Add a watermark to an image

How to use it

1. Choose an image — PNG, JPG and WebP all work.
2. Type your watermark text and set the size, colour, opacity and position.
3. When the preview looks right, press Apply — the watermarked image downloads in the same format.

A watermark deters reuse; it does not prevent it

Anyone with an image editor can clone a corner mark away in under a minute, and content-aware fill has made that easier still. Treating a watermark as protection sets the wrong expectation. What it does reliably is raise the effort above the threshold of casual reuse, which is where most unauthorised copying happens: a right-click and a repost.

The second function is provenance. A mark carries a name or a copyright line into every copy, so an unattributed version is visibly derived rather than merely similar, which is a useful position to argue from in a takedown or a client dispute. Placement follows from that. A mark tucked into a corner is croppable; one crossing the subject cannot be removed without damaging the part that was worth taking.

Opacity and tiling trade the picture against the mark

Every setting here is a compromise with the image underneath. A near-opaque mark is hard to work around but ruins the photograph for the client review it was made for. A very faint one reads as a fault rather than a signature, and heavy JPEG compression can dull it further until it barely registers. Mid-range opacity across a busy area is usually the workable middle.

Tiling changes the calculation. Repeating the text diagonally across the frame means no single crop removes every instance, which is why proofs, previews and stock comps are marked that way. The cost is deliberate: a tiled image is no longer usable as a picture, which is exactly the point for a proof and exactly wrong for a portfolio piece, where one discreet mark near an edge does the job.

Questions