Add Watermark to Image
Protect your images with custom text watermarks. Adjust opacity, size, and position.
Drag & drop image here
or click to select file
About Watermark Image Tool
Burn a line of text into a photograph — a copyright line, a studio name, a URL, or the word PROOF. The image is drawn onto a canvas at full resolution, the text is painted on top in bold Arial at the size, colour and opacity you set, and the whole canvas is encoded back to a file. The preview on screen is that canvas, just displayed smaller, so there is no gap between what you approve and what you download.
Size is measured in image pixels
This trips people up more than anything else here. The size slider, which runs from 10 to 200, sets the text height in the coordinate space of the image itself — not of your screen. On a 1000 px web graphic, 40 px of text is a bold caption. On a 6000 px camera export, the same 40 px is a whisper you can barely find. Judge it against the preview and expect to push the slider high on large files. The upside of working this way is that the mark keeps its proportion to the picture wherever the picture ends up.
Placement changes what the mark is for
Center lays the text diagonally at 45 degrees across the middle of the frame. It sits over the subject, so it cannot be cropped off and is awkward to paint out — the right choice for client proofs, portfolio previews and anything you are showing but not releasing. Bottom Right and Top Left draw a small horizontal caption near a corner, which reads as attribution rather than protection and stays out of the composition. Pair either with a lower opacity; something around 30-50% white is legible on most photographs without competing with them.
What a watermark can and cannot do
It establishes authorship and makes casual reuse inconvenient. It does not make an image theft-proof: a corner caption can be cropped in seconds and a faint one removed with a content-aware fill. If deterrence genuinely matters, use the diagonal centre placement at a substantial opacity and publish at a lower resolution — an image that is not useful at full size is not worth stealing. Keep your unwatermarked master, because the mark is burned into the pixels and cannot be lifted back out.
The download keeps your source format: a JPEG comes back as a JPEG, a PNG as a PNG, named watermarked- plus the original filename. A JPEG source picks up one extra generation of compression across the whole frame, which is invisible in practice but is a reason to work from the original rather than a file you already watermarked once. Everything runs in this browser tab, so unreleased photographs are never uploaded anywhere to have a copyright line added to them.
Preview is the real output
The canvas you see is the finished image at full resolution, just displayed smaller. What you download is that exact canvas — there is no separate render step that can differ from the preview.
Four controls that matter
Text, size in pixels, opacity from 10% to 100%, and colour from a full picker. Enough to make a mark legible on a dark photo or barely-there on a light one.
Three placements
A diagonal mark across the centre for maximum nuisance value to a thief, or a discreet caption in the bottom-right or top-left corner for attribution.
Original format preserved
A JPEG comes back as a JPEG and a PNG as a PNG, saved as watermarked- plus your filename, so nothing downstream needs to know the file was touched.
No upload, no queue
The text is drawn onto a canvas in this tab. Unpublished photographs never leave your machine, and there is no per-file limit because there is no server paying for it.
Frequently Asked Questions
My watermark is tiny on a big photo. Why?
The size slider is measured in image pixels, not screen pixels. On a 6000 px-wide camera export, 40 px of text is about half a percent of the width — genuinely almost invisible. Push the slider up towards its 200 px maximum for large photos. The reason it works this way is that a mark defined in image pixels stays the same relative size no matter how the picture is later displayed.
Can I use my logo instead of text?
No — this tool draws text only. There is no image overlay. If you need a logo mark, composite it in an image editor, or export your logo as text-like artwork and use a bold short string here as a stand-in. Only the typed characters are rendered.
Which position should I choose?
Center draws the text diagonally at 45 degrees across the middle of the frame. It is hard to crop out and hard to clone away, which is what you want on a proof or a preview being sent to a client. Bottom Right and Top Left place a small horizontal caption near a corner — the right choice for attribution on something you are happy to have shared.
What opacity actually works?
For attribution, 30-50% white on a photograph is usually readable without dominating. For a proof you do not want reused, 60-80% across the centre is more effective. Below about 20% the mark is easy to remove with a content-aware fill, so it is decorative rather than protective.
Can I change the typeface?
Not currently. The mark is drawn in bold Arial at whatever size you set, which renders consistently across Windows, macOS, Linux and mobile browsers. That consistency is the reason it is fixed — a font that is not installed on the visitor's machine would silently substitute and change the result.
Does watermarking degrade the photo?
The pixels under the text are permanently changed — that is the whole idea. Everything else depends on format. A PNG is re-encoded losslessly and is pixel-identical outside the mark. A JPEG is re-encoded, which means one extra generation of compression on the whole image; at normal quality that is not visible, but always keep your unwatermarked original.
Does a watermark actually stop image theft?
It raises the cost and it establishes attribution, which is the realistic goal. It will not defeat someone with a modern editor and ten minutes. A diagonal mark across the subject is far harder to remove than a corner caption, and a lower-resolution export makes reuse less attractive in the first place.
Can I watermark a whole folder at once?
Not on this page — it handles one image at a time so you can position and size each mark against the actual picture. For batch work where the images all share dimensions, watermark one, confirm the settings look right, and repeat; the controls keep their values between files in the same session.
Is the photo uploaded to add the watermark?
No. The image is decoded into a canvas in this browser tab, the text is drawn on top, and the canvas is encoded straight back to a download. Nothing is transmitted, which is exactly what you want when the pictures being watermarked are unreleased client work.