It is becoming near impossible to find relevant information from search engines. Duckduckgo, SearXNG, Bing, Google, and so many more mainstream engines have a significantly high noise to signal ratio, and it is getting worse.
Here are a collection of the best search engines I know, please add more to the list.
- Forum Search Engine: https://crowdview.ai/
- Non-commercial Search: https://search.marginalia.nu/
- Libre Meta Search Engine: https://librey.devol.it/
- Golden Age Search Engine: https://www.wiby.org/
- Yandex: https://yandex.com/
If no more high quality search engines exist, would it be possible to host your own?
EDIT: Some new discoveries. The addon uBlacklist and filters can block super SEO sites from appearing in search.
I find all the kagi mentions to be very suspicious
It’s not, really, I switched from Google some years ago and had accepted my faith with DuckDuckGo, but then tried out Kagi. I use search so much daily for work, the relief of getting quality results again is immense and probably saves me hours per week. I get much better results from Kagi than I got at the end from Google, and I can tune them to my liking:
- block Pinterest results when I search for images,
- downprioritize shopping results,
- rewrite all Reddit links to go to old.reddit.com,
- unamp google AMP links
- summarize long texts / documents
- quick answer from the top 5 results
…and so on and so on. It’s just so effective.
… suspicious
You can redirect old reddit with the tampermonkey extension
Yeah there are some good solutions to achieve this, I’m a big fan of the libredirect project, but currently I’m just setting up the redirects I want directly in Kagi so any URLs in the search results are already rewritten to my liking.
I always fear it comes across that way when I recommend it to people here. I’m just a very happy user and want to see them succeed.
I find it expensive for what it is (given that I still get a limited number of searches) and I’m not comfortable with some of their ways (I don’t want anything to do with AI, and I the idea they have of being nonpolitical seems dangerously naive to me). I also don’t like supporting non-FOSS projects all that much.
Still, it’s the best search I’ve found, and I’m paying every month until I find something better. It’s worth it.
If I had stock/investments in a search engine, you better fucking believe Id also have a bunch of bots crawling for the terms “What is the best search engine” and immediately hijack the convo with bots upvoting my search engine.
I cannot explain how easy it is to do this.
Your comment belongs higher. When given the opportunity to make money by social media advertising sometimes in the thousands or millions, companies, share holders, and conflict of interest groups take it. Cablemod’s burning adapters, cryptocurrency scams, payed positive youtube reviews are some great examples. In general there is no honor system and its best to assume anything that can be abused will.
Also manipulating votes is incredibly effective towards swaying public opinion due to the bandwagon effect. Spend a days worth of effort making fake accounts and downvoting any opinion you see as undesirable and most people will follow suit. This is especially bad in echo chambers like on twitter, reddit, etc.
I wish a broader audience could be aware of this. The best I can do is try to spread the word.
Sources:
- The Problem with Linus Tech Tips: Accuracy, Ethics, & Responsibility: https://www.youtube.com/watch?v=FGW3TPytTjc
- Cablemod Recall: https://old.reddit.com/r/cablemod/comments/18o7bnv/planned_voluntary_safety_recall_of_cablemod/
- FTX Scandal: https://en.wikipedia.org/wiki/Bankruptcy_of_FTX
- Bandwagon Effect: https://en.wikipedia.org/wiki/Bandwagon_effect
- Echo Chamber: https://en.wikipedia.org/wiki/Echo_chamber_(media)
same. specially considering how privacy invasive kagi is.
How is it privacy invasive?
I know nothing about Kagi, except that it requires an account and is paid. So I assume all searches are tied to that account and there is no way to do an anonymous search.Disregard I’m a fool
We care about data protection: We will be good stewards of any personal information you share with us. We do not log or associate searches with an account. More at our privacy policy.
Literally the first paragraph on their website
The Patriot Act and Snowden’s leaks have shown companies will go against their privacy policy to appease governments. Search engines especially are targeted by five eyes with the PRISM program where copies of all your data, linked to your payment, are sent to Five Eyes and stored. Gag orders and legal threats prevent disclosure, as has been done with prior tech companies who have tried to push back against this.
Be wary of trusting corporations with your data as monetization is a powerful incentive.
Lemmy is different though right?… right?
