Analytics cookies
We use optional Google Analytics to understand which tools are useful and improve the site. Analytics stays off unless you accept. Privacy Policy
Choosing AVIF, WebP, JPEG or PNG
There is no format that wins for every image. AVIF often makes the smallest lossy files and keeps transparency, but it is the slowest to encode. WebP is usually close behind, encodes quickly and also keeps transparency. JPEG has no transparency and tends to be larger at a similar look, but it is the format every image tool and service has handled the longest. PNG is lossless: ideal for flat graphics, screenshots and anything that must keep every pixel, and usually far larger for photos.
Format Lab encodes your image in all four formats at the same preset and shows the measured sizes, so the choice comes from this image rather than a rule of thumb.
Quality versus file size
The quality setting is each encoder’s own control, from 1 to 100 (99 for WebP, because browsers switch to a separate lossless encoder at 100). The same number doesn’t look the same in different formats, so the tool never carries a quality number from one format to another.
Safe, Balanced and Aggressive are starting points set per format: AVIF 72, 60 and 45; WebP 88, 78 and 62; JPEG 90, 80 and 65. Safe favors fidelity, Aggressive favors size. They aren’t visual-quality guarantees: check the comparison, zoomed in on edges, text and smooth gradients, before you ship.
Using a target file size
Target size finds the highest quality whose real encoded file is at or under the size you enter, at the current dimensions and format. It encodes and measures a handful of qualities, then checks the ones next to the answer, because file size doesn’t always rise with every quality step. The result is always a size it measured, never an estimate.
It never shrinks the image to reach a target. If even the lowest quality it tries is too large, it says so: reduce the dimensions, choose another format or use a larger target. PNG has no quality to search, so target size applies to AVIF, WebP and JPEG.
Transparency and metadata
AVIF, WebP and PNG keep transparent areas. JPEG can’t store transparency, so when an image has transparent areas and you choose JPEG, they are filled with a background color you pick, and the comparison shows that fill before you download. Transparency is read from the file and, when the file can hold it, confirmed on the pixels.
Every export is newly encoded from the pixels, so it doesn’t copy the source file’s EXIF, XMP or other metadata, such as camera details or a location. The tool shows which kinds of metadata the source contains, without reading their values. A photo’s rotation tag is applied to the pixels, so exports are saved upright. Colors are converted by the browser and re-encoded, so an embedded color profile isn’t carried over byte for byte.
Your image stays on your device
Your image is decoded, compared and encoded in your browser, and is never uploaded to GetKineticHub or anywhere else. WebP, JPEG and PNG use your browser’s own encoders. AVIF uses an encoder that comes with this page as WebAssembly, loaded from this site the first time you choose AVIF and run in a background worker, so the page stays responsive during long encodes.
Images up to 8192 px on each side and 30 MB are accepted. If you accept optional analytics, the tool reports only that a download or share finished, its format and whether the quality came from a target size. Your image, its name, sizes and settings are never included.