QR Code Decoder

No. Decoding happens locally — Canvas reads the pixels and jsQR analyses them in your browser. The image and decoded data never leave the page.

Make sure the QR fills a reasonable portion of the image, is in focus, and has good contrast. Heavily blurred or tilted codes may need a clearer source.

Often yes — jsQR scans the whole image. If multiple codes are present it returns the first one it finds.

The decoded text is returned as-is. For Wi-Fi or vCard codes you'll see the raw payload (e.g. WIFI:T:WPA;S:Name;P:secret;;) which you can paste into another tool to act on.

Yes. The decoder attempts both polarities on every image, so a code printed light-on-dark, one lifted from a dark-mode screenshot, or a photo of a backlit display all work without you flipping anything first. That is not universal — plenty of phone camera apps only try the normal polarity and give up on inverted codes — so if a code scans here but not on your handset, inversion is a likely reason rather than damage to the code.

Because the image is scanned at its native resolution and nothing is downscaled first. A twelve-megapixel photo means twelve million pixels to binarise and search, which is slow, and the code itself may occupy only a small patch of that frame, giving the locator a lot of noise to reject. Crop tightly to the code and shrink the long edge to somewhere around 600-1000 px before uploading. What no amount of cropping fixes is a code on a curved surface, a strong perspective angle, or the moiré of a screen photographed at a bad distance.