Hacker News Reader

Comment by _pjee | Hacker News Reader

Been hearing a lot about Kagi on here, might reach out. My concern is that they're trying to beat Google at its own game by being better at scraping, document understanding and question-answering.

View on HN

From what I heard from their CEO (Vlad). No they don't try to do that and they don't even invest heavily in web scraping.

I don't think so. I was surprised to learn on their Discord that they rely heavily on paid results from Google's API and other search APIs. Not sure what they do with it but it's pretty amazing how they improve the results.

Don't think Google has a paid Search API. Bing does though with Azure.

From Kagi's FAQ (https://kagi.com/faq):

  > Where are your results coming from?
  >
  > *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.

Emphasis mine.

Right, that's scraping results, not from an API that doesn't exist. What exists is Custom Search from Google and that's an API to search results on your own website.

I was looking into that, looks like they have an API but it's limited to 10k requests per day, and TOS says you can't blend the results with other sources.

Reply on news.ycombinator.com