Private by design
image processing stays in your browser. ResizeNest does not need your original image uploaded to an application server.
Image EditorConvert
Convert ImagesJPG to PNGPNG to JPGJPG to WebPWebP to JPGPNG to WebPWebP to PNGJPG to AVIFAVIF to JPGPNG to AVIFAVIF to PNGHEIC to JPGHEIC to PNGSVG to PNGPNG to ICOJPG to ICOImage to PDFPDF to ImageMore tools
Rotate ImagesFlip ImagesImage EnlargerImage EditorAdd a Watermark to ImagesPick Colors from an ImageRemove Personal Metadata from ImagesDocument tools
Image to WordWord to ImageImage tool
Add a text or logo watermark to one image or a local batch. ResizeNest composites the watermark into the actual output pixels in the browser worker, while the preview remains a non-authoritative editing aid.
Start here
No account is required. Choose a file and work directly in this browser.
image processing stays in your browser. ResizeNest does not need your original image uploaded to an application server.
Results come from the actual processing pipeline, with measured dimensions, bytes and format instead of placeholder estimates.
Process one image or a local batch, then download current successful results.
Three simple steps
Choose, drop or paste JPG, PNG, WebP or AVIF base images, then choose a text watermark or load a supported logo.
Set placement, opacity, relative or absolute size, rotation, margins or tiling. Relative sizing keeps a consistent visual scale across portrait and landscape batch images.
Process the current recipe, review the worker-produced result, then download one current result or current successful results as a ZIP without leaving /watermark-image.
Useful details
Understand what the tool does before you download the result.
Text uses ResizeNest's self-hosted Inter font baseline with deterministic glyph handling. Logo mode keeps logo bytes separate from the base image recipe and accepts verified raster logos plus a conservative sanitized SVG subset with no scripts or external resource fetches.
Nine anchors, normalized image-space coordinates, relative or pixel margins, rotation and alpha opacity are converted into source-image geometry. A watermark that extends outside the image is cropped as placed instead of being silently moved back inside.
Relative sizing is the batch default, so an 8% watermark means 8% of each image width. You can switch an individual file to an absolute pixel size, and reruns always start from immutable original bytes so watermarks do not stack.
The watermark operation runs after base image operations and before one final encode. Downloaded images contain actual composited watermark pixels, and ZIP export includes only current successful results from the same /watermark-image route.
A visible watermark can communicate ownership or branding, but it does not technically prevent copying, cropping, editing or removal. ResizeNest does not present watermarking as guaranteed copy protection.
Questions, answered
Yes. Relative size and shared placement make consistent local batch watermarking practical, while per-file overrides remain available when one image needs different placement or size.
No. Processing and retry reread the immutable original image and rebuild the current recipe, so editing a watermark does not watermark a previous watermarked output.
A conservative standalone SVG subset can be rasterized locally after safety validation. Scripts, event handlers, external resources, recursive references and font-dependent SVG text are rejected instead of being executed or silently fetched.
No. A watermark can mark or brand an image, but it does not guarantee copy protection and may still be cropped, edited or removed.
No. Settings, processing, retry, result review, single download and current-success ZIP export stay on /watermark-image.