Google quietly announced that Programmable Search (ex-Custom Search) won’t allow new engines to “search the entire web” anymore. New engines are capped at searching up to 50 domains, and existing full-web engines have until Jan 1, 2027 to transition.
If you actually need whole-web search, Google now points you to an “interest form” for enterprise solutions (Vertex AI Search etc.), with no public pricing and no guarantee they’ll even reply.
This seems like it effectively ends the era of indie / niche search engines being able to build on Google’s index. Anything that looks like general web search is getting pushed behind enterprise gates.
I haven’t seen much discussion about this yet, but for anyone who built a small search product on Programmable Search, this feels like a pretty big shift.
Curious if others here are affected or already planning alternatives.
UPDATE: I logged into Programmable Search and the message is even more explicit: Full web search via the "Search the entire web" feature will be discontinued within the next year. Please update your search engine to specify specific sites to search. With this link: https://support.google.com/programmable-search/answer/123971...
I know that duckduckgo uses Microsoft Bing Custom search and honestly it is a much more robust system since you don't have to worry about Google axing it. https://www.customsearch.ai
Instead you worry about Microsoft axing it? Sure, it might take 3 years instead of 6 months, and the shutdown period would be 1 year instead of 1 month, but hardly either are long-term solutions.
What are some of the niche search engines build on Google's index affected by this?
Kagi
> Kagi This seems to be true, but more indirectly. From Kagi’s blog [0] which is a follow up to a Kagi blog post from last year [1].
[0]> Google: Google does not offer a public search API. The only available path is an ad-syndication bundle with no changes to result presentation - the model Startpage uses. Ad syndication is a non-starter for Kagi’s ad-free subscription model.[^1]
[0]> The current interim approach (current as of Jan 21, 2026)
[0]> Because direct licensing isn’t available to us on compatible terms, we - like many others - use third-party API providers for SERP-style results (SERP meaning search engine results page). These providers serve major enterprises (according to their websites) including Nvidia, Adobe, Samsung, Stanford, DeepMind, Uber, and the United Nations.
I’m an avid Kagi user, and it seems like Kagi and some other notable interested parties have _already_ been unable to do get what they want/need with Google’s index.
[0]> The fact that we - and companies like Stanford, Nvidia, Adobe, and the United Nations - have had to rely on third-party vendors is a symptom of the closed ecosystem, not a preference.
Hopefully someone here can clarify for me, or enumerate some of these “third-party vendors” who seem like they will/might/could be directly affected by this.
[0] antibabelic > relevant https://blog.kagi.com/waiting-dawn-search [1] https://blog.kagi.com/dawn-new-era-search > [^1]: A note on Google’s existing APIs: Google offers PSE, designed for adding search boxes to websites. It can return web results, but with reduced scope and terms tailored for that narrow use case. More recently, Google offers Grounding with Google Search through Vertex AI, intended for grounding LLM responses. Neither is general-purpose index access. Programmable Search Engine is not designed for building competitive search. Grounding with Google Search is priced at $35 per 1,000 requests - economically unviable for search at scale, and structured as an AI add-on rather than standalone index syndication. These are not the FRAND terms the market needs
I believe they try to indirectly say they are using SerpApi or a similar product that scrapes Google search results to use them. And other big ones use it too so it must be ok...
That must be the reason why they limit the searches you can do in the starter plan. Every SerpApi call costs money.
Google is also suing SerpAPI
And I can't prove correlation but they refused to index one of my domains and I think it _might_ be because we had some content on there about how to use SerpAPI
Kagi does not use Google's search index. From their post which made the front page of HN yesterday [1]:
> Google does not offer a public search API. The only available path is an ad-syndication bundle with no changes to result presentation - the model Startpage uses. Ad syndication is a non-starter for Kagi’s ad-free subscription model.
They then go on to say that they pay a 3rd party company to scrape Google results (and serve those scraped results to their users). So their search engine is indeed based on unauthorized and uncompensated use of Google's index.
But since they're not using/paying for a supported API but just taking what they want, they indeed are unlikely to be impacted by this API turndown.
Congrats on saying that in the most one-sided way possible. Google makes it literally impossible for them to pay for access to search results to make the product they want (customizable subscription search with no ads), and Google also is the de-facto globally sanctioned crawler because they are the only search engine anyone gives a shit about, and also sites need to be indexed by them to survive. In short, Google owns the river and sells the boats, and the public built a wall around it. Google is in a monopoly position in search.
