with each file keeping its original name and its new extension. There is no upload, no queue and no cap on how many you can do, because there is no server deciding what you are allowed.
Common questions
Does the watermark go on at the original resolution?
Yes. The preview is scaled down to keep the page responsive, but the actual output is drawn at the image's full pixel dimensions — a 6000-pixel photo comes back 6000 pixels wide with the text rendered crisply at that size. The size and margin are proportions of the image, so what you see in the preview is what you get, just larger.
Can I use my own logo instead of text?
Not on this page — it does text watermarks only. A logo watermark needs its own controls for scale and transparency handling, and doing it as an afterthought here would do it badly. Text covers the common cases: a credit line, a copyright notice, a “PROOF” or “DRAFT” stamp, or a website address.
Will a watermark stop people stealing my photos?
No, and it is worth being clear about that. A corner mark can be cropped off in seconds and a tiled mark can be removed by modern editing tools with some effort. What a watermark does is deter casual reuse, make the credit travel with the image, and make it hard for anyone to claim they did not know whose work it was. If an image truly must not circulate, publish it small.
Why does my transparent PNG come back with a white background?
Because you saved it as JPG, and JPEG has no transparency channel at all. Every transparent pixel has to become some colour, and white is chosen here deliberately — the default in a lot of tools is black, which looks like a bug. Choose PNG or WebP as the output format to keep the transparency.
Does re-saving lose quality?
If the output is JPG or WebP, yes, slightly: those are lossy formats and every save is a new generation. The output is encoded at quality 0.92, which is high enough that the difference is not visible in normal viewing, but you should always watermark from your original file rather than from a copy that has already been compressed once. PNG output is lossless.
Is there a limit on how many images I can do at once?
No. Drop in as many as you like — they are processed one after another in this page and come back as a ZIP. The only real constraint is your device's memory, since every image has to be decoded at full size. Very large batches of very large photos are slower on a phone than on a laptop.
Are my photos uploaded?
No. They are read with the browser's own file API, drawn to a canvas in this page, and handed straight back to your downloads folder. Nothing is transmitted, stored or logged, and there is no server to receive it. This is the reason to use an in-browser tool for photographs you have not published yet.
Does it remove EXIF data from the photo?
Yes, as a side effect. Because the image is redrawn on a canvas and re-encoded, the output carries no EXIF at all — no GPS coordinates, no camera serial, no timestamp. That is usually welcome when you are about to publish, but it does mean your copyright and author fields go too. If you want to keep the metadata, watermark first and write the fields back afterwards.