PDF QR Generator

Yes — every QR generated here follows the ISO/IEC 18004 standard, the same spec used by Apple Camera, Google Lens, and every QR scanner app. The QR is rendered at 4× the display size and then embedded as a PNG inside the PDF, so when the PDF is viewed on screen or printed at A4 the pixels stay crisp enough to scan from a normal phone distance (20–40 cm). For a printed QR you want at least a 25×25 mm physical size; on screen, the default 200 pt (~70 mm) leaves plenty of margin. If a scan fails, raise the error-correction level to H or increase the size slider.

On screen, 150–250 pt works for almost every URL. For print, plan the physical size around the scan distance: at 30 cm reading distance, a 25 mm square QR scans reliably; at 1 m (a poster), bump it to 80–100 mm. Longer text needs a denser QR (more modules), so its minimum readable size grows — a 50-character URL is easy at 25 mm but a 500-character vCard might need 40 mm. When in doubt, oversize. There is no downside to a slightly larger QR, and there is a hard floor below which it will not scan at all.

You can change the foreground (the dark modules) and background (the light modules) colours via the colour pickers. Keep contrast high — pure black on pure white scans best, and anything with a contrast ratio below ~4:1 will fail on cheaper phone scanners. Logo-in-the-middle QR codes (centre cut-out) are not supported here yet because they rely on H-level error correction plus a precise overlay — adding them as a separate tool is on the roadmap. If you need a branded QR today, generate the QR at error-correction H, drop it into your design tool (Figma, Canva), and overlay a small logo covering no more than 25% of the centre.

No. The QR is generated inside your browser tab by the `qrcode` JavaScript library — a single ~30 KB module that does the maths locally and returns a PNG. The PDF is then assembled by pdf-lib, also in the browser. Nothing about your text, URL, or attached PDF travels over the network. You can verify in DevTools: open the Network tab before clicking Generate, click it, and observe that no POST request with your data goes out. This matters for sensitive QR payloads — Wi-Fi credentials, internal URLs, BitLocker recovery codes — that you would not want logged on a third-party server.

Error correction is the QR's built-in tolerance for damage and obstruction. L recovers up to 7% of lost modules, M up to 15%, Q up to 25%, H up to 30%. For a clean digital PDF that you open on screen, M is plenty. For a printed QR that might get folded, smudged, or photocopied, pick Q. For a sticker that might lose a corner, or a QR with a logo overlay in the middle, pick H. Higher error correction means the QR has more modules and is denser at the same display size — so trade-off readability versus durability. The default M is the right pick 90% of the time.

The QR spec caps a single code at 4,296 alphanumeric characters or 2,953 bytes in byte mode (the mode this tool uses for URLs and Unicode text). In practice you start running into scanning issues well before that — a 1,000-character URL scans, but it produces a dense Version-40 QR that needs a 50 mm physical size and a steady hand. For long content, the right answer is usually to link to a hosted page that holds the data, and let the QR carry only the short URL. If your text is over 2,953 characters, the tool will refuse and ask you to shorten it.