I've been using the Kagi search engine for months now and I'm not impressed. I bought into it because there were a lot of posts saying that it was "just like old Google" but this has not been my experience. It's the same as new Google, you can type in what you're looking for exactly and you'll get random sort-of related websites.
I remember when you could half-remember a comment from a website, type that into Google, and get taken to the article you were looking for. That was back in like 2010. To me that's the old, and useful, search engine that I want.
olejorgenb 40 minutes ago [-]
Off topic, but ref
"I remember when you could half-remember a comment from a website, type that into Google, and get taken to the article you were looking for"
It's funny to me that (to my knowledge) no browser (mainstream?) implement this functionality yet. Seems like a no brainer to index what the user have actually seen... (Could even be restricted based on viewport - I don't think it's that crazy of an idea)
I know there's a a number of third party programs which does though. Of course - multi-device being the norm - complicates things.
hirako2000 7 minutes ago [-]
That's called search: in history.
catlover76 26 minutes ago [-]
[dead]
the_pwner224 2 hours ago [-]
I switched about a year ago. At the time it did seem like a step up from Google results. But there's been an increasing prevalence of low quality results. Blogspam, AI websites, etc. Obviously not blaming Kagi here, web search has gotten hard recently.
Is Kagi still better than Google? Probably, I don't really know because I don't use Google anymore. But at this point I feel like I'm with them out of inertia more than being an avid supporter. One of these days I'll re-evaluate Google and decide whether to switch back or not.
It does occasionally surface interesting results from small sites that you wouldn't get on Google. I do find that to be useful.
Kagi definitely isn't a bad search engine by any means. Honestly if you haven't used it, try the 100 search free trial on one device. Maybe you'll like it. This feels more like a general decline of the open web.
ArtificeAccount 2 hours ago [-]
I feel the same way. I'm probably going to end my subscription at some point, but right now the effort involved is what's keeping me with Kagi.
8cvor6j844qw_d6 2 hours ago [-]
Is usage based pricing available?
I'm thinking of trying it out Kagi, but adding another monthly commitment is what's holding me back.
A single credit top-up and occasional usage until the credits run out sounds good to me.
*Do you plan to allow purchasing privacy pass tokens without having an account?*
Yes, this makes sense. This is possible because technically the extension does not care if you have an account or not. It just needs to be 'loaded' with valid tokens. And you can imagine a mechanism where you could also anonymously purchase them, eg. with monero, without ever creating an account at Kagi. Let us know *here* ( https://kagifeedback.org/d/6163-kagi-privacy-pass ) if you are excited about this, as it will help prioritize it.
Personally I don't like being signed in during searches, this seems like a good solution.
mpalmer 1 hours ago [-]
Everyone has to answer for themselves why they would be OK with Google hoovering up their data in order to deliver substandard results, vs Kagi actively working to remove low-quality results all while collecting no personal data.
jmye 1 hours ago [-]
Yes. I use both (Google only at work) and Kagi is certainly no worse and comes with the massive benefit of simply not being Google. It's worth paying for for that reason alone, even if the engineers at Google are constantly working on making sure I'm tracked anyway.
sheepscreek 2 hours ago [-]
It’s definitely not Kagi’s fault. The AI slop is simply taking effect and I feel sorry for them. I never expected them to match Google’s quality, but I was impressed with how close it was when I used it a few years ago.
I've been using Kagi for ~18months and your description doesn't match my experience at all.
Querying for something like "snowflake json from variant?" in both engines and in google I get a sort-of-right-but-not-really-that-helpful ai summary about "parse_json" function. In Kagi I get an actually useful summary with code examples of parse_json, but also the colon-based syntax for accessing values inside nested objects without needing to parse anything.
I very rarely need to go into a page, I use Kagi quick search summary with the "?" suffix and it almost always gives me a useful answer in one-shot.
singpolyma3 1 hours ago [-]
But then you're not using kagi search just the LLM
mpalmer 1 hours ago [-]
No, the responses are backed by searches.
