Skip to main content

​
Per-user rate limit

Enforced per external_user_id, not across your whole tenant.

​
Per-key request limit

Each tenant API key also has its own rate_limit_per_minute setting. This is enforced across tenant memory, user, agent, API-key, and tenant routes and is independent of the per-user and plan limits. A 429 caused by this boundary returns details.scope: "api_key"; the shared tenant write ceiling returns details.scope: "tenant". Use separate keys per backend service so one integration cannot consume every other service’s key-level capacity. Redis failure leaves requests available; quota, identity, and database authorization checks still run normally.

​
Plan limits

Monthly calls and tokens reset at the start of each billing month. Write calls are add() requests. Retrieval is unlimited on all current plans.

​
When you hit a limit

Behavior depends on your tenant’s overage policy — see Degradation modes for the exact quota mode handling and response headers.