>In short, Google owns the river and sells the boats, and the public built a wall around it.
That would be a monopoly if there was only 1 river in the whole world.
They published this the other day:
https://blog.kagi.com/waiting-dawn-search
Which saw some discussion on HN.
> some discussion
~450 score, ~247 comments and still on /best ("Most-upvoted stories of the last 48 hours"):
https://news.ycombinator.com/item?id=46708678 - "Waiting for dawn in search: Search index, Google rulings and impact on Kagi"
I think Kagi buys search engine results from SERP vendors who typically scrape Google’s results and offer an API experience on top of it.
No wonder Kagi is angry.
Google is a monopoly across several broad categories. They're also a taxation enterprise.
Google Search took over as the URL bar for 91% of all web users across all devices.
Since this intercepts trademarks and brand names, Google gets to tax all businesses unfairly.
Tell your legislators in the US and the EU that Google shouldn't be able to sell ads against registered trademarks (+/- some edit distance). They re-engineered the web to be a taxation system for all businesses across all categories.
Searching for Claude -> Ads in first place
Searching for ChatGPT -> Ads in first place
Searching for iPhone -> Ads in first place
This is inexcusable.
Only searches for "ChatGPT versus", "iPhone reviews", or "Nintendo game comparison" should allow ads. And one could argue that the "URL Bar" shouldn't auto suggest these either when a trademark is in the URL bar.
If Google won't play fair, we have to kill 50% of their search revenue for being egregiously evil.
If you own a trademark, Google shouldn't be able to sell ads against you.
--
Google's really bad. Ideally we'd get an antitrust breakup. They're worse than Ma Bell. I wouldn't even split Google into multiple companies by division - I'd force them to be multiple copies of the same exact entity that then have to compete with each other:
Bell Systems -> {BellSouth, Bell Atlantic, Southwestern Bell, ...}
Google -> {GoogleA, GoogleB, GoogleC, ...}
They'd each have cloud, search, browser, and YouTube. But new brand names for new parent companies. That would create all-out war and lead to incredible consumer wins.
Could probably argued that search access is an essential facility[1], though it doesn't appear antitrust law has anywhere near the same sort of enforcement it did in the past.
[1] https://en.wikipedia.org/wiki/Essential_facilities_doctrine
> If you own a trademark, Google shouldn't be able to sell ads against you.
This is frustrating even from a consumer perspective. Before I ran adblock everywhere, I couldn't stand that typing in a specific company I was looking for would just serve ads from any number of related brands that I wasn't looking for that were competitors.
what stops Kagi from indexing internet and makes them pay some guys to scrape search results from Google? one guy at Marginalia can do it and entire dev team at a PAID search engine can't?
I don't know about others, but we have special rules for Google, Bing, and a few others, rate-limiting them less than some random bot.
The problem is scrapers (mostly AI scrapers from what we can tell). They will pound a site into the ground and not care and they are becoming increasingly good at hiding their tracks. The only reasonable way to deal with them is to rate-limit every IP by default and then lifting some of those restrictions on known, well behaving bots. Now we will lift those restrictions if asked, and frequently look at statistics to lift the restrictions from search engines we might have missed, but it's an up hill battle if you're new and unknown.
As we've seen here on HN on the AI boom, it's not wonderful when a bunch of companies all use bots to scrape the entire web. Many sites only allow Google scrapers in robots.txt and the public will fight you hard if you scrape them without permission. It's just one of those things where it would be better for everyone if search engines could pay for access to the work that's done only once.
> Many sites only allow Google scrapers in robots.txt and the public will fight you hard if you scrape them without permission.
This just lets a monopoly replace the website instead of distributing power and fostering open source. The same monopoly that was already bleeding off the web's utility and taxing it.
> “search the entire web”
TIL they allowed that before. It sounds a bit crazy. Like Google is inviting people to repackage google search itself and sell it / serve with their own ads.
You know, back in the days, the web used to be more open. Also - just because you CAN do something, doesn't mean you HAVE to.
It basically means that Google is now transitioning into a private web.
Others have to replace Google. We need access to public information. States can not allow corporations to hold us here hostage.