0xdeafbeef 49 minutes ago [-]
Try g.ai. It's stupid fast and uses google indexes. Kagi? sometimes doesn't correctly parse intent, in Google thing you can just ask function doing this and gives you it, with examples, grounding and extremely fast. I'm paying for kagi since the begging and I guess id cancel it because it gives not so much added value
DuzAwe 51 minutes ago [-]
I’ve been using kagi for about eight months now as well and at least in Europe it’s a significantly better search engine than Google by a long shot. The results are significantly more accurate. I don’t get listicles I don’t get AI spam. I get what I’m searching for, it’s refreshing.
The assistant is a nice addition but it’s search is superior for me.
BrunoBernardino 21 minutes ago [-]
I won't add links so it doesn't look like I'm spamming or promoting a service (though I am, but it seems in line with what you're talking about), but there's a product I've built with my wife which has made things a little bit better in our experience because it gives you an option to choose different providers/indexes, thus tailor results to your personal preference. You can find it from my personal website (my username . com).
muyuu 58 minutes ago [-]
this is hard to evaluate, but we cannot replicate the old web search experience not just of Google, but Altavista, Lycos or Yahoo, when most of the web is siloed and increasingly botted - simply because the stuff you see in the siloed internet is actively "protected" out of your control
perhaps the best we can do is this "small web" thing which can be seen as some sort of retrofuturistic solution, but of course the siloed internet is a black hole of content and effort, and of course if the small web gets enough traction, astroturfed generative AI content will target it
spicycode 1 hours ago [-]
I've switched as of a few years back and it definitely works like pre-AI/search index degradation for me. But I def understand search is very user specific based on how you search and what you are targeting.
alpineman 18 minutes ago [-]
Disagree, I love it, at least as good as Google
LatticeAnimal 8 minutes ago [-]
I think that's the problem. I used to find it far superior to google. Now, there are a lot of queries where I am unimpressed with the results and end up trying google just to get better results. (like I used to do with DDG)
I've had a few experiences now where someone is standing over my shoulder asking me to look something up, and I search kagi, find nothing, then search google and find what they asked me to look up. Then when they ask "what was that other search engine you used first?" I don't feel compelled to vouch for kagi :(.
windowliker 2 hours ago [-]
Funny to look back and recall how useful web search actually was at one point. Ahh the good old days.
ArtificeAccount 2 hours ago [-]
The enshittification of everything has really put a damper on the technological optimism that seemed to be the norm back then.
rusty-peanuts 2 hours ago [-]
[dead]
cjohnson318 41 minutes ago [-]
I typed in my dentist's full business name and location, "<name> family dentistry <city> <state>", and it was still #5 in the results. I still, out of habit, tapped the first link and called that number instead. It's ludicrous. In 2010 that would have been the top hit, next to the Wikipedia page on dentistry.
hrmtst93837 2 hours ago [-]
What you describe sounds more like a large ElasticSearch like full-text index over the entire internet.
ArtificeAccount 2 hours ago [-]
Whatever it was called, it was way better than whatever Google is doing currently.
sodapopcan 2 hours ago [-]
In fairness, there wasn't the level of utter crap back then that there is now. Not that there wasn't a LOT back then, but there is even more now.
112233 41 minutes ago [-]
How comes yandex.com can show me results that contain my search term? Most egregious example: I am searching for name of abandoned blogspot domain: yandex shows me 1 result, which is that domain. Goolge shows me "no results" fishing monster. Blogspot is google service !!!
ArtificeAccount 2 hours ago [-]
People say that, but on the other hand companies like Google have a lot of much better ways of categorizing things now than they did in the past. I'm not sure I buy the excuse of "gosh, it's just too hard for us :(" from this international company worth trillions employing geniuses.
It really feels either intentional or egregiously incompetent.
sodapopcan 1 hours ago [-]
That's fair.
littlestymaar 2 hours ago [-]
We literally have internal emails[1] showing that Google simply gives no shit about the quality of its search engine's results.
