Kagi search reached 20k paying members

Original Article ↗ Hacker News Discussion ↗

Privacy-focused search engine Kagi has reached 20,000 paying users, prompting debate over whether a subscription-based, ad-free search model can remain financially sustainable at its current scale. Commenters generally praise Kagi’s search quality, customization features, and lack of SEO spam compared to Google and DuckDuckGo, but question choices like spending a large share of recent investment on free t‑shirts for early customers and partially relying on third-party indexes such as Brave, Bing, Google, and Microsoft. Many see Kagi’s small, paid user base as both a protective moat against spam and a limitation on growth, with some arguing that its approach only works as a niche alternative rather than a mass-market rival to dominant search engines.

Overall Sentiment & Adoption

  • Many commenters are paying users and report being very satisfied; several say it’s the first time they’ve been able to fully leave Google or DuckDuckGo.
  • Others tried Kagi and didn’t see enough benefit over DDG/Startpage or now rely mostly on LLMs instead of search.
  • Several stress that a paid, ad‑free search product “for power users” is appealing amid perceived “enshittification” of big platforms.

Search Quality & Features

  • Common praise: results feel like “old Google,” with less SEO spam and better relevance, especially for technical queries.
  • Highly valued features: pin/boost/downrank or block specific domains, lenses like “Small Web,” URL rewrite rules (e.g., redirect Reddit/Twitter), Vim keybindings, Quick Peek, and stats/leaderboards of most‑blocked sites.
  • Some users report a noticeable drop in quality and slower results recently, plus more AI‑generated spam, and a few reverted to self‑hosted search.

Business Model, Pricing & Sustainability

  • Kagi is said to have ~20k paying users and ~430k queries/day, roughly 20–30 searches per user.
  • Back‑of‑the‑envelope calculations (using Kagi’s past financial blog) suggest per‑query API costs are significant; commenters argue that ~$2/month pricing would be unviable.
  • A Kagi representative states they are close to breakeven, have ~4+ years of runway, and are largely bootstrapped aside from a ~$670k SAFE round.
  • Some find $10/month steep or unnecessary given low personal search volume; others say the time saved easily justifies the cost.
  • There is a 100‑search free trial, but some find it too limited; others note that overage beyond plans is only a small per‑query fee.

T‑Shirt Campaign Debate

  • Major thread: concern that spending roughly low‑six‑figures and building a full merch pipeline (company, warehouse, logistics, site) is a distraction and wasteful, especially given early‑stage economics.
  • Defenders see it as:
    • Customer appreciation instead of ad spend, potentially reducing churn and strengthening word‑of‑mouth.
    • Morale‑boosting for the team and community.
  • Design choice (mascot dog without prominent “Kagi” text) is polarizing:
    • Some say it’s a missed branding opportunity; others say they only wear shirts without overt logos and will happily evangelize when asked.
  • Users must claim shirts via credits and pay shipping; unclaimed stock may later become general merch.

Indexing Partners & Politics

  • Kagi uses multiple external indexes (including Brave and Bing); some users strongly dislike the Brave association for political or ethical reasons, others explicitly don’t care and focus only on quality/cost.
  • A subset blames quality regressions on the Brave API; others note costs may have driven this integration and hope further index diversification improves results.

Branding, UX, and Ecosystem

  • Mixed reactions to the yellow “g” logo: some find it instantly recognizable in tabs, others find it confusing or too reminiscent of Google.
  • Some appreciate tight integration with privacy‑oriented browsers (e.g., Orion) and the inclusion of AI tools via special bangs like !quick, !fast, !code, and !expert in higher tiers.