Set exact pixels, a percentage or a social media size. Runs on your device.
A LinkedIn profile banner is very wide and very short: LinkedIn's help page recommends 1584 by 396 pixels, a shape four times wider than tall. Almost every photograph is taller than that, so the first question is which strip of your picture should be the banner. This page opens with the LinkedIn banner preset and Cover selected, which takes a wide strip from the middle of your picture and shrinks it to 1584 by 396. A 4000 by 3000 photo keeps only the middle third of its height under Cover, so a skyline or an office wide shot works better than a portrait. Contain keeps the whole picture and fills the space on both sides with a colour, which suits a logo on a plain background. If your image is narrower than 1584 pixels it is enlarged and a warning shows. The tool does not choose the strip for you, so crop first if the middle is not the part you want.
Use Pixels when a form or a design gives you exact numbers, such as 1080 by 1080. With Lock aspect ratio on, you type one side and the other follows the shape of your picture, so nothing is squeezed. Use Percentage when you only want smaller or larger: 50 percent turns a 4000 by 3000 photo into 2000 by 1500, and a value up to 400 enlarges it. Use Preset when the target is a social media size and you do not want to remember the numbers. A preset has a fixed width and height, so your picture's shape may not match it, and that is what the Fit setting is for. With a batch, every image gets the same treatment: in Pixels mode with the lock on, each picture keeps its own shape and is sized from the side you typed.
Fit decides what happens when the new shape differs from the old one. Cover fills the whole frame and cuts off whatever sticks out, taking the middle of the picture. A 4000 by 3000 photo at 1080 by 1080 keeps a centred square of 3000 by 3000 pixels. Contain keeps the whole picture and fills the leftover space with a colour you choose, or with transparency for PNG and WebP files. The same photo at 1080 by 1080 becomes 1080 by 810 with bands of 135 pixels above and below. Stretch squeezes the whole picture into the new frame, so faces and circles look distorted unless the shapes already match. Fit is hidden when the shape cannot differ, in Percentage mode and in Pixels mode with the lock on.
A browser that shrinks a large picture in one jump skips over many pixels, and the result looks jagged or full of shimmer on fine detail. This tool halves the picture again and again while the next half is still at least as big as your target, then makes the final small adjustment. A 4000 by 3000 photo going to 1000 by 750 passes through 2000 by 1500 first. Each step averages neighbouring pixels, which is why text and edges stay clean. Enlarging is done in one step, because there is nothing to average. The output keeps the format of the input: JPEG and WebP are saved at quality 90, and PNG is saved without loss.
Making a picture bigger cannot add detail that was never captured. A 400 by 400 picture stretched to 1080 by 1080 is 2.7 times larger and will look soft. The tool allows it and shows a warning whenever the picture is enlarged more than 2 times. The largest size you can enter is 16384 pixels on a side. Your device may allow less: if the result would be too big for this browser's canvas, the tool makes it smaller to fit and tells you the real size before you start. This tool does not use AI to invent detail, it does not choose a crop around faces, and it does not change the file format or the print resolution stored in the file. For a manual crop use the crop tool, and for a different format use the converter.
Cover removes everything outside the 4 to 1 strip. A 4000 by 3000 photo keeps only 1000 pixels of its height before it is shrunk. Use Contain to keep the whole picture with coloured bands on the sides, or crop the photo first so the strip you want is in the middle.
Yes. Add the PNG, choose Contain and tick Transparent next to the pad colour. The bands stay transparent in the PNG.
Making a picture smaller keeps it sharp because the tool shrinks in steps. A JPEG or WebP is saved again at quality 90, which is a small change you rarely see on screen. PNG is saved without loss. Making a picture larger always looks softer, because the new pixels are estimated from the old ones.
Cover fills the whole frame and cuts off the parts that do not fit. For a 4000 by 3000 photo at 1080 by 1080, it keeps the centred 3000 by 3000 square and shrinks that. Choose Contain if nothing may be cut off, and Stretch if you would rather squeeze the whole picture.
When the picture is enlarged more than 2 times, for example 400 by 400 up to 1080 by 1080, the extra pixels are guesses and the result looks soft. The warning does not block anything. The finished row also shows how many times it was enlarged, such as 2.7×.
Yes, up to 50 at a time. Every image gets the same settings and is processed one after another. With Lock aspect ratio on, each picture keeps its own shape and shows auto for the side you did not type. Download all (.zip) saves names like photo-1080x810.jpg.
The presets were checked against each platform's help page when this tool was built, and platforms change their recommendations. Before an important post, look at the platform's current help page. If it gives different numbers, choose Pixels and type them.
No. A JPEG stays a JPEG, a PNG stays a PNG and a WebP stays a WebP. If your browser cannot save WebP, that one file is saved as PNG and its row says so. To switch formats, use the image converter.