I feel like I can still do this with Google if I use quotes.
Kagi I've been using and it's fine. Better than DDG for sure. But sometimes I still go back to google to find something kagi is struggling with.
user3939382 2 hours ago [-]
For over two years I’ve maintained the practice of using Kagi and falling back to Google if I couldn’t find something. I can count the number of successes doing that on one hand. In the meantime I get to support a company which actually respects me as a user and isn’t doing things like tying accounts to browsers, AMP (trying to take over the web), trying to kill adblock, etc.
Terretta 2 hours ago [-]
In comparisons (often shared here) among SERPs, kagi has tended to have fewer blatant results campers crowding out original authoritative sources.
And yes, Google's founders were right that web ads would kill that experience you want.
freetonik 3 hours ago [-]
On a similar note, I maintain and grow a manually curated collection of personal blogs with valid RSS feeds: https://minifeed.net/blogs
The criteria is simple: human-written (as much as I can validate myself), in English (for now), with valid RSS feed, and not a micro-blog (so, more than just feed of links or short tweet-like messages).
Similar to Kagi's Small Web viewer, or StumbleUpon-style viewer: you can get a random listing of blogs [1] or a random listing of posts from all blogs [2]. Feeds and posts are indexed, so full-text search works across all blogs. When possible and permitted by robots.txt, text is scraped for searching, so even if some text is omitted in the RSS feed by the author, search should work.
Though I do plan to implement a similar "view one random post at source" kind of view, soon.
UPD: Feel free to submit a blog, including your own! [3]
Jokes aside, it's really nice and I can totally see becoming addictive. Kudos to Kagi team for an other user oriented product. (as a side note, I am using Kagi daily and i didn't know about this tool)
coopreme 1 hours ago [-]
SU was amazing. Stumbling upon Ze Frank? It was internet gold!
erremerre 4 hours ago [-]
I like the idea, but would like to be able to select a language and see the small web of that language. There are more languages than English, and this tool could make them thrive.
Also somehow if they are clever, they could use this for those translation system they are using, but please let us select our own language without feeding automatic translation like youtube does).
8organicbits 3 hours ago [-]
I think the problem is that it's hard to curate feeds in a language you don't understand. I've been building an uncurated index of OPML blogrolls, with no language restriction. The OPML blogrolls are curated by their owners, so someone decided they met some inclusion criteria, but the overall list is uncurated.
The first random page it returned to me was this — https://gaultier.github.io/blog/how_to_make_your_own_static_... — which was about building one's own static site generator, which I really liked. I did not realise when I closed that page how hard it would be to find it again, because, of course every new visit to Kagi returns a different page :-)
arscan 3 hours ago [-]
I do love the concept, but a little part of me died
each time I came across an article with a very strong AI voice. That just feels antithetical to the ‘small web’ ethos because it obscures the ‘neighbor’ behind it.
windowliker 2 hours ago [-]
Welcome to 2026 when the next door neighbour is an AI datacentre using up all your groundwater.
hristian 2 hours ago [-]
[dead]
input_sh 2 hours ago [-]
Could've at least checked if the website even allows embedding before embedding it, I found two by randomly clicking around that don't.
codethief 2 hours ago [-]
Yeah, many links in the embedded blog posts don't work either, presumably because the target website doesn't allow embedding. On mobile I always have to open them in a new tab for them to work.
lelanthran 23 minutes ago [-]
This is fantastic - how long do I have to press F5 before my blog shows up?
:-)
rusakov-field 50 minutes ago [-]
My blog was getting traffic from that domain! So that is what that is.
emehex 4 hours ago [-]
StumbleUpon?
rdmuser 3 hours ago [-]
Personally my favorite spiritual successor to stumbleupon has been cloudhiker.net. I found kagis to be too personal blog focused for my tastes. I love that kagi is doing so much of this out in the open though.
How do we keep getting surprised by enshittification!?
The worst case scenario is that AI runs everything, we have no skills, and are completely dependent on it...and it shows us crummy commercials and subtly steers us to paid placement with no recourse whatsoever. I hate this possible future, but this is where the money will lead.
unbindableisaac 3 hours ago [-]
Bit bummed. The first random page I landed on was a really interesting article for me. The custom cursor (well why not) had me struggling to following a link, and instinctively I refreshed the page. I ended up somewhere else in the haystack with ostensibly no way back to that particular article.
Perhaps I'm yelling into the void here, but what would be great is when first landing at kagi.com/smallweb, the url query parameter would be somehow set, as it is when "Next Post" is clicked.
bjord 3 hours ago [-]
doesn't solve the root problem, but maybe try searching for the topic in kagi with the small web lens?
unbindableisaac 2 hours ago [-]
I think it would, so long as the redirected URL with the search parameter was diarized into browser history. It would however introduce a behavior change that may be undesired (users need to know to press "Next Post" instead of refreshing).
In any case, my Kagi search for the article containing the memorable phrase "rare as rocking-horse s*t" came up empty. Perhaps it's not yet been indexed.
sam_goody 3 hours ago [-]
So, basically, a random site from their index of ~30,000 sites.
You can choose similar sites by index.
But what are the criterion to have your site listed here, or how it will prevent this from just becoming a massive gamified advertising index, or anything more about "why these?" is not obvious to me.
Can anyone explain what is special about these sites specifically, or where this project is going?
I run a Hugo blog and I get more interesting referral traffic from Kagi's small web index than from Google at this point. 5,000 curated sites is small enough to be useful most "indie web" directories are graveyards unfortunately..
4 hours ago [-]
apples_oranges 3 hours ago [-]
A bit off topic, but I noticed I hardly ever use search anymore. It's just google.com/ai in 99% of cases. I believe in the future, search engines must go in this direction ..
myth_drannon 1 hours ago [-]
Too many AI or relegion related sites.
bryanhogan 13 minutes ago [-]
Tried it myself, first three I got were also about AI.
Curious if there are any statistics on which topics people are writing about.
jwelten 3 hours ago [-]
Interesting, really like the idea. Maybe in the future a possibility to use it in multiple languages
4 hours ago [-]
4 hours ago [-]
WhereIsTheTruth 3 hours ago [-]
Kagi wants to exist in a world that doesn't need it anymore
freetonik 3 hours ago [-]
We can create worlds, not just inhabit ones created by corporations.
I remember when you could half-remember a comment from a website, type that into Google, and get taken to the article you were looking for. That was back in like 2010. To me that's the old, and useful, search engine that I want.
"I remember when you could half-remember a comment from a website, type that into Google, and get taken to the article you were looking for"
It's funny to me that (to my knowledge) no browser (mainstream?) implement this functionality yet. Seems like a no brainer to index what the user have actually seen... (Could even be restricted based on viewport - I don't think it's that crazy of an idea)
I know there's a a number of third party programs which does though. Of course - multi-device being the norm - complicates things.
Is Kagi still better than Google? Probably, I don't really know because I don't use Google anymore. But at this point I feel like I'm with them out of inertia more than being an avid supporter. One of these days I'll re-evaluate Google and decide whether to switch back or not.
It does occasionally surface interesting results from small sites that you wouldn't get on Google. I do find that to be useful.
Kagi definitely isn't a bad search engine by any means. Honestly if you haven't used it, try the 100 search free trial on one device. Maybe you'll like it. This feels more like a general decline of the open web.
I'm thinking of trying it out Kagi, but adding another monthly commitment is what's holding me back.
A single credit top-up and occasional usage until the credits run out sounds good to me.
Also, from the Kagi privacy pass FAQ at https://blog.kagi.com/kagi-privacy-pass#faq:
Personally I don't like being signed in during searches, this seems like a good solution.Querying for something like "snowflake json from variant?" in both engines and in google I get a sort-of-right-but-not-really-that-helpful ai summary about "parse_json" function. In Kagi I get an actually useful summary with code examples of parse_json, but also the colon-based syntax for accessing values inside nested objects without needing to parse anything.
