Your users need an identity to use features that modify data, like chatting,
commenting, reacting, and voting. (Reading is anonymous.) There are three ways to
provide one — this page covers each and when to use it.
Your app’s API key is a separate identity, always present; this page is about the
end user’s.
The simplest start. You have no user system of your own and don’t want to
build one. We give you the user database — you register each user with their
email or a display name, and we store them. Drop-in widgets mean no backend
of your own. Your data is yours: export it anytime, no lock-in.
2 · You mint the identity
You already manage your own users. Your server makes one call to mint
a token for a signed-in user. From then on, anyone using our widgets and
components acts as exactly the identity you handed us. Your server stays the
source of truth.
3 · Your identity provider
You use a third-party login — Auth0, Amazon Cognito, Firebase, Okta,
Google, or any OIDC provider. Point us at it once, and the token your
users already carry just works. No second login, no token-minting code.
If part of your audience only needs to watch the conversation — a public
liveblog or a watch-along feed — they can read room events with your application
key alone, no user token required. Reads work anonymously; posting, reacting, and
reporting need one of the three models above.