On today’s commercial internet, search results feel less like a map of the web and more like a crowded mall food court, where every stall is shouting for attention. Kagi’s Small Web project pushes in the opposite direction, treating a handpicked network of human-made sites as a quiet back alley where readers can finally hear themselves think.
By expanding the Small Web to mobile apps, browser extensions and a curated catalog of more than 30,000 non commercial sites, Kagi is betting that there is real demand for an internet that behaves more like a library than a slot machine. The stakes are bigger than one niche search engine: the experiment tests whether independent writing, comics and projects can still compete in a web increasingly flooded with AI generated content and enshittified search results.
The Small Web turns curation into a form of resistance
The basic pitch is deceptively simple. Instead of indexing every possible page, Kagi maintains a vetted allow list of sites that meet strict criteria: personal ownership, human authorship, minimal or no advertising and recent updates. According to reporting from techcrunch.com, the index has grown from a few thousand entries to more than 30,000 blogs, webcomics, independent video channels and personal project pages as the feature has matured and moved to mobile.
Inside the Small Web interface, users can browse by topic, tap through random picks in a shuffle mode reminiscent of StumbleUpon and filter for specific types of content like essays, comics, code repositories or photography. That framing matters. Where mainstream search has been captured by affiliate spam and SEO farms, a curated catalog is a quiet statement that not every page deserves equal prominence simply because it can be crawled.
Supporters argue that this constraint is exactly the point. Bloggers and independent developers quoted in coverage of the feature say they are seeing traffic they never expected to get in 2026, when traditional Google search is so saturated that small sites vanish to page eight overnight. By turning discovery into a deliberate choice rather than a side effect of gaming an opaque ranking system, Kagi is nudging readers toward a healthier relationship with the web itself.
But even the resistance has blind spots and trade offs
The Small Web is not a pure utopia, and critics have been quick to point out its limits. Commentators on OSNews and in independent blog reviews have raised sharp questions about Kagi’s overall approach to privacy, its reliance on external indexes and some odd business decisions, from a widely mocked T shirt marketing campaign to a messy tax catch up that forced sudden price rises.
On a more practical level, writers in the indie web community have documented frustrations with how the Small Web catalog is built. Sites often have to meet requirements around active RSS feeds and recent posts before they can be included. That policy keeps obviously abandoned projects out of the stream, but it also means that unusual or slow cadence sites — the kind of long running personal pages that defined the early web — can be excluded even when their material is timeless.
There is also the problem of scale. Kagi still licenses results from larger indexes like Bing and other providers for its main search product, and Small Web entries sit alongside those links rather than replacing them entirely. That hybrid design is pragmatic, but it highlights how dependent even a resistance style project remains on the same corporate infrastructure it is reacting against.
The indie web needs discovery tools, not just nostalgia
Supporters of the independent web have been warning for years that discovery is the missing piece. Essays on sites like creativerly.com and developer blogs focused on IndieWeb search note that while more people than ever are building personal websites, very few readers will ever stumble across them without help. Google was effectively the only discovery engine for the long tail of the web, and as its results became clogged with AI written sludge and aggressive shopping modules, that bridge quietly collapsed.
Kagi’s Small Web tries to rebuild that bridge in a more intentional way. By giving users a literal button that says show me something from the Small Web, it acknowledges that discovery is a use case in its own right, not just a side effect of typing the right magic keywords into a traditional search box. The new mobile apps extend that idea into dead time moments like commuting or waiting in line, when a quick random article or comic can replace doomscrolling a social feed.
Yet discovery alone will not save independent sites if readers still expect everything on the internet to be free, instant and frictionless. Small Web entries exist inside a paid search product that costs real money every month. That paywall is precisely what allows Kagi to avoid the ad driven incentives that hollowed out big tech search, but it also means only a subset of web users will ever see this curated indie layer at all.
What This Actually Means
Seen in isolation, Kagi’s move looks like another niche feature for power users. Taken seriously, it is closer to an argument about what kind of web we want to live with for the next decade. A subscription funded search engine that injects human scale sites into results is one of the few concrete attempts to reverse the trend toward generic, AI written, engagement bait content.
The Small Web will not topple Google or fix every problem with discovery, but it gives writers and readers a working alternative to a feed determined by whoever bought the most ads. If it succeeds, other search providers will be pressured to copy the model or at least carve out their own protected spaces for individual creators. If it fails, the lesson will not be that the indie web was a bad idea, but that it was left to a tiny company to fight battles the rest of the industry dodged.
What is Kagi’s Small Web?
Kagi’s Small Web is a curated index of personal, non commercial websites that meet a basic standard of human authorship and low friction design. Instead of scraping and ranking everything, Kagi invites site owners to opt in, then uses a mix of manual review and automated checks to decide whether a blog, comic or project belongs in the catalog.
- Sites are typically owned by individuals or very small teams, not large media conglomerates.
- Content is written, drawn or recorded by humans, with AI generated pages explicitly excluded.
- Advertising, tracking scripts and dark pattern design are heavily discouraged or disqualifying.
- Entries are organized into categories such as essays, technology, gaming, art and outdoors to help readers browse.
- The same Small Web index feeds into Kagi search results, where eligible pages can be highlighted as human first options alongside mainstream links.
Who benefits from a Small Web layer?
The most obvious winners are independent creators whose work would otherwise be buried. Personal blogs about niche programming topics, long form cultural criticism, hobbyist photography and one person webcomics all have a better chance of being found when they are not competing directly with commercial publishers and content farms.
Readers benefit as well. For people exhausted by social algorithms and engagement bait, a space where every click lands on something that exists because someone cared enough to make it is its own kind of relief. techcrunch.com framed the Small Web launch as a play for users who miss the randomness and texture of the early web; the mobile apps simply make that feeling accessible from the couch or the bus.
There is a collective benefit too. The more that tools like Kagi, indie search projects and human curated directories send traffic to small sites, the more incentives there are to keep building those sites in the first place. An internet where genuine personal pages are visible is one where it is at least possible to resist total consolidation into a handful of ad driven platforms.
How does this compare to traditional search?
Traditional search engines index and rank pages based on a mix of backlinks, engagement metrics and opaque quality signals. That approach scales, but it also creates direct incentives for publishers to chase whatever content structure the algorithm currently rewards. The result, as even mainstream outlets have documented, is an explosion of low value pages optimized for ad impressions rather than reader understanding.
Kagi’s Small Web layer inverts that logic. Instead of asking how a page performs in an attention market, it asks whether the page is worth reading in the first place. Human curation replaces pure ranking, and the subscription model stands in for ad auctions. techcrunch.com and other tech publications have been clear that this will not work for everyone, but they also note that it meaningfully changes what surfaces when you search for certain topics.
For example, a search about building a personal knowledge base or learning a niche framework is more likely to serve up someone’s detailed blog post than a generic content farm guide. That shift is not an accident; it is a direct consequence of Kagi charging users instead of advertisers and taking the Small Web concept seriously enough to invest in tools around it.