With lemmy you are trusting whatever instance your account is on, and really any federated instances since they could choose to hold onto your posts and comments
I don’t know if I believe that. It’s a paid service, so the only way to enforce that unpaid users cannot search is to take a search request and check if it is coming from your account. Same with basic things like rate limiting requests. You literally need to associate your requests to an account to make basic functionality like this work.
If they do this but just don’t log it, then that means there is no way for their devs to ever debug issues users have or to monitor their services. I’m highly skeptical.
Also, “trust us” is something I’ve heard too many times.
It’s a paid service, so the only way to enforce that unpaid users cannot search is to take a search request and check if it is coming from your account
That’s not the same as logging.
You literally need to associate your requests to an account to make basic functionality like this work.
They just need to check the session of the user on the fly during the search operation. Once the search is done they don’t need to persist any record linking the search and the user.
They don’t need to tie the searches to an account. They log them anonymously. From their privacy policy:
Absent from our logs are any identifying information about your client. As such, any query or traffic logging that we do cannot be tied back to your account, ensuring that Kagi developers are the only people that the logs will ever be useful to.
Thanks, edited the comment!
and am I supposed to believe such a bold claim? the only reason they give is “trust me, bro. I pinky promise I’m not logging anything”.
You have one account, every search query you make is associated with that account. And even if they aren’t selling that ultra sensitive data, I’m sure they are keeping logs to prevent abuse and fix bugs which could be used when a third party gains access to their servers (malicious actors, law enforcement, etc).
And that’s assuming that Kagi is not mining and or selling any data themselves, which is a bold assumption given how little we know about their proprietary product. If at least they published the source code, but no. I’m supposed to trust a proprietary black box which could potentially be linking every search query back to me.
I don’t trust or ever plan on getting Kagi, but in their defense, the “trust me bro” is a large portion of privacy services. I use Mullvad VPN and think they have a great reputation that have proved themselves. I have no however, personally checked the servers to verify myself what’s running, so I am trusting then. Even when running open source software, I know none of us here have actually looked into every line of code of our browsers or our phones to see what’s all running. It’s simply unfeasible, so trust and reputation is still required at the end of the day.
That’s absolutely true. The problem is that, to make use of VPN services, it’s required to have an account or other identifier.
But that’s no true for search engines. If I wanted to, I could make completely anonymous searches using SearXNG or DDG from different IPs and they would not have any way to correlate the search queries.
That’s not true with Kagi and it’s a completely unnecessary privacy risk you’re taking when using it.
How is it privacy invasive? For example, compared to competition like Google?
Not op, nor i have any experience with kagi, but i suspect there is no way to do an Anonymous search with kagi.
copypasting the other comment I made in this thread:
and am I supposed to believe such a bold claim? the only reason they give is “trust me, bro. I pinky promise I’m not logging anything”.
You have one account, every search query you make is associated with that account. And even if they aren’t selling that ultra sensitive data, I’m sure they are keeping logs to prevent abuse and fix bugs which could be used when a third party gains access to their servers (malicious actors, law enforcement, etc).
And that’s assuming that Kagi is not mining and or selling any data themselves, which is a bold assumption given how little we know about their proprietary product. If at least they published the source code, but no. I’m supposed to trust a proprietary black box which could potentially be linking every search query back to me.
I don’t have any skin in this game. I just wanted to point out that you went from “given how privacy invasive this particular entity is”
To
“… assuming… how little we know… could potentially”
That’s a pretty big leap from a bold and confident assertion that an entity is doing something all the way to saying that entity maybe could be doing something but we don’t know. It’s just a weird logical leap to me, and I felt compelled to mention it.
I’m glad I’m finally not the only one feeling this way. I’ve been seeing them aggressively pushed seemingly out of the blue for months now. Especially for what is such an awful deal and zero evidence to their claims, just citing the marketing page at face value.
Why? It’s a great search engine that a lot of people find extremely useful.
I’ve mentioned it a few times because it’s actually good.
Technically the best one was Altavista.
But they are long gone because they came from the old academic & idealistic internet and they never learned to survive in that internet where money rules.
From a very old memory, Google blew AltaVista out of the water no?
I mean we all switched for a reason and it wasn’t the cute logo.
I believe yahoo killed altvista before google came and finished it off
Altavista was ahead of their time. The modern internet desperately needs a technical search engine.
