There's a certain irony in how search engines have evolved. They were supposed to help us find things. Now, finding one that doesn't track every query feels like the harder task.
Enter Uruky, a Portuguese search operation run by a husband-and-wife team who believe privacy tools don't need sprawling feature sets to justify their existence. The company recently unveiled Image Search and URL Rewrites—two features that, judging by the reception on Hacker News, struck a nerve with developers tired of compromising on privacy. The announcement drew 231 points and 220 comments on the Show page, with founder Bruno Bernardino self-reporting that the service had crossed 100 monthly active accounts.
Not a staggering number, perhaps. But for a metasearch engine operating out of Portugal with EU-only infrastructure and a deliberate rejection of AI, it signals something: there's an audience for search that doesn't try to be everything at once.
The Kagi Question
Bruno and his wife Patrícia, who operate Uruky through their Portuguese company I'm Blue, Lda., aren't exactly coy about their positioning. Their website includes a dedicated page titled "Kagi alternative," laying out the differences between their €5-per-month service and Kagi's $10 unlimited plan.
Kagi, for context, has become something of a gold standard among privacy-minded users—a US-based paid search engine that's expanded into browsers, AI summarizers, and a growing ecosystem of tools. It launched URL Redirects earlier this year, a feature that automatically swaps links in search results with privacy-friendly alternatives.
Uruky's version does much the same thing. Want every x.com link replaced with xcancel.com before you click? The URL Rewrites feature handles that across all search results, not just social media. It's the kind of quality-of-life improvement that matters if you've already done the work of cataloging privacy-respecting frontends but would rather not manually redirect every time.
Where Uruky diverges is in what it doesn't offer. No browser. No AI-powered anything. Bruno stated flatly in the Hacker News thread: "We won't have Generative AI." That's positioning, certainly, but it also reflects a bet that some users are fatigued by the race to bolt AI onto every product.
Image Search, Proxied
The other new feature, Image Search, exited beta in June after a May trial run. Uruky proxies all images through its own servers, pulling from Pixabay and Serper—Pexels was included during the beta phase but later removed—meaning your IP address never reaches those third-party providers. Images load in search results without pinging external trackers.
It's not groundbreaking technology. But it's the kind of thoughtful implementation that matters when the goal is keeping data within EU borders and out of tracking networks.
Both features include a date filter, controllable via settings or a "w" query parameter. Small touches, maybe, but they add up for users who've grown accustomed to search engines that prioritize engagement over utility.
Traction, By the Numbers
Uruky launched its public beta in January 2026, starting with Mojeek and Marginalia as search providers. By late spring, Bruno self-reported in an interview with The Privacy Dad that the service had 60 monthly paid accounts. The milestone of 100 monthly active accounts came shortly after.
Revenue figures remain undisclosed, but the model is straightforward: €5 buys a month of unlimited searches, paid via one-time top-up through Mollie, an EU payment processor. No auto-renewing subscriptions. A 14-day money-back guarantee. A two-hour free trial gated by a proof-of-work captcha, added in May to deter abuse.
For customers who stick around—those who've paid for a full year—there's an unusual offer: request a copy of the codebase under NDA. The plan, as of the recent announcement, is to shift to a PolyForm Shield 1.0.0 license within six months. Source-available, but not open source.
It's a curious middle ground. Transparent enough to satisfy the privacy-conscious crowd, but not so open as to invite clones.
The Upstream Dependency
Uruky is, fundamentally, a metasearch engine. It aggregates results from providers like Mojeek, Marginalia, EUSP (the Ecosia/Qwant partnership), Linkup, and Serper. That means it's beholden to those upstream sources, though users can prioritize which providers to query first.
The company also runs a small opt-in index called Uruky Site Search, though Bruno acknowledged in the Hacker News thread that it's not yet robust enough to serve as a primary provider. The goal is around 100 million URLs—an indieweb index rather than a Google replacement.
There's a separate product by the same name, priced at €30 for six months, aimed at site owners who want white-label, privacy-first internal search. It's a side business, really, though one that underscores the company's deliberate focus: search and search-adjacent tools, nothing more.
No Logs, No Plans for Expansion
Uruky doesn't log searches. Doesn't collect personal data beyond an account number, Mullvad-style. Keeps all data processing within the EU. The privacy policy was updated in early February, the terms in late January, both reflecting the no-tracking commitments and EU-only storage.
There are no plans for email, cloud storage, or the kind of ecosystem expansion Kagi has pursued. Bruno and Patrícia seem content—or at least committed—to keeping the operation narrow.
Whether that's sustainable long-term is an open question. A hundred paying users might cover costs for a small team in Portugal, but it's not exactly venture-scale growth. Then again, maybe that's the point. Not every product needs to become a platform.
For now, Uruky occupies a specific niche: EU-based, AI-free, search-focused. Five euros a month, no frills, no apologies. Whether the market for that is 100 users or 10,000 remains to be seen. But the Hacker News thread suggests there's at least curiosity—and perhaps more than the founders expected—for a search engine that simply searches.
