this post was submitted on 09 Aug 2026
44 points (94.0% liked)
Linux
14633 readers
492 users here now
A community for everything relating to the GNU/Linux operating system (except the memes!)
Also, check out:
Original icon base courtesy of lewing@isc.tamu.edu and The GIMP
founded 3 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
~~Check my post history~~, ~~I've~~ someone else has done research to discover Kagi just uses ~~Bing~~ Yandex as its indexer:
There is no way a fucking startup can catch up to Google that has been indexing the internet for the past 30 years. Microsoft had to purchase the indexer from Yahoo for billions of dollars just to catch up.
Kagi has always been upfront about using other search providers' indexes.
Per Kagi's current documentation (source), their search results
That particular page dates back to at least Oct. 2022. It was also mentioned in their FAQ back when they were still in closed beta
SearXNG does the same exact thing for free. What are the idiots paying for Kagi getting with their subscriptions?
It looks like you get what you pay for with SearXNG.
While SearXNG claims to aggregate results from "more than 70 search services", the "Engines" tab show that most instances only support a small number of engines, and many of these are degraded. That matches my experience of trying SearXNG, where I entered a few queries on a service that randomly picks an instance: Half of those queries failed due to the instance having been rate-limited by the search providers and the other half gave inconsistent and often poor responses. I'm not sure what engine(s) those instances used.
I guess I could self-host, and configure and maintain just the engines I want access to, but the time required to do so would very quickly exceed the amount of hours I have to work to pay for a Kagi subscription, so it doesn't make much sense economically.
However, one area where I think SearXNG does have a distinct advantage is privacy, provided that you self-host and only use engines that do not require an API key to access. But that's a tall ask for most people
Ecosia and Qwant have invested lots of money and they still only slowly on board the use of their own engine (staan).
A small player has no chance to be anywhere nearly as good.
I thought I remember a blog post by them explaining how it's impossible to catch up with Google results due to the control they have and money involved, so they have their own indexing service that also pulls from others that they pay for access to (other than Google).
I may be wrong but that's how I've understood Kagi's results as I've used it for the last 2 years...
Edit: Found the article
https://blog.kagi.com/waiting-dawn-search
Yeah I feel like they have always been open about their results pulling from a range of other search engines.
That blog post is promising, and a little hilarious. "A search index is human knowledge, it belongs to everyone, that's why ours is proprietary".