4 ms·

Last I saw Kagi was highly dependent on Google for their search results and seemed much more interested in LLMs and other side features than in replacing that c

• voltaireodactyl 2y ago
  FWIW, my understanding is that there just isn’t a real drop in replacement for the Google Index. Kagis focus on LLM seems targeted to me, and I wonder if they’re trying to figure out how to put it to work indexing for themselves, more than providing the search results to the user.
• berdon 2y ago
  I believe Kagi uses most major search indexes as well as its own. Additionally, I don’t think it’s fair to say it’s more interested in LLMs than focusing on search. I think it’s fair to say they’re interested in ensuring they’re offering a better, non ad-based search replacement. Disclaimer: Not affiliated with Kagi in any way, just a long time happy user.
• nine_k 2y ago
  In 1998, the Web was incomparably smaller. Google could put their whole infra into a dozen boxes, and then grow along with the Web, gradually. Modern competitors don't have this advantage. By now, crawling and indexing is a herculean task, and also quite expensive, due to the sheer size. There is Common Crawl [1]; at 400 TiB it is huge, but at 60 days refresh interval it's far from being very comprehensive or very fresh. Good for research, but likely not good for a commercial search engine. [1]: https://commoncrawl.org/ https://commoncrawl.org/