The biggest file-size win is often dimensional: a 4000px source serving a 1200px slot is pure waste. This resizer caps the longest edge at 1920 pixels, then you can optimise the result for a tiny, sharp file.
The saving optimisation can't give you
Optimisation rearranges how a JPEG stores its pixels; it cannot remove pixels that shouldn't be there in the first place. If your source is 4000 pixels wide and the page only ever shows it at 1200, three-quarters of the data is pure overhead, and no amount of optimisation recovers it — the encoder dutifully packs detail nobody will see. Resizing addresses the root cause. Because file size tracks pixel area, cutting the longest edge from 4000 to 1920 removes a large majority of the data before any other tool touches the file. This resizer caps the longest edge at 1920 pixels, the practical ceiling for full-width display on high-DPI screens, so you keep visual quality where it shows and shed the waste where it doesn't.
Resize, then optimise — the canonical order
The sequence is deliberate, and reversing it wastes work. Resize to display size first so the optimiser only has the pixels you'll actually serve; then optimise so those pixels are stored as tightly as possible. Run optimisation first and you've carefully encoded data you're about to discard. The clean flow is:
- Resize the oversized source here to a 1920px maximum edge.
- Optimise the result with JPEG optimisation for a lean, sharp file.
- Convert to WebP when you want the smallest modern delivery.
Each stage is a single upload, and together they turn a heavy camera original into a build-ready asset.
Uniform scaling keeps geometry honest
Resizing here scales width and height by the same factor, so the image never stretches or squashes — straight lines stay straight and proportions stay true. Applying the cap to the longest edge means portraits and landscapes both end up sized sensibly without you specifying dimensions by hand. For a content pipeline that's one less thing to get wrong. Once a source is right-sized, the byte reduction continues downstream: optimise it for the baseline, or push to WebP for the lightest modern file. The full set lives together on jpegoptim so the resize-then-optimise habit stays frictionless.