Zippy is open source — here's the whole engine
The redirect engine that springs the real app instead of the in-app browser is AGPL on GitHub. Why we opened it, what's in the box, and how to self-host it.
The core of Zippy — the redirect engine that takes a tap inside Instagram's webview and springs the real native app — is open source, under AGPL, at github.com/zippylink/zippy. Not a demo repo. Not "open core" with the good parts held back. The engine that serves our production redirects is the engine in the repo.
⚡ Zippy: the whole engine, in the open. no "email us for the good parts." zip it, fork it, keep it.
Why open the thing we sell
A short link is a trust product with a brutal failure mode: if the service behind it dies, every link you ever posted dies with it — in old posts, in bios, in printed QR codes you can't recall. The graveyard of dead shorteners is long, and everyone who's been burned remembers.
Open source is the strongest permanence guarantee we can make. Our pricing page says links never expire; the repo is what makes that promise survivable even in the worst case. If we vanished tomorrow, you could stand up the engine and point your domain at it. Your links outlive us. That's the deal.
There's a second reason, less noble and just as true: deeplink schemes are a moving target, and a community that self-hosts is a community that notices when TikTok quietly changes its Android intent format. Fixes land in the engine, and everybody's links — hosted and self-hosted — keep working.
What's in the box
- The redirect engine. Platform detection (which webview is this tap coming from?), the per-platform deeplink routing table (iOS Universal Links, Android intents, URL schemes), and clean web fallbacks when the native app isn't installed.
- Link management. Slugs, custom slugs, the works.
- The routing table itself. The accumulated, annoying, hard-won knowledge of how to escape each platform's webview — LinkedIn, Instagram, WhatsApp, Reddit, Product Hunt, YouTube, TikTok, X — versioned in the open where it can be reviewed and fixed.
Why AGPL
Plain GPL wasn't enough for a server-side product: a company could take the engine, run it as a closed hosted service, and give nothing back — the classic SaaS loophole. AGPL closes it. Run the engine as a service, and your modifications must be shared back with the people using it.
For you, self-hosting for yourself or your company: nothing to worry about. Run it, modify it, keep your changes private as long as you're not offering it as a service. The license targets exactly one actor — someone reselling a closed fork of the engine — and nobody else.
⚡ Zippy: run me, tweak me, keep it to yourself — all good. the AGPL only bites if you resell a closed-up me. don't be that guy.
Self-hosting in one paragraph
Clone the repo, follow README → self-host quickstart: bring your own domain, run the
engine, point your DNS at it. Everything the redirect path needs ships in the repo.
When the routing table updates upstream, pull and redeploy — that's the whole
maintenance story.
So why pay for the cloud?
Because the hosted cloud just means you never have to. zipthe.link runs the same engine with the ops handled — the always-current routing table, the analytics (geo, device, platform, referrer), custom slugs, QR codes, editable "living links", uptime someone else gets paged for. Self-hosting gets you the redirects; the cloud gets you the product around them.
Both paths keep your links alive forever. One of them you have to babysit.
⚡ Zippy: self-host = free redirects + you hold the pager. cloud = same me, someone else holds the pager. pick your fighter.
FAQ
Is Zippy really open source?
Yes. The redirect engine that serves Zippy's production traffic is AGPL-licensed on GitHub at github.com/zippylink/zippy — the same code, not a stripped "community edition." Platform detection, the deeplink routing table, and link management all ship in the repo.
What does the AGPL license let me do?
Run it, modify it, and self-host it for yourself or your company, keeping your changes private. The AGPL only requires sharing modifications when you offer the engine to others as a hosted service. For everyone self-hosting for their own use, there's nothing to publish.
Can I self-host Zippy for free?
Yes. Clone the repo, bring your own domain, run the engine, and point your DNS at it — the whole redirect path ships in the box. You pay for the hosted cloud only if you'd rather not maintain the routing table and uptime yourself.
What do I lose by self-hosting instead of using the cloud?
The ops. Self-hosting gets you the redirects; you keep the routing table current as platforms change their deeplink schemes, and you own uptime. The hosted cloud runs the same engine with the always-current routing table, analytics, custom slugs, QR codes, and editable "living links" handled for you.
Star the repo, read the engine, file the issue when a platform breaks something. The routing table is better with more eyes on it.