Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I don't know how to make this more meaningful than just an anecdote, but I love the idea of Kagi but just cancelled my subscription. All the issues about google search becoming more and more useless are absolutely true, but I still continue to get much better results for most topics with Google than Kagi. Same for Kagi's LLM products compared to directly using Claude or others.


I agree, but they aren't _that_ much better and I'd rather do a little more digging than support Google. I still find what I need and make good money. I can also rationalize it that I'm not always just taking the first result and actually reading multiple solutions to my problem but yanno, YMMV there.


My rule of thumb that I’ve found is that for the most part Kagi is on par with Google for anything non location-aware. Oftentimes it’s noticeably better actually since there’s less pumping of useless results to the top. For anything location-aware, however, it ranges from significantly worse to steaming hot pile of garbage depending on the search.

That’s what eventually made me cancel. Loved it for what it does give but found myself rather annoyed about having to switch to Google to get actual usable results on the significant amount of location based queries. Never got into the habit of using bangs unfortunately.

FWIW DDG is flawed in nearly the exact same way.


I am wondering if you reported any of these issues/example queries to Kagi/kagifeedback.org?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: