Kagi
Kagi · Search · US · kagi.com
services/kagi.yamlVendor data-handling terms
“When using third party AI model providers, we are using privacy-respecting services that do not save the data or use it for training, whenever possible. / Unless stated below, and in a similar vein as Search, requests to AI tools are not saved or associated with a customer account (they may be retained for a short period of time as a part of request debugging, explained above).”
“By default and unless otherwise stated, Kagi Assistant automatically deletes conversation threads after one day. / We may store web requests made by user browser temporarily, with strict retention periods, for debugging purposes, and in a manner that they are not linked to an account.”
No hosting region is named; Kagi is a US company.
“When viewing images, videos or other media results, or using the Kagi Assistant, we protect your privacy by proxying all connections and data through our servers.”
Domains and endpoints
observed means seen in Unseen deployments; vendor-documented means listed by the vendor. Vendors do not publish complete lists.
| Host | Role | Source |
|---|---|---|
| kagi.com | app | observed |
Assessment
Reasoning: A paid search engine with an assistant that fronts several models. Requests are not tied to accounts, threads expire after a day, and providers are selected for no-training terms with an explicit "whenever possible" caveat. Prompt text still reaches third-party models.
Flags: Provider no-training terms are best-effort, not guaranteed · Region not stated · Assessed 2026-09-16. The assessment is Unseen's; the terms above are the vendor's.
Changelog
- 2026-09-16Documented. Provider, retention and proxying terms verified against the Kagi privacy policy.
- 2026-09-15Listed from the Unseen catalogue with observed domains.
Corrections
Spotted something wrong or out of date? Send it with the source. Changes are checked and recorded in the changelog.