Batch image resizer
Drop in a folder, choose how the images should be sized, and process the whole set at once. Nothing is uploaded.
Drop images here
JPG · PNG · WebP · AVIF · GIF · BMP
- Processed on your device
- Built for large batches
- No account, no upload
Six ways to specify a size
Most batch jobs fail for a boring reason: the images are not all the same shape. A folder from a phone, a mirrorless camera and a screenshot tool contains portrait, landscape and square files, and a single width-and-height pair means something different for each one. The resize modes exist so you can describe the outcome you want rather than one rectangle.
- Width or Height — set one edge, the other follows the original ratio. The safest choice for mixed batches.
- Exact size — pin both edges, then decide with fit, fill, stretch or pad what happens to images that are a different shape.
- Percentage — scale everything by the same proportion, useful when the originals are already consistent.
- Longest side — cap the biggest edge, so portrait and landscape images end up equally large on screen. This is what most galleries and portfolios actually want.
- Shortest side — guarantee a minimum edge, useful when a platform enforces a floor.
Fit, fill, stretch and pad
These four only apply when both dimensions are pinned, and they answer one question: what should happen to an image whose shape does not match the box?
Fit keeps the whole image and lands inside the box, so the output may be smaller than requested in one dimension. Fill covers the box exactly and crops the overflow, using the focal point to decide what survives — this is what social platforms do to your uploads anyway. Stretch forces the exact numbers and distorts anything of a different ratio; it is occasionally what you want for textures, and almost never what you want for photographs. Pad fits the image and fills the leftover space with your background colour, which is how you standardise a catalogue on white without cropping products.
Enlarging is not the same as resizing
Making an image smaller discards information, which is a normal and lossless-looking operation when done in stages. Making one larger has to invent information that was never captured. This tool interpolates cleanly, but a 400-pixel image scaled to 2000 pixels will look soft because the detail does not exist. The “don’t enlarge smaller images” option is on by default for that reason: images already below your target are left alone rather than blown up to match.
What happens to each file
Every image goes through the same sequence: decode, apply EXIF orientation, rotate or flip if you asked, crop, resize, add background and padding, draw any watermark, then encode to the output format. Large reductions are done by halving repeatedly rather than in one jump, which is what keeps edges clean. Processing runs in worker threads so the interface stays usable while a long batch works through.
Questions
- Are my images uploaded anywhere?
- No. Files are read from your device, processed by your browser, and saved back by you. There is no image server and no processing API. You can verify this in your browser's network panel while a batch runs.
- How many images can I process at once?
- There is no fixed cap, and there is no honest single number either. Batches of several hundred medium-resolution photos are routine on a modern laptop; very large images or a phone with limited memory will run into trouble sooner. Processing runs a few images at a time and reports any file that fails.
- Which formats can I use?
- JPEG, PNG, WebP and, in browsers that decode it, AVIF. GIF and BMP usually decode too, but only the first frame of an animation is used. Output is JPEG, PNG, WebP, and AVIF where your browser can encode it — the option is disabled when it cannot.
- Can I resize without losing quality?
- Making an image smaller always discards pixels, but done well the result is visually clean. Images are downscaled in stages rather than one jump, which avoids the jagged edges you get from a single large reduction. Enlarging is a different matter: no tool can invent detail that was never captured.
- Do I need an account?
- No. There is no sign-up, no login, and nothing to install. Your settings and saved presets are kept in your browser's local storage; your images are not stored at all.
- Can I download everything at once?
- Yes. Processed files can be downloaded individually, as a selection, or packaged into a single ZIP built in your browser. In browsers that support the File System Access API you can also write straight into a folder you choose.
- What happens to my images when I close the tab?
- They are gone. Files are held in memory for the session only. Nothing is written to local storage, IndexedDB, or a cache, so a refresh means starting again — which is also why nothing lingers.
- What is DPI, and does changing it improve my image?
- DPI describes how many pixels are packed into an inch of physical print. It tells a printer how large to print a given number of pixels. Changing DPI alone does not add detail; what matters is whether you have enough pixels for the size you want. A 4 × 6 inch print at 300 DPI needs 1200 × 1800 pixels.