Many people keep asserting that DuckDuckGo is as useless as Google, but I haven’t had a single issue with it. Could it be a regional thing?
DDG is ok for most searches, but they have definitely hit a plateau. Programming search results are quite poor, for instance.
I’ve started paying for kagi. Their results are just way better at this point.
I don’t live in the US and the regional results for my country are worse on DuckDuckGo than on Google.
But still, I noticed Google’s quality drastically falling down in recent years.
It is as useless. After all, it’s just Bing. But if the results are good enough for you, then why bother finding something else.
Some results come from Bing.
most are. just do a side by side comparison. for most queries it’s literally the same results.
Hmm, I did notice a sudden severe drop in quality recently. Perhaps they are A/B testing something.
ddg has been going downhill for awhile now. they changed something significant a couple years back that just made results, especially after the first half-page, absolute shit.
I think it was because they dropped Yandex results.
Ddg cannot filter out results. If you don’t want pages containing the term, term you add -term to your search and those results should not be included.
Ddg doesn’t do this. I did a brief test of many search engines, and only google and mojeek filtered results correctly.
Edit: yadex seems to have working filtering.
Google doesn’t do that properly anymore, either, much like the “literal search” (quotes) - used to work, now it’s a crap shoot.
It’s not as bad as Google yet, but I find myself getting terrible or no results quite a few times.
Ex: if I’m looking for a niche blog post from example.com, just entering the keywords doesn’t return the right result, if anything at all. I have to add “site:example.com” and the right link shows up on top.
It’s kinda amusing when this happens, but I keep using ddg anyway because bing and Google had the same issue for the same keywords when I ran into the issue.
I self-host a SearXNG instance and it works really well.
Correct answer
Why pay for a search engine? I’m getting tired of Kagi astroturfing
Thank you!
Ooo this looks super interesting
Yandex was way better for searches in russian sources, but it came to shit just like google and also excludes whatever russian government don’t like at that point. I searched for some software in it multiple timea and the first link was some noname, probably malware site. It also promotes it’s own malware like browser with questionable russian security sertificates and their own Alexa. I’d honestly not include it in any list.
I like DDG and don’t switch from it that much. I’ve also heard Kagi as paid search engine is good, but I’ve never tried it.
Here are some options I use in my rotation.
Brave Search (skip the browser)
Mojeek
Qwant (French)
Yandex (Russian)
Mullvad Leta (Mullvad VPN subscription required)
MetaGer (German meta search)
Startpage (Private Google results)
DuckDuckGo (Private Bing results)
SearXNG and similar self hosted options are awesome, but I’ve found them unreliable.
Be skeptical of Kagi… It’s promoted pretty heavily around here for something that’s not FOSS.
I’m convinced at this point that Kagi is astroturfing on Lemmy, it’s unnatural. All they do is just cite the marketing page when questioned about the quality, or as to why they supposedly think it’s worth paying for compared to much more mature as-driven services that do respect your privacy.
I think you meant to write ass-driven
I’ve been using Kagi for 3 months now I think? And I’ve found it to be extremely useful for research…to a point, my only real issue is that after the first or so series of results it either doesn’t offer anything further or just no longer is showing relevant results to my search. Honestly though I’ve been very happy with Kagi and continue to pay the 10/month but if I do find something even better I’d happily switch. I’m not married to any particular search engine (well …least not to the extent of using Google services).
If you’re skeptical just give it a try I think they have a trial system in place.
THANK YOU for mentioning startpage! I have not heard of them but I haven’t liked using DDG very much so I’ve begrudgingly using google in private mode (and then I always forget to switch back and forth)
Took wayyy to long to find something not recommending Kagi.
Kagi
I’ve found duckduckgo works fine for things that aren’t recent.
Controversially the bing gpt chat bot works alarming well.
The bing chat bot scares me a bit. I don’t want to use bing out of principle, but at this point they’re just flirting too damn hard.
DDG is just Bing, so …
Are you a bot?
Are you?
Haha I wish, it’s just your response seems like you didn’t read the message.
My point is that you were talking solely about Bing.
You don’t believe that DDG is powered by Bing?
I self host a searxng instance and I find the combination of bing, duckduckgo and qwant as the source engines to return decent results. You can use a public instance and choose those engines in settings.
I’m specifically looking to replace DDG, they have really dropped in quality lately and are very clearly going back on their word for not having you in a bubble with specific to you results.
