Skip to content
On this page

Kagi AI

Kagi has long heritage in AI, in fact we started as kagi.ai in 2018 and we've previously published products, research and even a sci-fi story about AI. While generative AI opens a new paradigm of search and a vast search space of queries that never previously existed we have taken special care to ensure a thoughtful user experience guided by this philosophy of AI integration.

Modern search engines can indeed satisfy the majority of existing searches. However, it is important to note that these searches only represent a small fraction of the total possible searches that can be made (probably less than 10%). The emergence of generative AI will enable a new paradigm in search that can unlock a whole new category of previously impossible searches.

To ensure that our solution is in line with our mission to humanize the web we consider that:

  • AI should be used in closed, defined, context relevant to search (not to be a therapist for example)
  • AI should be used to enhance search experience, not to create it (similar to how we use JavaScript, meaning Kagi works perfectly fine without it)
  • AI should be used to the extent that does not diminish our humanity (AI is a tool, not a replacement for brain)

With our advancements in AI and search you are able to expand the possibilities of what you can search, asking harder and more advanced questions such as:

  • Do more people live in Rome or Budapest?
  • Give me the names of all Palo Alto based CEOs that run companies in the search space with less than 50 employee?
  • Summarize the latest research on lung cancer for me
  • Who is known as the father of Texas?
  • During world cup 2022, Argentina lost to France by how many points?
  • What is the name of Joe Biden’s wife’s mother?
  • Which of these compute the same thing: Fourier Transform on real functions, Fast Fourier Transform, Quantum Fourier Transform, Discrete Fourier Transform?

Kagi is thrilled to introduce next-generation AI into our product offering:

  1. Assistant
  2. Quick answer
  3. Summarize page
  4. Ask Question about Document

For a longer discussion about our AI journey please read our blog "Kagi's approach to AI in search" by our founder Vlad.