Getting unhappy about DuckDuckGo

I haven't noticed any issues with the quality of DDG results, but if you feel the results are lackluster, you could try a metasearch engine like SearXNG. You can self-host it or use one of the many public instances maintained by the community. The main advantage, apart from the privacy focused aspect of the project, is that you can pull results from multiple search engines with a single query. It's highly customizable too. You can configure it exactly how you want.

I like SearXNG, you get all of all worlds
- I've only used searx[ng] for several years. searx.space is pretty recommended to look for searxng working instances, as well as the ones that you might prefer depending of the country of the instance and so for. Public searx but no searxng working instances are really uncommon now a days.
  
  Every now and then your preferred instance becomes useless (whether google finds its way to block it, or to apply an aggressive rate limiter, or the instance gets unmaintained), so one needs to look for another one.
  
  DDG doesn't give bad results, but when I realized the majority of its results come from bing, meaning it's mostly a metasearch as well with a few entries of its own (that might have varied from that time), I then started to only use searx, and then when searx working instances were really hard to find I moved to searxng, and I'm happy with those instances. Again, at times I need to move to a different instance, though I've been using the last one I chose for more than a year now...
Trying it these past days and I'm impressed!
Public instances need to fight against the giants, but running your own local version is easy if you learn to use Docker. It just takes around a hundred of megabytes of memory. I have been super happy with it.
And for anyone who doesn't know: SearX is pronounced "Search", and it's successor, SearXNG, is pronounced "Searching".

(In many languages "x" signals a "ch" sound)

Note: But maybe don't go around saying "Have you Searching'd it yet?"
Tried using public instances, but 1/3 searches was failing. Found it unusable. That was like a year ago when I was looking for DDG alternatives.
- I've had good luck with searxng.site, but yeah, public instances aren't going to have the uptime and reliability of Google or DDG. Think of public instances as a test drive. You get a taste of what it can do but it's a much better ride when you self-host.
- This is unfortunately the same type of experience I've had with SearXNG. 🙁
This shit right here!

I use Qwant instead and I'm pretty happy wiith that.

At some point, DuckDuckGo stop handling boolean logic properly in search terms. I've been using it for more than a decade, and the quality has definitely gotten worse over the last few years

I like qwant.com from France. They are using the Bing index but started to collaborate with Ecosia to build their own.

I'll try qwant!
+1, Qwant is awesome
Do they still pop up a modal dialog asking you to disable your ad-blocker? That was what killed Qwant for me.
- I have never seen that. So not sure if qwant got better... or the adblockers.
- I only saw that briefly. I can't remember if it went away on it's own, or if I blocked it myself.
I’ve been using this for about a month now and have been happy with it too

There's only a handful of companies out there actually spidering. A lot of third party offerings are just re-scraping the existing spiders. I wouldn't be surprised if deficiencies in quality were cat and mouse games between google/bing/et all and DuckDuckGo.

I've been self-hosting SearXNG. It's fantastic for everything except local hits, business hours, stuff where Google maps data is being referenced.

I think the problem with free search is that somebody needs to pay for it. There's more people block both ads and anonymize themselves, the more free options will eventually wither.

And while I'm perfectly willing to pay for ad-free anonymity, capitalism dictates that all services need to have exponential growth or fail, and eventually all that data can just be sold or otherwise make it into the wrong hands.

I'm kind of hoping that at some point you can purchase distilled search content in a locally hostable AI model. It could post ad free and complete anonymous access, and you just need to pay for updates to the search model.

I've been using SearXNG during the last day and I'm quite impressed too so far!

True what you say about the problems behind net search. It's actually a very complex problem. In my opinion part of the problem is that there's a lot (most?) of rubbish out there. It's like a library with useful books of different genres all mixed together, and mixed with an even larger amount of nonsense books. Maybe a solution would be something completely different from indexing – but I have no idea what.
- It's an old problem. From the very start of the net, you had to sort the wheat from the chaff. Back then, the BS was human-generated. Now we have the addition of AI crap. But anyway, they solved it already. Its called wikipedia. (Or any other community curated data source as well.) I'm not some wiki fan, but that's the world's answer to encroaching bad data. An army of real, very corruptible, infighting, weird-as-hell wiki editors is our last stand against the BS.
Right on. I'm running searxng and whoogle. Whoogle is a low resource option, and it only sources Google. I like searxng for the deep results, all kinds of weird stuff pops.

I was recently recommended to check out YaCy. Haven't done it yet.