I very rarely need to go into a page, I use Kagi quick search summary with the "?" suffix and it almost always gives me a useful answer in one-shot.
The assistant is a nice addition but it’s search is superior for me.
perhaps the best we can do is this "small web" thing which can be seen as some sort of retrofuturistic solution, but of course the siloed internet is a black hole of content and effort, and of course if the small web gets enough traction, astroturfed generative AI content will target it
I've had a few experiences now where someone is standing over my shoulder asking me to look something up, and I search kagi, find nothing, then search google and find what they asked me to look up. Then when they ask "what was that other search engine you used first?" I don't feel compelled to vouch for kagi :(.
It really feels either intentional or egregiously incompetent.
[1]: [The Man Who Killed Google Search](https://www.wheresyoured.at/the-men-who-killed-google/)
Kagi I've been using and it's fine. Better than DDG for sure. But sometimes I still go back to google to find something kagi is struggling with.
And yes, Google's founders were right that web ads would kill that experience you want.
The criteria is simple: human-written (as much as I can validate myself), in English (for now), with valid RSS feed, and not a micro-blog (so, more than just feed of links or short tweet-like messages).
Similar to Kagi's Small Web viewer, or StumbleUpon-style viewer: you can get a random listing of blogs [1] or a random listing of posts from all blogs [2]. Feeds and posts are indexed, so full-text search works across all blogs. When possible and permitted by robots.txt, text is scraped for searching, so even if some text is omitted in the RSS feed by the author, search should work.
Though I do plan to implement a similar "view one random post at source" kind of view, soon.
UPD: Feel free to submit a blog, including your own! [3]
[1] https://minifeed.net/blogs/by/random
[2] https://minifeed.net/global/random
[3] https://minifeed.net/suggest
https://github.com/kagisearch/smallweb/blob/main/smallweb.tx...
There is also Small Comic:
https://kagi.com/smallweb/?comic
https://github.com/kagisearch/smallweb/blob/main/smallcomic....
And Small YouTube:
https://kagi.com/smallweb/?yt
https://github.com/kagisearch/smallweb/blob/main/smallyt.txt
https://hcker.news/?smallweb=true
https://news.ycombinator.com/item?id=46618714 (Ask HN: Share your personal website, 2414 comments)
Jokes aside, it's really nice and I can totally see becoming addictive. Kudos to Kagi team for an other user oriented product. (as a side note, I am using Kagi daily and i didn't know about this tool)
Also somehow if they are clever, they could use this for those translation system they are using, but please let us select our own language without feeding automatic translation like youtube does).
https://alexsci.com/rss-blogroll-network/
:-)
There are a surprising amount out there: https://blog.woblick.dev/en/2025/best-stumbleupon-alternativ...
http://cloudhiker.net
https://www.offscopes.com
Newsletter version if you prefer: https://randomdailyurls.com
Previous post 7-sept-2023 https://news.ycombinator.com/item?id=37420281 185 comments. And https://news.ycombinator.com/item?id=39476015 23-feb-2023 36 comments
https://blog.kagi.com/small-web
The worst case scenario is that AI runs everything, we have no skills, and are completely dependent on it...and it shows us crummy commercials and subtly steers us to paid placement with no recourse whatsoever. I hate this possible future, but this is where the money will lead.
Perhaps I'm yelling into the void here, but what would be great is when first landing at kagi.com/smallweb, the url query parameter would be somehow set, as it is when "Next Post" is clicked.
In any case, my Kagi search for the article containing the memorable phrase "rare as rocking-horse s*t" came up empty. Perhaps it's not yet been indexed.
You can choose similar sites by index.
But what are the criterion to have your site listed here, or how it will prevent this from just becoming a massive gamified advertising index, or anything more about "why these?" is not obvious to me.
Can anyone explain what is special about these sites specifically, or where this project is going?
Curious if there are any statistics on which topics people are writing about.