I tried it and contributed to searx. It didn't give the same result as Google, and it also have 10k request rate limit (per month I believe). More than that you'll have to "contact us"
I built my own web search index on bare metal, index now up to 34m docs: https://greppr.org/
People rely too much on other people's infra and services, which can be decommissioned anytime. The Google Graveyard is real.
Number of docs isn’t the limiting factor.
I just searched for “stackoverflow” and the first result was this: https://www.perl.com/tags/stackoverflow/
The actual Stackoverflow site was ranked way down, below some weird twitter accounts.
I don't weight home pages in any way yet to bump them up, it's just raw search on keyword relevance.
Google's entire (initial) claim-to-fame was "PageRank", referring both to the ranking of pages and co-founder Larry Page, which strongly prioritised a relevance attribute over raw keyword findings (which then-popular alternatives such as Alta Vista, Yahoo, AskJeeves, Lycos, Infoseek, HotBot, etc., relied on, or the rather more notorious paid-rankings schemes in which SERP order was effectively sold). When it was first introduced, Google Web Search was absolutely worlds ahead of any competition. I remember this well having used them previously and adopted Google quite early (1998/99).
Even with PageRank result prioritisation is highly subject to gaming. Raw keyword search is far more so (keyword stuffing and other shenanigans), moreso as any given search engine begins to become popular and catch the attention of publishers.
Google now applies other additional ordering factors as well. And of course has come to dominate SERP results with paid, advertised, listings, which are all but impossible to discern from "organic" search results.
(I've not used Google Web Search as my primary tool for well over a decade, and probably only run a few searches per month. DDG is my primary, though I'll look at a few others including Kagi and Marginalia, though those rarely.)
<https://en.wikipedia.org/wiki/PageRank>
"The anatomy of a large-scale hypertextual Web search engine" (1998) <http://infolab.stanford.edu/pub/papers/google.pdf> (PDF)
Early (1990s) search engines: <https://en.wikipedia.org/wiki/Search_engine#1990s:_Birth_of_...>.
PageRank was an innovative idea in the early days of the Internet when trust was high, but yes it's absolutely gamed now and I would be surprised if Google still relies on it.
Fair play to them though, it enabled them to build a massive business.
Anchor text information is arguably a better source for relevance ranking in my experience.
I publish exports of the ones Marginalia is aware of[1] if you want to play with integrating them.
[1] https://downloads.marginalia.nu/exports/ grab 'atags-25-04-20.parquet'
Though I'd think that you'd want to weight unaffiliated sites' anchor text to a given URL much higher than an affiliated site.
"Affiliation" is a tricky term itself. Content farms were popular in the aughts (though they seem to have largely subsided), firms such as Claria and Gator. There are chumboxes (Outbrain, Taboola), and of course affiliate links (e.g., to Amazon or other shopping sites). SEO manipulation is its own whole universe.
(I'm sure you know far more about this than I do, I'm mostly talking at other readers, and maybe hoping to glean some more wisdom from you ;-)
Oh yeah, there's definitely room for improvement in that general direction. Indexing anchor texts is much better than page rank, but in isolation, it's not sufficient.
I've also seen some benefit fingerpinting the network traffic the websites make using a headless browser, to identify which ad networks they load. Very few spam sites have no ads, since there wouldn't be any economy in that.
e.g. https://marginalia-search.com/site/www.salon.com?view=traffi...
The full data set of DOM samples + recorded network traffic are in an enormous sqlite file (400GB+), and I haven't yet worked out any way of distributing the data yet. Though it's in the back of my mind as something I'd like to solve.
Oh, that is clever!
I'd also suspect that there are networks / links which are more likely signs of low-value content than others. Off the top of my head, crypto, MLM, known scam/fraud sites, and perhaps share links to certain social networks might be negative indicators.
You can actually identify clusters of websites based on the cosine similarity of their outbound links. Pretty useful for identifying content farms spanning multiple websites.
Have a lil' data explorer for this: https://explore2.marginalia.nu/
Quite a lot of dead links in the dataset, but it's still useful.
Very interesting, and it is very kind of you to share your data like that. Will review!
Google’s biggest search signal now is aggregate behavioral data reported from Chrome. That pervasive behavioral surveillance is the main reason Apple has never allowed a native Chrome app on iOS.
