Arbitrary and quite possibly unique searches over substantially all knowledge ever produced by humanity seem a lot harder than sending an unchanging large file over a network.
And since Bing rasing prices affects their price so much, I'd argue that Kagi isn't investing enough into their own crawler/index, so that they can ultimately bring down the price of search.
Instead they're integrating yet another third-party (OpenAI), thereby rasing the price even more and tying their price to third-party API pricing even more.
Instead they're integrating yet another third-party (OpenAI), thereby rasing the price even more and tying their price to third-party API pricing even more.