Kagi Search – Paywalled articles indicator and improved weather widget
A paid, privacy-focused search engine, Kagi, is drawing interest for features like paywalled-article indicators, customizable domain ranking, and integrated AI tools that many users say make it feel like a cleaner, more relevant version of early-2010s Google. Supporters argue the ad-free model, spam resistance, and controls such as bangs, lenses, and redirects justify the subscription, while critics question result quality in some regions, weak local business and maps integration, pricing, and long‑term sustainability of flat‑rate plans. Privacy trade-offs around account-based access and payment, plus comparisons with alternatives like DuckDuckGo, Google, GPT-based tools, and Mullvad-style anonymous models, frame a broader debate about how web search should be funded and governed.
Overall sentiment
- Many commenters say Kagi has replaced Google/DDG for them and feels like “Google circa 2011,” with fewer ads, SEO spam, and AI garbage.
- Others find it only comparable to DDG or worse than Google, especially for some use cases, and don’t see enough benefit to justify paying.
- Free trials sometimes go unused by light users who then conclude they don’t need a paid search engine.
Search quality vs other engines
- Praised for:
- More relevant, less spammy results; better for technical searches and documentation.
- Ability to promote/demote/block domains, creating highly personalized results.
- Lenses (e.g., academic, forums, PDFs) and bangs for quick context switching.
- Criticized for:
- Weaker local/business searches, maps, shopping, and some “instant answer”/widget scenarios.
- Occasionally missing niche domains or giving irrelevant foreign Amazon/regional results.
- Some users outside the US report latency and odd regionalization (e.g., UK vs AU, Germany vs international).
Paywall indicator & content payments
- Paywall indicator is seen as useful for saving time and avoiding surprise walls; some see it as also highlighting potentially higher‑quality content.
- Several wish for an ecosystem where a monthly pool (e.g., $20) is automatically split among sites visited, ideally integrated with Kagi/Orion.
- Others mention redirect rules to send paywalled links to archive sites, though rate limits and ethics aren’t deeply discussed.
AI & advanced features
- FastGPT and the Universal Summarizer get strong praise; citations are seen as a differentiator vs other chatbots.
- Some say LLM-style tools (ChatGPT, Perplexity) have already replaced most traditional search for them, so Kagi feels like a step back.
- Kagi staff note LLM costs are now relatively small compared to core search.
Pricing, quotas, and sustainability
- Debate over value: $10/month unlimited is worth it for heavy users; light users wish for cheaper or metered plans.
- “Range anxiety” over limited-search tiers and confusing trial consumption (e.g., “More results” using another query) frustrates some.
- Concern from some that flat-rate, feature-rich plans may be financially unsustainable; Kagi staff respond that unit economics are sound and pricing has adjusted as costs dropped and user mix broadened.
Privacy & identity
- Kagi states it does not log/tie searches to accounts; history is opt‑in and currently disabled.
- Some want a Mullvad-style, account-number model without email/password; others think existing options (burner email, bitcoin) are sufficient.
- A few users are uneasy about needing to sign up at all for a “privacy-focused” search engine.