Image Tools

Batch Image Resizer

Resize multiple PNG, JPEG and WebP images locally. Fit each within a maximum size or scale by percentage, preserve aspect ratios and download successful results together in a ZIP.

Your content stays on your device Free, with no account
One set of settings. A whole batch. Local images · local ZIP

No file selected.

Up to 20 static PNG, JPEG or WebP images · 20 MiB each · 80 MiB total

No images selected. Choosing again replaces the selection.

PNG and WebP retain transparency. JPEG fills transparent areas with this color.

Aspect ratios are preserved with pixel rounding. Fit mode adds no padding and performs no crop. Input: 40 megapixels / 16,384 px per side. Output: 16 megapixels / 8,192 px per side per image.

YOUR RESIZED BATCH

Select images and run the batch to see previews.

Ready. Choose images to begin.

Images are processed sequentially. The ZIP contains successful outputs and report.json listing every outcome. Failed images are skipped. Stop or changing settings discards the pending ZIP. Output images have new encoding and do not preserve original metadata.

A FEW SIMPLE STEPS

How to use Batch Image Resizer

  1. Choose or drop up to 20 static PNG, JPEG or WebP files. Choosing again replaces the entire selection.
  2. Choose maximum width and height for Fit mode, or set a percentage applied to each original. Aspect ratios are preserved.
  3. Choose output format and quality, then select Resize batch. Review each file’s result or error.
  4. Download the ZIP containing successful images and report.json. Stop or changing settings discards the pending batch.
PRIVATE BY DESIGN

Why use this tool?

Your images are never uploaded to any server; decoding, resizing and ZIP packaging run locally in your browser. Sequential processing keeps only one full decoded source active at a time. No image history or account is needed, and you control when results are downloaded.

Tool inputs stay in your browser. Only your theme preference is saved on this device.

A few things worth knowing

How many files can I resize at once?

Select 1–20 files, each nonempty and up to 20 MiB, with a combined input limit of 80 MiB. Each source is limited to 40 megapixels and 16,384 pixels per side. Each output is limited to 16 megapixels and 8,192 pixels per side. Encoded images in a batch are limited to 100 MiB in total; files that would exceed it are reported as failed. Browser memory may impose lower practical limits.

Does Fit mode crop, stretch or pad images?

No. Fit mode scales each image to fit inside both maximum dimensions while keeping its aspect ratio, with integer-pixel rounding. No padding or crop is added. Do not enlarge smaller images is enabled by default and applies only to Fit mode. Percentage mode scales each original independently, accepts 0.01–1,000% with up to two decimals and rounds tiny dimensions to at least one pixel.

Which file formats and export options are supported?

Static PNG, JPEG and WebP are supported. SVG, GIF, AVIF, HEIC and detected animated PNG/WebP are not supported. Keep each original format retains its type, but re-encodes the pixels. You can instead select a single output type for the entire batch. JPEG uses the selected background for transparency; PNG/WebP keep alpha. Quality applies to JPEG/WebP and does not promise a particular file size.

What is included in the ZIP?

Successful outputs use indexed, sanitized filenames to avoid collisions and archive paths. report.json contains original filenames, settings, success/error details, dimensions and byte sizes. Failed images are omitted from the image outputs but remain in the report and result table. The ZIP stores already-encoded images without another compression pass. If all files fail, no ZIP is offered.

Can I stop a batch or change its settings?

Yes. Stop, new file selection, setting changes and Clear discard pending results and disable the previous ZIP. A browser decode already underway cannot be forcibly aborted, but its result is discarded and released when it returns. Image encoding has a 15-second response limit. Previews show up to six successful images as small thumbnails rather than full decoded output images.

Are metadata and original quality preserved?

No. Canvas export makes a new raster image; original metadata, exact encoded bytes and color-profile behavior are not guaranteed. The browser handles image orientation and interpolation, and lossy re-encoding can change quality. Upscaling cannot restore missing detail. The tool never uploads or stores selected image content as application history.