I switched to Kagi and am beyond satisfied. If your goal is to strictly degoogle, it fits the bill, but it still does if you are looking for better privacy, as it now comes with an implementation of Privacy Pass. The algorithm is leagues above Google's and DDGs, IMO, and the "lens" feature allows you to seamlessly filter the results to specialized sources, including the Fediverse. "Small web" is a fun feature for when you're bored running unit tests at work, too

I tried Kagi for a while, but it was giving me less useful results than DDG, so I simply left it. I think it depends a lot on what kinds of searches one does, and Kagi is more useful for other users.
- AFAIK the algorithm for Kagi is really alien compared to Google and Bing/DDG, so the results do look a little weird at first, the main difference being just the sheer reduction in quantity of results.
  But I guess if you didn't like it, you didn't. Maybe it is worse and I'm biased because I already paid
Forgot to mention that it unfortunately is a US-owned company, so it would be off the table for the full-on US boycott crowd, especially because it's a paid service.
Though they seem to be a genuinely good company that consistenly provides good customer support and improves the product in tangible ways. Privacy Pass was implemented because of customer feedback, for example, and so were crypto payments, and both were publicly discussed on the forums with good transparency. They also actively promote the decentralization of the internet: with that Small Web feature I mentioned, with Fediverse and Usenet Archives search being implemented by default, by providing an interface to use any LLM model through their assistant... So I wouldn't want to boycott them, and I don't

You can find a good Searxng instance or run your own (it can also run on a RPi)

This.

I'm using perennialte.ch, they redirect reddit urls to redlib which is a nice touch.

I went ddg > kagi > searxng and this is the set up I'm happiest with.
- Nice initiative besides the search service! Thanks for sharing.

For you, I would recommend the following:

I use SearXNG, which I self-host.

Didn't know about several of these, cheers!

https://www.qwant.com/

Have you also experienced this worsening of DuckDuckGo?

Yes.

Which other more privacy-respecting alternatives do you recommend?

I'm in the same boat. I'll be trying out these answers.

Same^2

I've definitely felt the enshittification of DDG. A couple of years ago they would start dropping hits related to my location into my search results, even when I had region off and private search by default. That gave me the impression that my IP address was being used and possibly passed on to Bing, but I don't have the chops to confirm it 🤷

Maybe I search for weird things, but my major gripe with DDG is that its autocorrect is way too aggressive. But SearXNG public instances work for me 99% of the time.

Kagi is pretty amazing. You have to pay but the peace of mind is worth it for the respect of your privacy. FastGPT is a phenomenally helpful tool that I use multiple times per day. Kagi.com

Thank you for the suggestion. I tried Kagi a couple of times, but it missed the useful results that DDG or Google were giving, so I dropped it.

It depends of course on what kinds of searches one typically needs. Probably there isn't a universally best search engine.

Been happy with Kagi for the past few months. So far no thoughts of switching back to either Google or DDG.

I have thought the same thing after about three years or so of using DDG, I've been using Qwant as of late and seem to get much better results

Same, I was a DDG user for years and switched to Qwant a month ago. Qwant results are a step up from DDG, and Qwant takes the same approach to privacy as DDG but it's based in France so it wins in that regard as well. I'm in the US and Qwant still does a great job of providing localized results.
I'll give Qwant too a try, cheers! Testing SearXNG for a couple of days first.
- I wish you luck with SearXNG, I don't know if I used the wrong instances or what but I didn't find the results i got that great. However that is the reason people say you should host you own instance. Hopefully it works out I really enjoy the idea of hosting my own instance

I haven't used DDG in years. I've been using Start Page and its fit all my needs (its basically old google before enshitification).

I've played with searxng which seems promising but I haven't given it enough time. But it seems like I might eventually move there.

Tried Startpage a while ago, but was put off but some alleged iffy dealings of the company behind it. Trying SearXNG now and I'm impressed!
- What iffy dealings are you referring to? Because they were acquired? They've been established, and restablished after being acquired, as being completely privacy focused.
  
  My only criticism is that they don't always play nice with VPNs.
  
  But yes, SearXNG is a great way to go too.

Searx is good enough if you set up plenty of engines - I do look up quite a lot of stuff and not once in the past 3 months did I go "yeah I need to use google for this".

I'm trying SearX today, after so many recommended it. It looks promising! Thank you for pointing out the multiple-engines setup.

One possible drawback: it seems I can't do "verbatim" searches; or at least, quotation marks don't seem to lead to verbatim searches – I'll try with "+". DDG was adamant with quotation marks, that's something I liked a lot about it.
- Interesting, verbatim searches work perfectly for me. Maybe it's some search engine that doesn't support them? I personally have bing/google/duckduckgo selected.

I mostly use startpage, but occasionally ecosia and mojeek.

I use StartPage because it’s the closest to Google 10+ years ago.

True about the good old-Google feeling! I want to find how much about of the shadiness claims about Startpage have been substantiated or denied.

