Enlarge an image 2x, 3x or 4x
Make a small image bigger with smooth interpolation instead of a blocky stretch. Honest about what it is: it cleans up the enlargement, it does not invent detail an AI would. Runs in your browser.
Drop a low-res image to upscale
JPG, PNG, WEBP, works best with images under 1500 by 1500
Smooth beats blocky
The same small image enlarged. Left is a raw blocky stretch, right is the smooth interpolation this tool uses.
Blocky stretch
This tool
Both started from the same tiny source. Notice the smooth side has no jagged steps, but it is not inventing fine detail either. That is the honest limit of interpolation.
What it can and cannot do
Good for
Bumping a 600px product photo to 1200px, making 1x UI assets retina-ready, or meeting a platform's minimum upload size.
Not magic
It will not turn a tiny, blurry photo into a sharp one. The detail that was never captured cannot be added back by enlarging.
Tip
For a big jump, run 2x twice instead of one 4x. The edges come out cleaner and the file stays more usable.
Frequently asked questions
It will look smoother and less pixelated, but it cannot bring back detail that was never captured. Enlarging guesses the in-between pixels; it does not add real information. For a genuinely blurry source, expect a cleaner version, not a magically sharp one.
AI upscalers invent plausible detail with a neural network, which is slower, needs an upload, and sometimes fabricates features that were not there. This tool uses fast, honest interpolation in your browser: instant, private, and it never makes things up.
2x is the safe default and looks best. For a bigger jump, running 2x twice usually gives cleaner edges than a single 4x.
Up to 20 MB in. After a 4x enlarge a very large source can hit the browser memory limit, so start from a smaller original when you can.
No. Upscaling runs in your browser. Your file never leaves your device.