buherator I agree with you, but I think the challenge is that Kagi combines results from several sources, including Google, Bing, (...and others...), and its own index, rather than relying on a single backend. It therefore has to reconcile, deduplicate, filter, and re-rank those results before presenting a unified results page. Since search operators and advanced features can behave differently across these backends, supporting them consistently may limit some of the flexibility Kagi would have if it relied exclusively on its own index. In my opinion, though, its own index alone is not yet sufficient, as it is less mature than Google’s or Bing’s (and many others') and would require substantial time and resources to approach their coverage of websites and pages. Combining sources therefore remains a worthwhile trade-off. At least, that is my understanding based on what the staff and developers have shared here.
For my work, which involves finding a large number of resumes, I still cannot fully replace Google with Kagi. When I use search operators, Google returns many resumes that do not appear in Kagi’s results (well, Google actually outclasses Bing and many other big known search engines here, too). I am not sure why this happens, but I suspect it may be related to how Kagi filters or processes results. Since Kagi also draws results from Google, I would expect it to surface those pages even if they are not included in its own index.
