The image resizer and compressor that never uploads your files

Presets

Picking dimensions for each platform

Every social platform re-compresses what you upload, and each one has a size it prefers. Feeding the platform exactly what it wants avoids a second, server-side resize you have no control over — and keeps text and fine detail sharp. The preset selector above fills the width and height fields with the current recommendations:

• Instagram Post — 1080×1080, the square feed format. Portrait posts work at 1080×1350, which you can type in manually.
• Instagram Story — 1080×1920, a 9:16 full-screen vertical.
• X / Twitter — 1600×900 for single-image posts, a 16:9 frame that displays without cropping in the timeline.
• Pinterest — 1000×1500, the standard 2:3 pin ratio the feed is built around.
• YouTube thumbnail — 1280×720, the minimum size YouTube accepts for custom thumbnails.
• LinkedIn — 1200×627 for link and post images, roughly a 1.91:1 ratio.

If you prepare images for several platforms at once, add the files, apply one preset, download the batch, then switch presets and run them again. Each pass takes seconds because nothing leaves your machine.

Compression

Quality vs. dimensions: what actually shrinks a file

An image's file size is roughly the number of pixels multiplied by the cost of describing each pixel. The dimensions control the first term; the quality setting controls the second. Halving both sides of an image quarters the pixel count, which is why reducing dimensions saves far more than nudging the quality slider — a 50% resize often beats dropping JPEG quality from 90% to 60%, and looks better doing it.

The trade-off works in the other direction too. If the dimensions are fixed — a 1280×720 YouTube thumbnail, say — quality is the only lever left. For JPEG and WebP, the 80–90% range is usually indistinguishable from the original at a fraction of the size, and WebP typically reaches the same visual quality at a smaller size than JPEG.

Target file size mode automates this balancing act. It binary-searches the quality setting to land under your limit, and only steps the dimensions down to 85% when quality alone cannot get there — then tells you the final quality and size it used, so there are no surprises.

Privacy

Why resize images client-side?

Most online resizers ask you to upload your images to a server, resize them there, and download the result. That upload is a copy you can no longer account for — of photos that often contain EXIF metadata revealing where and when they were taken, or AI metadata describing how they were generated.

This image resizer does the work where the file already is: your browser reads it with the File API, redraws it on a local canvas and saves the result straight back to your device. There is no upload step, so there is nothing to intercept, log or retain. You can open your network inspector, or disconnect from the internet entirely, and the tool keeps working.

A useful side effect of the canvas re-encode is that metadata does not survive it. Resizing here doubles as a quick pass with a metadata cleaner — and for a verified clean with a full before-and-after report, the MetadataRemover metadata cleaner is one click away.