I run my own instance of SearX. Very reliable.

Just setup my own instance! Happy so far!

you can search through brave or google using mullvads leta:

https://leta.mullvad.net/

I use it all the time, just wish they had an image search tab. That's the only thing I switch search engines for now.

You can try hosting a searxng instance. If not try some public searxng instances. https://searx.space/

Duckduckgo gives you Bing results. If you like Microsoft they are up the alley. If not tough luck.

DDG is often but not always a lot worse than Google in my experience.

I didn't know about this – that may explain the problem.

Been using Qwant for about a month. Feels like of like old Google did, very happy.

Only place it’s lacking is image search, but Google has been pretty shite for the past year too.

Same, but I've found Quants image search to be good enough.
- It’s passable so I don’t really mind it. It uses Bing’s image search if I remember right.

I think in the last few years DDG has been improving and google has been worsening for general searching. Because I have nearly stopped using !g before I used it constantly.

I still use google at work as the results there match a bit better.

I’ve tried other search engines any times. Despite the worsening google search somehow is still much better than alternatives.

i have noticed that i get fewer results from ddg lately, and local (server-side geoip driven) and totally irrelevant shit frequently thrown in starting below the fold on page 1.

but ddg has been my go-to for years. and very rarely do i need to look elsewhere for a different 'perspective'--picking from the others configured in my browser: ecosia, startpage, qwant, mojeek. i have a couple instances of self-hostable meta engines in there too, but those are too unreliable to count on for regular use.

I use DDG and can't say I've had an uptick in the amount of !g I have to do. The only one recently was for an image search but that's pretty normal when looking for something obscure.

Don't have any solutions, but figured another input might be interesting.

It's very likely to depend on the kinds of searches I do, indeed. Although I think it's the same as in the previous years. Could also be just a subjective impression, so I'll try to keep count of how often the "!g" really leads to better results.

I haven't really experienced a worsening of DDG, and this is a bit off topic, perhaps, but—

I have yet to find a better alternative to Google's video search. Google Books also remains valuable in many ways, since it will give you different "search inside the book"-type search results than the Internet Archive will (and they also have some books that IA does not).

What's annoying to me is that StartPage, which is supposed to have Google search results, and by and large does, does not give the same video results as Google (go ahead, try it). It would also be nice to have an Invidious or FreeTube type front end for Google Books, and I believe there used to be something like that, but not any longer.

Some Google products still have definite value, it's just important that they derive no benefit from us using them.

I remember reading that there was something "fishy" with Startpage, which is why I abandoned it quickly. But don't remember where.

Luckily I don't often search for videos, but I agree that it seems difficult to find good search alternatives there.
- For what it's worth, Mullvad recently opened up Leta to people without Mullvad VPN accounts. It's a google/brave search proxy so pretty much what Startpage claims to do, but I trust Mullvad to actually respect my privacy more than I do Startpage.
- This?

I've noticed this is the case but not that it's declining. I haven't used it that long.

Admittedly it could be a temporary fluctuation of some kind.

4get.ca

How is it for porn?
Really funny about page 😂 Cool that they share the git code!

The vast majority of the quality comes from the result ranking (that is what people generally refer as indexes albeit indexes are slightly different conceptually). In practice there are two main indexes/ranking (a) Google (b) Bing. DDG, Ecosia, yahoo are, at best, tuning the ranking (not sure if they make things better or worse with that).

This is the reason if you want meaningfully better quality you need searXNG. Because combining results from multiple search engines you grt both main indexes.

There is a ranking/quality problem there as well because how do you join the results of multiple search engines? I think it simply uses a “voting” system where it bumps the results from most search engines. This also means that you want a “balanced” number of search engines based on Google and Bing otherwise you’ll get mainly Google or mainly bing results. Or you may just enable only Google and Bing, really 🤷‍♂️

If there will be new indexes (e.g. the new European one) you may want to enable them as well but this kind of problems take time to become competitive with giants like Google and Bing (albeit with enshittification this gap is being reduced)

Thank you for the explanation about rankings, and the SearXNG suggestion. I'm trying it now and it looks promising!
(Oh and of course if you don’t want to self host and you don’t want your queries to be proxies by random Joe with the public instances, go for the engine you want to “donate” your attention to. I would go for Ecosia just because they seem the most ethical of all, including DDG…. But your choice)

Btw https://mwmbl.org/ is an interesting project, although not usable yet

Crawling is not really the elephant in the room, really. The issue is ranking and I don’t see how they plan to address that for having good quality

Startpage uses Google's results, similar to like DDG uses Bing's, I find to be splendid replacement for spying engine.

Try www.nogoo.me