20 Nov, 2025
TL;DR
Today we’re releasing two research assistants: Quick Assistant and Research Assistant (previously named Ki during beta).
Kagi’s Research Assistant happened to top a popular benchmark (SimpleQA) when we ran it in August 2025. This was a happy accident. We’re building our research assistants to be useful products, not maximize benchmark scores.
What’s Kagi Quick/Research?
Kagi Quick Assistant and Research Assistant (documentation here) are Kagi’s flagship research assistants. We’re building our research assistants with our philosophy on using AI in our products in mind: *Humans should be at the center of the experience,* and AI should enhance, not replace the search experience. We know that LLMs are prone to bullshitting, but they’re incredibly useful tools when built into a product with their failings in mind.
Our assistants use different base models for specific tasks. We continuously benchmark top-performing models and select the best one for each job, so you don’t have to.
Their main strength is research: identifying what to search for, executing multiple simultaneous searches (in different languages, if needed), and synthesising the findings into high-quality answers.
The Quick Assistant (available on all plans) optimises for speed, providing direct and concise answers. The Research Assistant focuses on depth and diversity, conducting exhaustive analysis for thorough results.
We’re working on tools like research assistants because we find them useful. We hope you find them useful too. We’re not planning to force AI onto our users or products. We try to build tools because we think they’ll empower the humans that use them.
... continue reading