QR Code Reader

The reader gives you two paths to the same result. Upload mode takes any picture that contains a code: a screenshot, a saved photo, a scan of a printed page, or product packaging. Drag the file in or pick it from your device, and the decoder looks across the whole image for every code it can find. Camera mode points your device at a code in the real world and decodes it as soon as the frame is sharp enough. Upload works best when the code already lives on your screen or in your photo library, since there is nothing to physically point at. The camera shines when you are holding the printed code in front of you and want an instant read. Both modes use the same ZXing-based engine, so a code that decodes from a photo will decode just as well live. You can switch between them at any time without losing your place, picking whichever route matches where the code happens to be.

A QR code is just a container for text, and that text follows patterns the reader recognizes. The most common payload is a URL, shown here as a clickable link so you can open it without retyping. Plain text comes through exactly as written, useful for notes, serial numbers, and short messages. Structured payloads carry more meaning: WiFi credentials list the network name, security type, and password; a vCard holds contact details you can save in one tap; email, SMS, and phone payloads prefill the matching app. Calendar events bring a title, date, and location. The reader detects which type it is decoding and presents it accordingly, so a contact looks like a contact and a network looks like a network rather than a wall of raw characters. Seeing the structure makes the result easier to trust and act on. If you want to build these payloads yourself, the QR Code Generator creates every type, and the Barcode Reader handles the linear codes on retail packaging.

When a code refuses to decode, the image is almost always the cause rather than the code itself. Blur is the top offender, so hold the camera steady or pick a sharper photo. Glare from a screen or glossy paper can wash out the pattern; tilt the surface slightly to move the reflection off the code. Low contrast, such as a faint print or a dark code on a dark background, leaves the decoder little to lock onto, so brighten the shot if you can. Cropping matters more than people expect: every QR code needs a clear margin around it, called the quiet zone, and cutting into it can break the read. Make sure the full code plus a little white space is inside the frame. If the code is small in a large photo, crop closer so it fills more of the image. Working through these one at a time usually turns a stubborn code into a clean result without any special equipment.

QR is the most familiar pattern, but it is not the only 2D code the reader handles. Data Matrix appears on small parts, electronics, and labels where space is tight, packing data into a compact square. Aztec codes show up on transit tickets and boarding passes, recognizable by the bullseye at their center. PDF417 is the tall, striped block found on driver licenses, shipping documents, and event tickets, storing more data than a single line barcode can. The reader detects which format it is looking at and decodes it the same way it handles a QR code, so you do not need to choose a mode in advance. This matters when a document mixes several code types, since one upload can surface all of them at once. For the strictly linear retail and logistics symbologies such as EAN and UPC, the Barcode Reader is the better fit, while this tool focuses on the 2D family.

Reading a code does not require sending anything to a server. The decoding happens in JavaScript inside your browser, so the image you upload or the camera frame you capture is processed on your own device and never leaves it. That design has practical benefits beyond privacy. There is no upload wait, which is why a read can feel instant even on a large photo. It works on a flaky connection, since the heavy lifting is local. And sensitive content stays sensitive: a QR code might hold a WiFi password, a private link, or contact details you would rather not hand to a third party. Because the camera stream is read frame by frame and discarded, nothing is recorded. The one requirement is that camera access needs a secure connection and your permission, a browser rule that protects you rather than the tool. Once you grant it, every scan runs entirely on your hardware, with the decoded text appearing only on your screen.