Kong · Rate Limits
Kong Rate Limits
Kong's pricing constrains Konnect through plan-level quotas (Plus = 1M API requests/month included, $200/M overage; AI Gateway = 5 LLM models included, $100/extra). The Konnect control-plane API itself does not publicly publish per-second / per-minute throttling thresholds in its developer documentation; consumers should code against the standard 429 / Retry-After contract. Kong's flagship product is itself a rate-limiting plugin set, applied to data-plane traffic the customer terminates.
8 Limits
Throttle: 429
Rate LimitingAPI GatewayAPI Management
Limits
Konnect Plus included API requests organization
1000000
Overage billed at $200 per additional 1M requests on the Plus plan.
Konnect Plus AI Gateway models organization
5
Each additional unique LLM model billed at $100/month on the Plus plan.
Konnect Plus Serverless gateways organization
5
Konnect Plus Hybrid gateways organization
2
Konnect Plus Dedicated Cloud gateways organization
2
Konnect Plus Developer Portals organization
2
Konnect Plus Published APIs per Portal portal
20
Konnect control-plane API throttling organization
see Kong Konnect API documentation; numeric thresholds not publicly published
Policies
Plan quota vs gateway throttling
Plus-plan request and model caps are billing thresholds, not hard throttles — exceeding them creates overage charges, not 429 responses to your end users.
Backoff
For 429/5xx from the Konnect control-plane API, retry with exponential backoff and jitter; honor Retry-After.
Data-plane rate limiting
Customers configure their own Kong rate-limiting/rate-limiting-advanced plugins on data-plane gateways to throttle their consumers; that configuration is independent of the Konnect platform's own control-plane behavior.