I got tired of QR code generators gatekeeping SVG exports, so I built one that doesn't

Here's a small thing that's been bugging me for a while: you need a QR code, you Google "QR code generator," you land on one of the usual suspects, you customize your colors and add a logo, and then — right as you go to download — you hit a wall. PNG export is free. SVG, the format you actually need for a print job or a vector pipeline, is locked behind a paid plan. Sometimes there's a "create an account first" gate too, just for good measure.

None of this is actually expensive to provide. Generating a QR code is cheap, deterministic, and well-understood. The friction is a business model, not a technical constraint. So I built QrFast.io — a QR code generator with no signup, no paywall on exports, and full SVG/PNG downloads from the first interaction.

What it does

QrFast is intentionally narrow in scope: you give it data, you customize the look, you download a file. That's the whole loop.

No login, no account, no saved history. Open the page, generate, download, done. There's nothing to sign up for and nothing tracking your generation history.

Multiple data formats, not just plain URLs — WiFi credentials, vCards, WhatsApp deep links, Bitcoin payment URIs, and calendar events are all supported as first-class input types, each with its own structured form instead of forcing you to hand-format a string.

Full visual customization — foreground/background colors, dot styles, and logo embedding, all rendered live as you type. No "generate then preview" round trip.

SVG and PNG exports, both free, both immediate. SVG specifically matters here because raster QR codes get blurry the moment you scale them for a banner, a poster, or a packaging insert. Vector output means the code stays crisp at any size, which matters a lot more for QR codes than people expect — a soft edge on a finder pattern is exactly the kind of thing that breaks a scanner.

Why I built it this way

The honest answer is that QR generation is a solved problem from an algorithmic standpoint — error correction levels, module placement, quiet zones, all standard. The differentiation isn't in generating a QR code, it's in not getting in the user's way while they do it. So the design constraints I set for myself were:

Every interaction should feel instant. No full-page reloads when you change a color or switch from URL to WiFi mode. The preview updates in place.

Nothing should require an account. If a tool's core function is "take input, produce file," account creation is pure overhead for the user and a data liability for me.

Vector output isn't a premium feature. It's table stakes for anyone using this for print.

The stack, for the curious

The frontend is Nuxt 3, the backend is Laravel. Nuxt handles the reactive form state and live SVG preview — the QR rendering recalculates client-side as you type, so there's no network round trip just to see what a color change looks like. Laravel sits behind it for the heavier lifting: persistent format validation, any server-rendered generation paths, and keeping the whole thing fast enough that SSR doesn't become a bottleneck.

One nuance I went back and forth on: how much of the QR encoding logic should live client-side versus server-side. Client-side keeps the live preview snappy and removes a network hop, but you end up duplicating validation logic (e.g., WiFi credential escaping, vCard field formatting) on both ends if you're not careful. I landed on doing the visual rendering and live preview client-side, with format-specific payload construction validated again server-side before the final export — mostly to make sure malformed input never makes it into a generated file, especially for formats like vCard where a stray special character can produce a QR code that looks fine but fails to parse on scan.

What's next / where I'd love feedback

QrFast is live now and free to use at qrfast.io. It's still early, and there are a few things I'm actively thinking about:

Rendering edge cases — very dense payloads (long URLs, full vCards with multiple fields) push toward higher QR versions with smaller modules, and I want to make sure the dot-style customization doesn't compromise scannability at that density.

UI/UX for format switching — right now each format (WiFi, vCard, WhatsApp, Bitcoin, Event) has its own form, and I'm curious whether that's intuitive enough or whether it needs better signposting for someone who lands on the URL tab by default and doesn't realize the other formats exist.

Logo embedding and error correction — embedding a logo eats into the QR's effective error correction budget, and I'd rather over-warn users about logo size than ship a code that fails to scan in the wild.

If you try it and have thoughts — on the rendering quality, the UX, or anything that broke — I'd genuinely like to hear it. The whole point of building it free and frictionless was to get it in front of more people who'd actually use it, so real feedback from real use cases is the most useful thing anyone could send my way right now.