It’s also why it is so hard to compete with Google. You guys are talking about techniques for analyzing the corpus of the search index. Google does that and has a direct view into how millions of people interact with it.
> That pervasive behavioral surveillance is the main reason Apple has never allowed a native Chrome app on iOS
The Chrome iOS app still knows every url visited, duration, scroll depth, etc.
Yes indeed, they have an impossibly deep moat and deeper pockets. I'm certainly not trying to compete with them with my little side project, it's just for fun!
Sure, but the point is results are not relevant at all?
It’s cool though, and really fast
I'll work on that adjustment, it's fair feedback thanks!
Unfortunately this is the bulk of search engine work. Recursive scraping is easy in comparison, even with CAPTCHA bypassing. You either limit the index to only highly relevant sites (as Marginalia does) or you must work very hard to separate the spam from the ham. And spam in one search may be ham in another.
I limit it to highly relevant curated seed sites, and don't allow public submissions. I'd rather have a small high-quality index.
You are absolutely right, it is the hardest part!
What do you mean they're not relevant? The top result you linked contained the word stackoverflow didn't it? It's showing you exactly what you searched for. Why would you need a search engine at all if you already know the name of the thing? Just type stackoverflow.com into your address bar.
I feel like Google-style "search" has made people really dumb and unable to help themselves.
the query is just to highlight that relevance is a complex topic. few people would consider "perl blog posts from 2016 that have the stack overflow tag" as the most relevant result for that query.
Confluence search does this, for our intranet. As a result it's barely usable.
Indexing is a nice compact CS problem; not completely simple for huge datasets like the entire internet, but well-formed. Ranking is the thing that makes a search engine valuable. Especially when faced with people trying to game it with SEO.
Unfortunately the index is the easy part. Transforming user input into a series of tokens which get used to rank possible matches and return the top N, based on likely relevence, is the hard part and I'm afraid this doesn't appear to do an acceptable job with any of the queries I tested.
There's a reason Google became so popular as quickly as it did. It's even harder to compete in this space nowadays, as the volume of junk and SEO spam is many orders of magnitude worse as a percentage of the corpus than it was back then.
I am definitely not trying to complete with Google, instead I am offering an old-school "just search" engine with no tracking, personalization filtering, or AI.
It's driven by my own personal nostalgia for the early Internet, and to find interesting hidden corners of the Internet that are becoming increasingly hard to find on Google after you wade through all of the sponsored results and spam in the first few pages...
There may be a free CS course out there that teaches how to implement a simplified version of Google's PageRank. It's essentially just the recursive idea that a page is important if important pages link to it. The original paper for it is a good read, too. Curiously, it took me forever to find the unaltered version of the paper that includes Appendix A: Advertising and Mixed Motives, explaining how any search engine with an ad-based business model will inherently be biased against the needs of their users[0]
[0] https://www.site.uottawa.ca/~stan/csi5389/readings/google.pd...
The input on the results page doesn't work, you always need to return to the start page on which the browser history is disabled. That's just confusing behaviour.
I guess you used the return key instead of clicking on the search icon? Seems to be a bug with the return key, I'll fix that this weekend sorry.
True, didn't occur to me, that I should click on the icon instead. Once I have clicked on the search icon once, enter also works. When I input a short query (single letter) it sometimes just shows a blank page, but maybe that is just HNs hug of death. Consider putting the query term more prominently in the front of the URL, so users can edit it. Also from the startpage, the URL in the URLbar isn't updated. As I already wrote, the browser shows completion for the searchbar on the result page, but does not for the one one the startpage. For my taste I would prefer less JS trickery, which would maybe already get rid of some of these issues.
Appreciate the detailed feedback! A lot of the JS trickery and URL shenanigans I'm doing is to prevent bot spam attempts, which was a real problem in the beginning.
This is pretty cool. Don't let the naysayers stop you. Taking a stab at beating Google at their core product is bravery in my book. The best of luck to you!
Thank you kindly! It's just for fun.
> it’s just for fun.
amazing, for real.
everything i’ve read and heard about the good internet is that it was good because sooooo many of the people did stuff for exactly that, fun.
i’ve spent some time reading through some of the old email lists from earlier internet folks, they predicted exactly what weve turned this into. reading the resistance against early adoption of cookies is incredible to see how prescient some of those people were. truly incredible.
keep having fun with it, i think it’s our only way out of whatever this thing is we have now.
