WebP files are smaller than JPG or PNG at equivalent visual quality, and every browser still receiving updates has displayed them without issue for years. That combination — real savings, no compatibility risk for viewers — makes converting to WebP one of the easiest high-payoff changes you can make to a batch of images.

There are two things most guides on this leave out, and both matter: WebP is really two formats wearing one extension, and there are still specific places that will refuse the file outright. Let's cover the conversion first, then both of those.

Converting JPG to WebP

The common case: photos, product shots, anything headed for a web page. Both formats are lossy by default, so this is a straightforward swap.

Encoding the same photograph both ways at matched quality — not at matched quality numbers, which measure nothing, but adjusted until the two look equivalent — we measured WebP landing about 36% below the JPG. That's in line with the 25–35% usually quoted, and on a page full of photos it's the difference between a fast load and a slow one.

Two practical notes:

Converting a JPG to WebP is a second lossy pass. Strictly, some additional detail is lost. At a sensible quality setting (80 or above) it isn't visible at normal viewing size, and the size win is worth it. It only becomes obvious if you go back and forth repeatedly or push the quality very low.
If you still have the original — the RAW, the camera export, the design file — export to WebP from that instead. One lossy pass beats two, always.

Converting PNG to WebP

A different job entirely: screenshots, logos, icons, line art, anything with flat colour, sharp text or a transparent background. Transparency survives the trip, which is the main reason people make it.

But here's the part that actually matters, and almost nobody explains it: WebP has two completely separate modes.

Lossy WebP works like JPG — it approximates, and it smears hard edges. Fine for photos, wrong for text.
Lossless WebP is a different design entirely, purpose-built for flat colour and sharp edges. It stores every pixel exactly, and on that kind of image it comfortably beats PNG.

For a screenshot you want the lossless mode. The complication is that browser-based converters, this one included, reach WebP through the browser's own canvas encoder, and which mode you get at the top of the quality range is browser-dependent: Firefox switches to lossless WebP at maximum quality, and Chromium has a lossless path there too, but this is implementation behaviour rather than something the page can request explicitly.

The practical upshot:

The PNG to WebP converter starts its quality slider at 100 rather than the photo default, precisely because this content punishes lossy encoding. At that setting the result is either lossless or so near it that flat colour and text survive intact.
If you need a guarantee of pixel-exactness — a logo going into a brand system, a screenshot that will be diffed or re-edited — keep the PNG. A guarantee is not something a canvas-based converter can honestly offer you.

Where WebP still gets rejected

This is the section that saves you the wasted conversion, and it's missing from most articles because it undercuts the pitch.

Browser support is a settled question: every browser still getting updates renders WebP. But displaying a format and accepting an upload of it are different problems, and the second one is not settled at all.

Places that still routinely refuse .webp:

Upload forms — government portals, job application systems, insurance and banking sites. Many validate against a hardcoded list of image/jpeg and image/png.
Email — a number of desktop and webmail clients won't render an inline WebP, so a newsletter image can arrive as a broken box.
Print services — photo printing, business cards, anything ending on paper generally wants JPG or TIFF.
Older desktop software — office suites, legacy design tools and internal corporate apps of a certain age.

None of this is a reason to avoid WebP on your own website, where it's the right default. It is a reason to keep the JPG around for everything that leaves your website. And if you've already converted and hit one of these walls, converting back is the fix — the WebP to JPG converter and WebP to PNG converter exist for exactly this, which tells you how often it comes up.

What the conversion throws away besides pixels

Converting through a browser works by decoding the image, drawing it onto a canvas, and re-encoding from those pixels. A consequence that surprises people: the metadata doesn't survive. EXIF — camera model, capture date, lens settings, and often GPS coordinates — is not carried across.

That cuts both ways, so decide which side you're on before you batch-convert:

Good, if you're publishing. Stripping GPS coordinates out of a photo before it goes on the internet is a real privacy win, and one most people never think about.
Bad, if you're archiving. Capture dates and copyright fields are worth keeping. Convert a working copy for the web and keep the originals untouched.

Colour profiles are in the same boat. For anything colour-critical — print work, product photography where the shade has to match — check one converted file before committing the whole set.

Animated GIFs are a different question

WebP supports animation and beats GIF badly at it: a fraction of the size, and thousands of colours instead of 256. But converting an animation is a different operation from converting a still, and our converter doesn't do it — it processes single frames, so an animated source will not come out the other side as a working animation.

If that's what you need, it's a job for a dedicated animation tool. Worth knowing before you feed it a folder of GIFs and wonder why the results are motionless.

Does using WebP affect SEO?

Not negatively, and slightly positively. Google indexes WebP images normally, and it's the format its own PageSpeed Insights recommends when it flags oversized images. Since image weight feeds into Largest Contentful Paint, which is a ranking signal, a lighter format helps indirectly.

The one thing to get right is the switch itself: if you change a URL from .jpg to .webp, that's a new address. Either keep serving the old one, or redirect it, so you don't lose whatever the original had accumulated in image search.

The quality number doesn't carry over

Worth knowing before you copy your JPG settings across. Quality 80 in a JPG encoder and quality 80 in a WebP encoder are inputs to different algorithms; they don't describe the same visual result. WebP's scale generally runs a little "hotter", so the same number tends to give you a slightly better-looking image and a slightly larger file than you'd predict from JPG habits.

Don't port your number. Export one representative image at a couple of settings, look at them at display size, and pick from what you see.

Doing it in bulk

Everything above applies the same to one image and to four hundred. Drop a folder into the JPG to WebP converter, set the format and quality once, and take the result as a single ZIP.

One caveat for mixed folders: photos and screenshots want opposite settings. Running both through at quality 80 leaves the photos slightly larger than necessary and visibly damages the text. Split them into two passes — it costs a minute and it's the difference between a clean batch and a redo.

Nothing is uploaded in any of this. The conversion happens on your own device, which you can confirm by opening your browser's Network tab while it runs.

Where this fits

For the full comparison, including where AVIF beats WebP and the specific cases where it doesn't, see WebP vs JPG vs PNG vs AVIF and AVIF vs WebP. If page weight is what sent you here, how image size affects website speed puts the format choice in the context of the whole loading budget.

How to Convert JPG or PNG to WebP