Convert PNG to JPG
Turn large PNGs into JPGs that are a fraction of the size. Converted in your browser, with control over what replaces transparency.
Drop PNG files here or paste from the clipboard.
Your files are processed on this device. Nothing is uploaded.
What actually changes
PNG stores every pixel exactly; JPG approximates them in a way tuned to human vision. For a photograph that trades a difference you can’t see for a file five to ten times smaller, which is the whole reason to do it. For a screenshot or a diagram the trade is bad, because JPG's approximations show up as halos around text and hard edges. Transparency doesn’t survive, since JPG has no alpha channel at all.
| PNG | JPG | |
|---|---|---|
| Compression | Lossless | Lossy |
| Transparency | Yes | No |
| Format can be animated | No | No |
Text and sharp edges are where this goes wrong
JPG was designed for photographs and it handles gradients beautifully, but it works by discarding high-frequency detail and a hard black-on-white edge is nothing but high-frequency detail. Screenshots of text, UI captures, line drawings and logos all pick up a visible grey fuzz around every edge, worst at low quality settings. If the image has text in it, convert to WebP instead, which stays sharp and is still far smaller than the PNG.
The screenshot trap
The most common reason someone converts PNG to JPG is that a screenshot is too large to email. It works, the file shrinks and the result looks slightly wrong in a way that is hard to name. That is JPEG’s ringing artefact: a faint grey shadow either side of every sharp edge, most visible around small text.
For screenshots specifically the better sequence is to reduce the dimensions first, since a 5K screenshot is carrying four times the pixels of a 1080p one and then save as WebP if the destination accepts it. If it has to be JPG, keep the quality at 90 or above and accept a larger file.
Where the size actually goes in a PNG
PNG compresses by predicting each pixel from its neighbours and storing the difference. On a screenshot with large flat areas, the differences are mostly zero and the file is tiny. On a photograph, every pixel differs slightly from its neighbours because of sensor noise, so there is nothing to predict and the file balloons. This is why a 2MB screenshot and a 30MB photograph can have identical dimensions.
Keep the original
Converting is destructive in one direction only. The JPG can’t be turned back into the PNG you started with and each further save as JPG degrades it again. If the PNG is the master copy of a logo or an asset, keep it and treat the JPG as an export.
Common questions
How much smaller will the JPG be?
For a photograph saved as PNG, usually between five and fifteen times smaller, because PNG is a terrible container for photographic data. For a screenshot or a flat-colour graphic, often only two or three times and occasionally larger, because PNG compresses flat colour extremely well and JPG doesn’t.
What happens to the transparent parts?
They’re filled with a solid colour, white by default, because JPG can’t store transparency. The background control on this page sets that colour. Choose the colour of whatever the image will sit on or convert to WebP if the transparency needs to survive.
Will the quality drop noticeably?
At the default quality, not on a photograph. On anything with text, flat colour or sharp edges, yes and often at a glance. The conversion itself is a one-way step, so keep the PNG if you might need to go back.
Should I use JPG or WebP?
WebP if you control where the image is going, because it is smaller than JPG at the same quality, keeps transparency and handles sharp edges much better. JPG if the destination is old software, a print shop, a picture frame or an upload form with a fixed list of accepted types.