From Kagi's FAQ: "Our searching includes anonymized requests to traditional search indexes like Google and Bing as well as vertical sources like Wikipedia and DeepL or other APIs. We also have our own non-commercial index (Teclis), news index (TinyGem), and an AI for instant answers."

So, apparently Kagi is scraping Google and Bing results via a browser extension. Is this legal ?

My interpretation of this is they pay for API access and sends anonymised queries to Google and/or Bing.

I'm still puzzled how they get so extremely much better results compared to using any of thise directly or using DDG which does the same only with Bing.

Also a bit worried because I think there was a time when DDG felt the same before they suddenly gave in and started serving answers for queries I didn't ask just like Google but I'm not really sure.

Anonymized request means that the client's IP is not passed to the server. Since both search engine have a paid search results API, it is enough to simply proxy the API request to achieve this claim.

Kagi is both paying Google for API access and using Google infrastructure (GCP), helping Google make revenue that is not ad-supported.

They are paying for results, that seems fair and fully disclosed in their FAQ. I'd be curious to see how it performs (waiting for an invite).

I use startpage.com that does the same. It's just a proxy for Google search.

Do you have example queries to share? I’m curious.

I have stored some result pages in Joplin and will try to post some comparisons soon, but lunchbreak is over so not until later today - at earliest ;-)

But basically the results are good and if I don't immediately get exactly what I want one pair of doublequotes around a key phrase sorts the rest.