This is mad but cool. Keep at it.
You should consider filtering by input language. Showing the same Wikipedia article in different languages is not helpful when I am searching in English. Also you may unify by entries by URL, it shows the same URL, just with different publish dates, which is interesting and might be useful, but should maybe be behind a toggle, as it is confusing at first.
Great feedback, agree I need to filter here. Some website localization is very hard to work around, because they will try to geo-locate the IP address of your bot and redirect it accordingly to a given language...
The issue I was having was with the query "term+wikipedia" it then shows the wikipedia article in Czech, Hungarian, Russian, some kind of Arab and other before finally showing the English version. Then also a lot of that occur 2,3,4+ times with the same URL, just differing in crawltime by a few minutes.
I made also something for my own search needs. It's just an SQLite table of domains, and places. I have your search engine there also ;-)
https://github.com/rumca-js/Internet-Places-Database
Demo for most important ones https://rumca-js.github.io/search
Thank you, will check it out!
I tested it using a local keyword, as I normally do, and it took me to a Wikipedia page I didn’t know existed. So thanks for that.
It will throw up weird and interesting results sometimes ;-)
Lol, a GooglePlus URL was mentionned on a webpage i browsed this week.#blastFromThePast
I still remember their circles interface ;-)
It's been clear for the last decade that we have to wean ourselves off of centralized search indexes if only to innoculate the Net against censorship/politically motivated black holing.
I can only weep at this point, as the heroes that were the Silent and Greatest generations (in the U.S.), who fought hard to pass on as much institutional knowledge as possible through hardcore organization and distribution via public and University library, have had that legacy shit on by these ad obsessed cretins. The entirety of human published understanding; and we make it nigh impossible for all but the most determined to actually avail themselves of it.
Relevant: Waiting for dawn in search: Search index, Google rulings and impact on Kagi https://news.ycombinator.com/item?id=46708678
This might be me reading it wrong, but isn't shutting down the full-web search going against the ruling mentioned in the Kagi post?
> Google must provide Web Search Index data (URLs, crawl metadata, spam scores) at marginal cost.
Maybe they're shutting down the good integration and then Kagi, Ecosia and others can buy index data in an inconvenient way going forward?
If I understand Kagi's blog post correctly, then here's what happened, chronologically:
Kagi makes deals with many search engines so they can have raw search results in exchange for money.
Google says: no, you can't have raw search results because only whales can get those. Only thing we can offer you is search results riddled with ads and we won't allow you to reorder or filter them.
Kagi thinks Google's offer is unacceptable, so Kagi goes to a third party SERP API, which scrapes Google at scale and sells the raw search results to Kagi and others.
August 2024: Court says Google is breaking the law by selling raw search results only to whales.
December 2025: Court orders that for the next six years, 1. Google must no longer exclude non-whales from buying raw search results, 2. Google must offer the raw search results for a reasonable price, and 3. Google can no longer force partners to bundle the results with ads.
December 2025: Google sues the third-party scraping companies.
January 2026: Google says "hey, the old search offering is going to go away, there's going to be a new API by 2027, stay tuned."
I don't really see any mentioning of a new API, beyond their Vertex AI thing, and I don't know how comparable that might be. Also it is capped at 50 domains (by default).
It is perhaps a clever legal workaround. They must sell access to their index, but the verdict didn't state how much of it you can buy access to at any one time. So they put a limit of 50 domains, because that accommodates everyone who's not a search engine, but effectively blocks Kagi and Ecosia, while not exactly refusing to sell to them.
> I don't really see any mentioning of a new API, beyond their Vertex AI thing
I was referring to the following statement about full web search where they don’t mention a 50-domains limit:
> if your use case necessitates full web search, contact us to express your interest in and get more information about our full web search solution. Your transition to an alternative solution needs to be completed by January 1, 2027.
It's only a "clever" workaround in a captured legal system that isn't interested in anti-monopoly outcomes. Any competent legal system would slap that shit down. Just the thought that they could "hack this ruling with one weird trick" is infuriating.
The 'Google Graveyard is real' sentiment captures something important: every dependency on a large platform is a loan that can be called in. The 34-million-document indie index project someone mentioned is the right response - own your core infrastructure. Easier said than done for whole-web search, but the same principle applies everywhere.
