File4Edit

Optimize images for the web

Optimize images for the web automatically, for free and without uploading them. Each image is measured, given a format and the lowest quality that still looks right, and every decision is listed so you can check it.

Chooses a format from what the image actually is, then finds the smallest file that still looks right. It shows you what it decided.

In detail

Auto optimize chooses the output format and the compression quality for each image by measuring the image itself, then lists every decision in a ledger under the queue. WebP is the usual format, because it is small and keeps transparency, and Safari and iPhone, which cannot write WebP themselves, get it from File4Edit's own encoder. Only a browser with no way to write WebP at all gets PNG for transparent or flat images and JPEG for photographs.

The quality comes from a search that encodes the image, decodes it back and compares brightness, settling on the lowest quality whose error stays inside the budget you picked: Smallest, Balanced or Best looking. The long edge is capped at 2400 pixels by default, which is often the biggest saving of all.

Optimize images for the web at a glance

Input
JPG, PNG, WebP, AVIF, HEIC, GIF (first frame), BMP, TIFF, SVG and other image formats
Output
WebP, or PNG and JPEG where WebP cannot be written
What it does
  • A format chosen for each image from its content
  • Quality search against a visual error budget
  • Three budgets: Smallest, Balanced and Best looking
  • Long-edge cap, 2400 pixels by default
  • A ledger explaining every decision
  • Batch processing with one ZIP download
Who it is for
Web developers and site owners who want small images without judging each one by eye.
Limits
  • In Safari and on iPhone and iPad, WebP comes from File4Edit's own encoder, so the quality search runs more slowly there.
  • An animated GIF comes out as a single still frame.
  • PDFs are not accepted; use Compress PDF for those.
Privacy
Runs entirely in your browser: no file is uploaded, and nothing is kept after you close the tab.

How to optimize images for the web

  1. Drop your images Add the files. Each one is read once and reduced to a small sample, so changing a setting later re-decides without decoding again.
  2. Pick a quality budget Smallest, Balanced (the default) or Best looking. This is the one number everything else follows from: it sets how far the pixels are allowed to move from the original.
  3. Set the largest edge The default caps the long side at 2400 px, which is usually the biggest saving of all. Clear the field to write every image at the size it arrived.
  4. Read the ledger Under the queue, each row names the chosen format, the chosen quality, the output dimensions and the reason — photograph or flat art, color count, and whether there is transparency.
  5. Apply and download Press Apply to apply the decisions at full resolution, then download. A batch comes back as a zip.

Optimize images for the web: common questions

How does it decide which format to use?

By what the image is and what this browser can write. WebP comes first, because it is small and keeps transparency; Safari and iPhone cannot write WebP themselves, so there File4Edit uses its own WebP encoder. Only a browser with no way to write WebP at all gets PNG for anything with transparency or too few colors for a lossy codec, and JPEG for a photograph. An Allow AVIF switch waits for the day a browser's own encoder can write AVIF; none can yet, so it changes nothing today.

How does it choose the quality number?

It encodes the image, decodes the result and measures the root-mean-square luminance difference from the original, then binary-searches for the lowest quality whose error stays inside the budget — about 3.2 luminance values for Balanced, 6.5 for Smallest and 1.6 for Best looking. The search runs on a copy scaled to 480 px on its long edge so half a dozen encodes stay quick; the chosen quality is then applied to the full-size export.

Can it handle PDFs and animated GIFs?

No. PDFs are turned away when you drop them, with a pointer to Compress PDF, and an animated GIF comes out as a single still frame — the page warns you when it sees one.