Kagi’s first 50 searches no longer need an account

Kagi has removed the account requirement from the first 50 searches on its website. For a paid search engine that depends heavily on recommendations, this is a small product change with a clear consequence: someone can follow a friend's suggestion and judge a results page before deciding whether to register.

The November 5 changelog says registration provides 100 further trial searches. The change does not make Kagi an unlimited free service, and it should not be confused with unrestricted access to all paid-plan features. Kagi itself describes the account-free discovery experience as reduced compared with a paid subscription.

The first comparison no longer needs an account

Search is a habit built around an unusually low-effort action: type a question and inspect what appears. Requiring registration before that first action asks a prospective user to commit before seeing whether the engine handles the queries they actually make.

Fifty searches are enough to test several recurring needs: a difficult product name, a technical error, a local query or a page remembered only approximately. They are not enough to establish that one engine is universally better. The useful comparison is between the results you need and the effort required to reach them, rather than a favourite demonstration chosen by either company.

Kagi says it does not run advertising and relies on word of mouth. Removing the initial account gate fits that distribution model: a recommendation becomes easier to act on. Whether it leads to a subscription still depends on the quality of the experience and the value a reader assigns to paying for search.

Android makes the remaining searches visible

The same release updates Kagi's Android app with a redesigned home screen that shows the remaining account-free search allowance. It also adds a first-run tour, a rebuilt voice-search screen with language selection, home-screen widgets and a voice-search Quick Settings tile.

These changes address a different source of friction: making a second search engine convenient enough to reach repeatedly. A browser default is powerful partly because it is always close at hand. A widget or quick tile cannot prove better results, but it can make a fair comparison less dependent on remembering to open another website.

The changelog also adds a warning when a browser unexpectedly logs out a paid user and returns them to the discovery flow. That is an important distinction for existing subscribers: a restricted result should not automatically be interpreted as the paid product losing a feature.

Kagi's offer is most useful as a low-commitment experiment with real queries. The new threshold asks users to judge the search before judging the signup form. For a service selling an alternative to an entrenched default, that is a sensible place for the decision to begin.