I found public instances often have issues connecting to many search engines at once, will look into self hosting it. Thanks!
This site collects quite a few web search engines: https://www.searchenginemap.com/
For everyone who uses searxng, is it great for day to day browsing? Do I require to host my own instance or the setup is as easy as requiring to add “searxng” option on my browser app?
I’m interested to move away from google as it becomes shitty everyday and loses its effectiveness for advanced query (based on my own result compared during 2013 up to pre covid). Bing have weird result on my region so cannot use it, ddg only for occasional use.
Thanks!
pick a host from searx.space
Might be a good idea to bookmark this too because sometimes some hosts go down.
Appreciate it, will read more in documentations 👍🏽
Is “super SEO sites” a catch all term for those 99% filler websites that have a tomato soup recipe (in theory) but actually start out with, “Historical evidence seems to suggest that the tomato was first cultivated in the territory that would eventually become Guam back in 1464…”
I’ve wondered if we had a common reference term for those? I wish it didn’t have a positive connotation though…
Those plus sites that are lists of buzz words that have nothing to do with the actual site.
Even though there’s a small monthly cost, the results have been consistent for Kagi. But consistency meets only half of my needs for search: I also want to make decisions quickly from what I find within the contents. If I were to to go to a link, wait for it to load, scroll the content, etc. – does that listed forum post have the answer I am looking for? Does this news article cover the nuances I have been tracking and would like to read more of? Kagi offers an AI-based summarize feature that helps. And that’s been meeting the other half of my needs, as well.
EDIT, an opinion: Search services may well be eventually replaced by small, niche LLMs trained to perform summerization tasks, such as Consensus, which I have used for work research, and Perplexity.ai. The AI summarize feature of Kagi is why I see the service as more useful than straight indexes, even when self-hosted. Kagi is a stepping stone toward this for me, and why I recommend it.
The ability to summarize pages directly on the search results page is a big win for me. I also use the kagi summarizer several times a day to summarize youtube videos that I’m interested in but don’t want to sit through the whole video.
Lol kagi bots spotted!
Beep boop, my dude.
Please solve the partial differential equation to continue watching the video:
“I am not a robot” captcha is getting too hard…": https://www.youtube.com/watch?v=ru6fi4O4lp4
How WOULD you differentiate bots from happy users, anyway? They certainly look like organic accounts to me
A 7 months old random fediverse account is a bot? You know that the entirety of the fediverse is like 0.001 percent of the user base of the social media giants?
That would be money really badly spent on astro-turfing bots.
Ooooh that’s crazy. Why are people not joining fediverse ? I think we need some marketing people to band together and create easy guides on how to sign up and share it on Facebook, YouTube etc
Kagi is the highest quality for sure. There isn’t a better one, I have tried all of them except perplexity, and that’s more like chat gpt rather than search.
Come on over to Kagi! You do have to pay but I use a search engine dozens of times per day so I’m not too bothered by it.
I’ve heard a lot of great things about Kagi, though the search limit and subscription is a little off-putting. A self-hosted Kagi would be amazing though!
Kagi has a free trial, 100 searches for free, so you can try it out and see if you like it or not.
Why would they do all that work to create a better search engine than Google and give it away for free? What is going to pay their bills?
Google pays their bills by selling your data and spying on your searches, not to mention they are morally corrupt. Kagi pays their bills with your monthly sub. Seems a lot better.
I’m worried eventually even Kagi will get enshittified. It has been a common trend that almost always occurs. Open source is the only way to ensure stability. Conflict of interest is what leads to companies either overcharging, or even accepting to get bought out.
Don’t get me wrong, Kagi seems to be a great company thus far! But for something as important as search it would be best to have an open source solution.
Probably they will, but it could take a long time. Valve is still good as an example of a company that managed to not become crap.
Open source is the only way to provide stability? Can you explain more about this?
Valve is great! Private and non-profit organizations give me some hope for a better internet one day.
Regarding Kagi, there are other potential concerns with privacy, data leaks and price gouging as well. The Patriot Act and Snowden’s leaks have shown that companies will lie in their privacy policy to appeal to authorities even if they claim they are not storing information. All your health related searches, sensitive personal details, private life, etc is also always linked to your payment method waiting for a potential data leak if they are lying. (Or a copy is just sent to five eyes)
All that is to say, be wary of trusting your privacy to companies. Monetization is a powerful motivator!
