Pixels, not inches
A digital image has no physical size. It is a grid of pixels, and the only thing a resize can change is the size of that grid. The DPI value you see in photo editors is a label saved in the file's metadata. It tells a printer how densely to place the pixels, but it does not add or remove any.
So for print, do the sum yourself: print size in inches multiplied by the DPI you want. A 4 × 6 inch print at 300 DPI needs 1200 × 1800 pixels. The same print at 150 DPI needs 600 × 900. The table gives a few common sizes using the same arithmetic.
| Print size | Pixels at 300 DPI | Pixels at 150 DPI |
|---|---|---|
| 4 × 6 in | 1200 × 1800 px | 600 × 900 px |
| 5 × 7 in | 1500 × 2100 px | 750 × 1050 px |
| 8 × 10 in | 2400 × 3000 px | 1200 × 1500 px |
For screens, dimensions are usually set by the page or app that displays the image. Platforms change their recommended sizes from time to time, so check the platform's current specification before you resize for it. Our guide to resizing images for social media covers the workflow.
Single image versus batch: how the fields behave
The same two fields and the padlock mean slightly different things depending on how many files you add.
- One image. The fields start at the original size. With the padlock on, typing a width sets the height from the original aspect ratio, and the other way round. With it off, the two values are independent. The output is always exactly the width and height shown, so an unlocked, mismatched pair will stretch the picture.
- Batch, padlock on (fit within). The fields become a maximum width and a maximum height. Each image is scaled by the smaller of the two ratios so it fits inside the box, and keeps its own proportions. Leave one field empty to limit only the other. A box of 1200 wide and no height limit gives every image a width of 1200 whatever its height.
- Batch, padlock off (exact). Every image is drawn at precisely the width and height you give, which is useful for uniform thumbnails only if your sources already share an aspect ratio.
Fit-within is not limited to shrinking. If the box is larger than an image, that image is enlarged to fill it, so choose a box no bigger than your smallest source unless you want that.
Shrinking is safe; enlarging has limits
Reducing an image throws away pixels you did not need at the new size. Enlarging has to invent pixels by blending their neighbours, which adds size but not detail. A 600 pixel wide photo stretched to 1800 pixels will look soft because there is nothing more to reveal. If a printer or platform wants more pixels than you have, a new, higher-resolution original is the only real fix.
Resizing also re-encodes the file. PNG output is lossless. JPG and WebP output is saved at a fixed 92% quality, so repeated resize-and-save cycles on a JPG will slowly add artefacts. Resize once from the original rather than from a copy you already resized.
What the tool keeps and what it limits
Files keep their original names, with the extension matching the output format. Each file can be up to 20 MB, a batch holds up to 50, and the size fields accept up to 10,000 pixels. Batches run one image at a time and finish as a ZIP. When a result is ready, the Continue with bar can pass it directly to another tool, such as the compressor, without uploading it again. Everything happens in your browser tab; nothing is sent to our servers.