jpegoptim — Frequently Asked Questions

Yes — free, no watermark, no signup.

Optimisation keeps quality high and visually unchanged; aggressive compression trades a little sharpness for size.

Yes, upload many files at once.

It re-encodes the file with optimised Huffman tables and progressive scan order, shrinking it and speeding rendering without a visible quality change.

The re-encode applies a high quality setting; it is near-visually-lossless, prioritising small size with no perceptible difference.

One that loads in successive passes from blurry to sharp, instead of line by line — it feels faster on slow connections.

Yes — batch uploads are supported and returned together.

Often 70-85% smaller than the original, depending on the image and how much softening you're willing to accept in exchange for the size cut.

Use aggressive compression for thumbnails, previews, and tight upload limits where small size matters most. Use optimisation when image quality has to stay high.

In fine textures and sharp edges, which soften first. Smooth areas like sky and skin hold up well, which is why small images look fine.

Yes, and it helps. Resize the source down first, then compress hard. The two reductions compound to beat strict size caps that one alone can't.

For file size, usually yes — WebP is typically smaller at the same visible quality. JPEG still wins on universal compatibility outside web browsers.

All modern browsers display WebP natively. Older or unmaintained clients may not, which is why keeping a JPEG fallback is still common practice.

No. Conversion can't recover detail JPEG already discarded. It re-encodes the existing picture into a smaller WebP at the same visible quality.

Serve both. Offer WebP to browsers that accept it and an optimised JPEG as a fallback, so modern clients save bytes and older ones still work.

Optimisation can't fix oversized dimensions. Resizing first removes pixels you were never displaying, so the optimiser only encodes the data you'll actually serve.

No. Width and height scale by the same factor, so the aspect ratio is preserved and nothing is stretched or squashed.

A lot. File size tracks pixel area, so halving the longest edge removes roughly three-quarters of the data before any compression runs.

1920 pixels is the practical limit for full-width display on high-resolution screens. Anything larger usually adds file size without adding visible detail.