You can pay with bitcoin, which might help keeping the account entirely anonymous for some.
After that power company got blackmailed in the American South and extorted for like 40000 Bitcoin, which they paid…but then within 24 hours the FBI had recovered all of it…which is something no one was supposed to be capable of doing, I’m going to go ahead and say Bitcoin anonymity is compromised.
That means the FBI or DHS has devoted the computational power necessary to track every.single.bitcoin.transaction and then indexed it all well enough to cross reference against metadata.
Bitcoin is as secure as they allow it to be. It’s most likely allowed for the same reason the Tor network is, because it’s an avenue to move resources or information to undercover assets overseas.
At this point, I’ve come to expect that all of the products I like are going to be ruined at some point, so it’s about establishing enough independence to more easily transition to the next service.
Kagi’s great, and I’ll worry about finding a better search engine once it gets worse, but I don’t expect that to happen before my next renewal, so I’m happy.
Self hosting is smart! Usually good things always come to an end, at least if they are not open sourced.
Plus self hosting an index big enough and fast enough for a useful search engine is ridiculous.
They now have unlimited searches for $10/mo. That’s what got me to try it out.
You are correct though. I really do not like having all of my search history tied to my credit card (and then me). What helps me justify that is that instead of me being the product like google, by paying I’ve become the customer. Hopefully that incentivizes keeping them on the up and up.
I did come across searnxg in this thread. It looks like that can be self hosted so I’m gonna give that a try as well.
That’s expensive as shit, what the hell.
Yeah, you may be fine with the $5 plan, but that’s the lowest tier available.
Afaik they are not really running a profit yet, just expanding, so that’s an eye opener to how expensive it is to run a search business and how much value Google and others estimate they get from your personal information.
For now though their user base seems fairly much leaning towards business users that can defend this expense as part of becoming more effective professionally. Hopefully over time they’ll grow large enough to provide cheaper plans for regular persons while staying privacy focused and ad free.
I did 1300 queries which I think would put it at .007 cents per query. Sounds reasonable to me.
Not when they collaborate with brave. Fuck that!
They don’t “collaborate” with Brave, just use its search API.
The owner of Kagi is a homophobic piece of shit.
Just use searxng.
If you are going to look for perfect companies, you will look forever, and probably use Google while you are looking.
It’s not worth it. Just use something better than Google even if it’s flawed, because everything had flaws.
Just use searxng.
Don’t give money to homophobes, we have an open source and better alternative already.
Kagi isn’t homophobic. Apparently the “controversy” is that it employs brave search as well; which is indeed made by a company with a homophobic ceo.
Though by that logic virtually everyone in the world is. Because at some point in time, in some place, someone built a piece of modern civilization upon knowledge or resources belonging to a homophobic person, I’m fairly sure.
No, that’s a seperate controversy. The Kagis owners views became talked about a lot more during that event though.
So many people here are quick to defend and normalise homophobia, yet refuses to just use an open source free alternative that’s without this baggage.
Are you sure none of the developers of the open source free alternative aren’t homophobic, transphobic, or in some other way bad people?
Can you link to some evidence?
I think it’s perfectly understandable that some people are homophobes, so it doesn’t worry me. You still have people being afraid of black people in this age even.
You can dislike that of course, but that’s the real world. You can walk up to any older person and they are likely to be a lot less open to gay marriage than a younger person also.
You interpret that as them being hostile. I think it’s more about being afraid of something that doesn’t feel natural to them.
I think it’s perfectly understandable that some people are homophobes
No it’s not. It’s not a justifiable position. It’s not unexpected sadly, but it is not understandable because it’s not a position based on anything other than ingrained hatred and fear
Yeah, get fucked mate :)
Where did you get that from?
It’s all over the Mastodon side of the Fediverse, shouldn’t be too hard to find with a quick search. I’m not at my pc atm so cbf using a phone to find some links sorry.
The controversy was about Kagi using search API from Brave, whose CEO is supposedly homophobic. Not Kagi’s CEO himself being homophobic.
No, it is also the Kagi founder as per the Mastodon topics.
Can you show some evidence of that?
The sources I find are solely from mastodon threads, and date from this month and the Brave thing. This is the best thing I’ve found