I have worked the Kagi assistant almost completely out of my workflow, replacing it with Perplexity. I would like to stick with using just Kagi, since it is nice for traditional web search in addition to LLM assistants, and their lenses are really nice (more flexible than the "focus" in Perplexity), but it's going to be tough for them to beat Perplexity in terms of quality.
I didn't realize who I was replying to, my apologies. I honestly haven't had much of a chance to play around with the new Kagi interface yet, so I can't speak too much to that. Though Perplexity Pro can pull in dozens of sources collected through a series of searches with the intent of finding different parts of the question. It also has support for a lot of things like generating charts and tables, running code (quite new, haven't done that myself yet), etc.
I have just found the quality of both the UI/UX as well as the answers themselves to be really great with Perplexity, far surpassing was I experienced with the previous (not this newly released) "Expert" assistant in Kagi.
Thanks for clarifying. We heard anecdotally people cancel Perplexity to use the new Kagi Assistant (so opposite direction of what you said) and hence I was curious to find out more about where you are coming from.