Much easier said than done, especially if you are serving users on scale.
Since the issue here is self-hosting and "core infrastructure", that isn't a problem, but everyone has their own search index, isn't credible either.
Meanwhile in Europe: Qwant and Ecosia team up to build their own search index: https://blog.ecosia.org/eusp/
It's a noble effort, but they're so late to the game that it's hard to see them making a significant dent. I hope I'm wrong.
They were:
> aiming to serve 30% of French search queries [by end of 2025]
https://blog.ecosia.org/launching-our-european-search-index/
Better late than never.
> The French index is at an advanced stage of completion, we have started creating the German language index, and the English one should start shortly. All progress is quickly integrated into the Qwant STAAN API.
They can build whatever they want with lots of #hashtags and public money, but that doesn't mean they'll get 30% of French people to use it.
But of course they managed to cut themselves a nice salary with EU funds, paid in part by me and you, so that's all that matters.
The French government managed to rein in Amazon so traditional French stores, both online and brick and mortar ones, don't go bankrupt due to Amazon's unending pockets.
If they deem it necessary to rein in Google, they will rein in Google. There's no lack of tools for this, ranging from obliging phones sold in French territory to offer the French search engine as the default, to forcing every Google search result to promote the local search engine prominently, to campaigns about how it's important for national security not to rely on an adversary/enemy country's services, to everything in between and beyond.
Oh no, someone is making money outside of the big American monopolists. Quick, the vapors!
> with lots of #hashtags
I missed this one. What was it about?
Click the link.
Which honestly no user cares about. They only care about whether it is good enough that they can use it. Marketshare only matters if you fear the vendor might shut it down, or if you are running ads.
I feel like soon there won’t even be a point having a search engine since almost the entire internet will be useless AI slop.
It's as though full-text search of websites you've never heard of was a mistake :)
PageRank wouldn't exist without webrings, directories, and forums you could only search individually, and we thrived on that Internet.
Welcome back, ye olde Internet.
A search engine doesn't have to search the entire internet. Most of them are extremely opinionated about what they index.
The old internet is still there. It hasn‘t gone away; it‘s just undiscoverable with ad-based search. The more slop there is, the more necessary it is to have good search engines.
If you haven't tried Marginalia Search yet, do so. It's a small web search.
Recently, I set up a fresh system on a laptop. Ahahahaaa, how utterly crap Google search results now are! It fills me with some stress and disgust to use that. Now one of the first things I do, right after emergency using duckduckgo to search for uBlock Origin and NoScript, is to get Kagi search installed as default search. Then I can continue setting things up more calmly.
Seriously though. Five years ago Google already became unusable without "site:reddit.com" which is actually hilarious for a search engine that's supposed to search the entire internet. Nowadays reddit is also shit, which means that the only use case for me to use Google or any search engine is to find products that for some reason I don't want to buy on Amazon.
Internet isn't a global village, it's a global ghetto, and it's becoming increasingly true that the only way not to lose is not to play.
This will significantly impact (quite possibly kill) Startpage and Ecosia, who are effectively white-label Google, right?
What alternatives are there besides Bing? Is it really so hard that it’s not considered worth doing? Some of the AI companies (Perplexity, Anthropic) seem to have managed to get their own indexing up and running.
Excuse the self-promotion but Mojeek offers a web search API (>9 billion pages): https://www.mojeek.com/services/search/web-search-api/
I built many products on Google PSE (Custom Search). Results were nowhere near as good as regular Google, but still useful. I usually needed to use another library to get the DOM content anyway. But it still was solid for grounding/checking data.
RIP, another one to the Google Graveyard.
This and agressive anti-bot at YouTube is Alphabet closing the AI data leaking
Regarding alternate search engines: I consider the idea of YaCy kind of interesting: a P2P search engine: https://yacy.net/
Although, it needs some more work and peers to be usable as a general-purpose search engine.
Dumb question:
I keep seeing posts about how ~"the volume of AI scrapers is making hosting untenable."
There must a ton of new full-web datasets out there, right?
What are the major hurdles that prevent the owners of these datasets from providing them to third parties via API? Is it the quality of SERP, or staleness? Otherwise, this seems like a potentially lucrative pivot/side hustle?
Not directly covered by this blog, but for low cost and good performance the combination of gemini-3-flash with search grounding is hard to beat, at least for the many small experiments I use it for.
