We're hiring!

Looking for a Technical Customer Success Associate. You'll help our community via email and Discord, troubleshoot issues, create clear documentations, and funnel feedback straight to the dev team.

Fully remote FT position: https://kagi.peopleforce.io/careers/v/134221-technical-customer-success-associate

Yes, Kagi's Orion browser delivers a modern WebKit browser as far back as Mojave and we do plan to enable High Sierra support as well. Just because it's hard doesn't mean it shouldn't be done

We get this question a lot: "Why pay for search when I can get it elsewhere for free?"

Spoiler: "free" isn't really free. But what's your take?

With Kagi, you can easily view a webpage's history through the Wayback Machine by opening any result in Web Archive via our quick actions menu:

Jonathan Bailey writes about switching to Kagi and the importance of being the customer rather than the product, especially in the context of search:

https://plagiarismtoday.com/2025/07/02/google-from-symbiote-to-parasite/

"I got phished from a sponsored Google search result."

https://www.reddit.com/r/Scams/comments/1m9z7ca/be_wary_of_google_i_got_phished_from_a_sponsored/

These scams add up over time. A Kagi subscription can save you money by eliminating the hidden costs of "free" search engines - like falling for fake ads, phishing attempts, or the time spent dealing with compromised accounts.

How to make Kagi your default search engine across various browsers:

https://help.kagi.com/kagi/getting-started/setting-default.html

"Behind the scenes, a machine starts building you. Piece by piece. Search by search. Until you’re not just a customer. You’re a product."

"You won’t stop it all. But you can make it harder. Ditch the apps that track by default. Use browsers that block fingerprinting and ads."

This is why we built Kagi. To serve you, not sell you.

Quotes from this post: https://untraceabledigitaldissident.com/the-data-collection-trap/

Search without the noise.

(Illustration by @chazhutton)

Lenses is one of the most beloved features of Kagi. It allows you to customize your searches by specifying which websites and other parameters you see in your results.

There are several default lenses to pick from, and you can create and fully customize your own!

With the Kagi for Libraries program, we'll offer free access to Kagi for public library patrons worldwide

If your library is interested or you know a local public library that could benefit, encourage them to apply and help us expand this program: https://kagi.com/libraries

Kagi offers various widgets intended to help you customize your search results. Each widget can be individually toggled, allowing you to tailor your results to suit your needs and preferences.

As my first vacation project I started to write my own "meta search engine".

The project is inspired by #Kagi and The custom search engine that "matdoesdev" created: https://matdoes.dev/metasearch

It is still very bare bones but:
- Custom Bangs
- Searching on Duckduckgo
- Blocking domains
Are already working.

There is no public instance but the source is open as always.
Be warned however I took the opportunity to test working with an LLM.
The code isn't great but it was an okay way to get the draft out of the door.

In addition the container image for the app currently only supports `arm64` and the one for Nginx only `amd64` because I really only need the arm64 one.
Support for both architectures will come however.

The app is written in #Django and packaged with #Nix

https://github.com/Nebucatnetzer/meta-search

TIL that when I prefix any URL with a question mark and a space, @kagihq will give me some context about it. For example, searching for `? https://zottmann.org/2025/04/17/sarcasm-is-poison.html ` summarizes my blog post and gives one or two links which *aren't in my post* but relate to the subject.

This is neat.

Why yes, I do endorse #Kagi , and for over two years now. https://mastodon.social/@kagihq/114886595957287479

Kagi's algorithms prioritize organic, non-commercial sources over ad-heavy or tracker-riddled sites. This really helps to surface unique, human-generated content from smaller blogs, authentic sites and forums.