One thing touched upon in comments here: I never understood how it was proper for 3rd parties to scrape Google search results and reuse/resell them.
Really off topic, sorry, but I am surprised that more companies don’t build local search indices for just the few hundred web domains that are important to their businesses. I have tried this in combination with local (small and fast) LLMs and I think this is unappreciated tech: fast, cheap, and local.
Google has consistently ruined its search engine in the last (almost) 10 years. You can find numerous articles about this, as well as videos on youtube (which is also controlled by google).
Not long ago they ruined ublock origin (for chrome; ublock origin lite is nowhere near as good and effective, from my own experience here).
Now Google is also committing towards more evil and trying to ruin things for more - people, competitors, you name it. We can not allow Google to continue on its wiched path here. It'll just further erode the quality. There is a reason why "killed by google" is more than a mere meme - a graveyard of things killed by google.
We need alternatives, viable ones, for ALL Google services. Let's all work to make this world better - a place without Google.
To me there are two eras of the Google Graveyard(tm). First, there's the we're a university research group with an ad company footing the bill era. That's the early Google era, and it was a consequence of its corporate structure. They valued new projects, market fit, profitability, and maintenance be damned.
We're in the second era. The era of the MBAs are shutting down the last remnants of openness the company ever had.
Are competing search indexes (Bing, Ecosia/Qwant, etc) objectively worse in significant ways, or is Google just so entrenched that people don't want to "risk it" with another provider (and/or preferences and/or inertia).
I suppose I'm asking whether this is actually a _good thing_ in that it will stimulate competition in the space, or if it's just a case that Google's index is now too good for anyone to reasonably catch up at this point.
The beauty about Google Programmable Search across the entire web is that it's free and users can make money by linking it their Adsense account.
Bing charge per query for the average user. Ecosia and Qwant use Bing to power their results, probably under some type of license, which results in them paying much less per query than a normal user.
Bing recently shut down their API product, which was already very expensive.
If you want programmatic access to search results there aren't really many options left.
I can manage fine with other search indexes for English language searches; weather that is because others got better or google got worse i cannot tell, though I suspect the latter.
But for searching in more niche languages google is usually the only decent option and I have little hope that others will ever reach the scale where they could compete.
Bing's index is smaller than Google's, and anecdotally I get fewer relevant results when using it, particularly from sites like Reddit that have exclusive search deals with Google.
Yes, for non English queries they are all rubbish. And that's billions of users.
I'm curious about what it would take to build my own "toy" search engine with its own index. Anyone ever tried this?
Yeah that's where I started out in 2021. Been at it for almost 5 years now, last three of which full time. I'm indexing about 1.1 billion documents now off a single server.
Hard part is doing it at any sort of scale and producing useful results. It's easy to build something that indexes a few million documents. Pushing into billions is a bigger challenge, as you start needing a lot of increasingly intricate bespoke solutions.
Devlog here:
https://www.marginalia.nu/tags/search-engine/
And search engine itself:
https://marginalia-search.com/
(... though it operates a bit sub-optimally now as I'm using a ton of CPU cores to migrate the index to use postings lists compression, will take about 4-5 days I think).
Curious on what (how much) hardware your running this.
Currently running off
AMD EPYC 7543 x2 for 64 cores/128 threads
512 GB RAM
~ 90 TB of PM9A3 SSDs across 12 physical devices
Storage is not very full though. I'm probably using about a third of it at this point.
Might find YaCy interesting. It’s meant to be a decentralised search engine where users scrape the internet and can search other users indexes in a kind of torrent like way.
I found it didn’t really work as a real search engine but it was interesting.
Good luck scraping websites without being blocked, if you're not Google.
Well you'll get blocked some places but it's not too big of a deal. If you're running an above board operation, you can surprisingly often successfully just email the admin explaining what you're doing, and ask to be unblocked.
Sounds very time consuming. Glad you're able to sustain yourself to be able to do it full time.
I had misread the title as "Google is ending (full-web search) for [aka in favour of] (niche search engines)"
The correct parsing is: "Google is ending (full-web search for niche search engines)"
"Google will discontinue third-party niche search engine access to full-web search" would be far clearer.
Given that the title supplied is effectively editorialised, and the original article's title is effectively content-free ("Updates to our Web Search Products & Programmable Search Engine Capabilities"), my rewording would be at least as fair.
HN's policy is to try to use text from the article itself where the article title is clickbait, sensational, vague, etc., however. I suspect Google's blog authors are aware of this, and they've carefully avoided any readily-extracted clear statements, though I'll take a stab...
Here's the most direct 'graph from TFA:
Custom Search JSON API: Vertex AI Search is a favorable alternative for up to 50 domains. Alternatively, if your use case necessitates full web search, contact us to express your interest in and get more information about our full web search solution. Your transition to an alternative solution needs to be completed by January 1, 2027.
We can get a clearer, 80-character head that's somewhat faithful to that with:
"Google Search API alternative Vertex AI Search limited to 50 domains" (70 chars).
That's still pretty loosely adherent, though it (mostly) uses words from the original article. I'm suggesting it to mods via email at hn@ycominator.com; others may wish to suggest their own formulations.
Are search engines like Kagi completely screwed by this or is there a way for them to keep operating?
Kagi doesn't have a partnership with Google - they work under adversarial interoperability, stealing results from Google against their will, and paying some third-party to enable this. They'd like to simply pay Google, but Google doesn't want their money.
Kagi is backed by russia so they will be fine.
You’re probably referring to the fact that their search results include entries from Yandex. That’s something entirely different from being “backed by Russia.” If anything, they pay Russia, not the other way around.
What do you mean "backed by"
Is this about the little Google Search Bar that is present on some websites? Or am I mistaking something
Kind of, however the Google Search Bar present on website is usually there to search across their domain, the search results are limited to their domain e.g example.com/page1, example.com/page2. Google will carry on supporting this.
What they are ending is their support for websites to search across the entire web. The websites that search across the entire web are usually niche search engine websites.
Ahh; so that's the difference. Thanks!
What examples are there of people using this?
There is literally thousands of independent search engines that use Programmable search to search the entire web. Many ISP providers use it on their homepage, kids-based search engines like wackysafe.com use it, also search engines that focus on privacy like gprivate.com etc
Also LLM tools. Programmable Search Engine API was a way to give third-party LLM frontends the ability to give LLMs a web search tool. Notably, this was a common practice long before any of the major LLM providers added search capabilities to their frontents.
Exactly, Google want every one depended on Gemini.
Never build a product with core feature depending on a third-party, you will eventually get fucked up for sure. always have a 70:30 rule for revenue where 70% is core independent features.
Soon you'll find that you cannot exist on the web without relying on third parties. Sometimes you'll even have trouble getting paid thanks to the painful existence of payment processors.
True, you can’t exist without 3rd parties. But you shouldn’t let them be your core moat/USP. Jasper is a great example, they depended too much on LLM access, then ChatGPT launched and ate the value. Using third party APIs is fine, but building a product whose core depends on them is suicide.
That's why I eschew HTTPS.
Antitrust do not work against large companies.
Just dissolve them in acid.
This is the type of monopoly abuse these laws were designed to target, and antitrust laws actually do work against large companies.
If you actually enforce them.
Unfortunately, during the Reagan administration, political sentiment toward monopolies shifted and since then antitrust law has been a paper tiger at best.
I heard when Bush came to power, the antitrust complaint against Microsoft monopoly driven by the government was dropped.
Is this perhaps to prevent ChatGPT, Claude and Grok to use Google Search? It would make sense for Google to keep that ability for Gemini.
I suspect its going to hurt the indie developers and small start-ups who do not have special licensing agreements.
They'll go adversarial interop through SerpAPI, just like Kagi does. SerpAPI will get the money instead of Google getting it.
"Why we’re taking legal action against SerpApi’s unlawful scraping" https://blog.google/innovation-and-ai/technology/safety-secu...
What's their angle here? Courts have been over the "Is scraping websites that don't want to be scraped OK?" question plenty of times. From
> SerpApi deceptively takes content that Google licenses from others (like images that appear in Knowledge Panels, real-time data in Search features and much more), and then resells it for a fee. In doing so, it willfully disregards the rights and directives of websites and providers whose content appears in Search.
it sounds like they are somehow suing on behalf of whoever they are licensing content from, but does that even give Google standing?
I guess I'm asking if they actually are hoping to win or just going for a "the process is the punishment"+"we have more money and lawyers than you" approach.