specification: API Commons Rate Limits
specificationVersion: '0.1'
schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/RateLimits
provider: Calico
providerId: calico
created: '2026-05-04'
modified: '2026-09-05'
generated: '2026-09-05'
method: searched
source: >-
  https://docs.tigera.io/calico/latest/reference/ ,
  https://www.tigera.io/tigera-products/calico-cloud-pricing/ ,
  openapi/*.yml
provenance_note: >-
  REPLACED 2026-09-05. Every previous version of this file was a bulk-sweep SCAFFOLD (roadmap#35)
  asserting free/professional/enterprise tiers at 10/100/1,000 requests per minute with
  X-RateLimit-* response headers. None of that is true of Calico. It publishes no rate limits, it
  returns no X-RateLimit-* headers, and it has no tiers on this API. The invented numbers are gone.
description: >-
  Calico publishes NO API rate limits, and the honest zero is the finding. The
  projectcalico.org/v3 API is served by the OPERATOR'S OWN Kubernetes API server — Calico is not the
  one answering the request, so there is nobody in Calico's position to impose a limit. Whatever
  throttling a caller meets is the cluster's own, configured by the cluster's own admins.
limit_count: 0
provider_published_limits: false
tiers: none — this API has no tiers; see plans/calico-plans-pricing.yml
limits: []
headers:
  limit: null
  remaining: null
  reset: null
  retryAfter: Retry-After
  policy: null
  note: >-
    No X-RateLimit-* or RFC 9331 RateLimit-* headers. `Retry-After` is returned by the Kubernetes API
    server on a 429 from API Priority and Fairness — that is a platform header, not a Calico one, and
    it is the only runtime throttling signal an agent will see.
responseCodes:
  throttled: 429
  note: >-
    429 originates from the Kubernetes API server's API Priority and Fairness (APF) admission
    control, not from Calico. It means the cluster's own flow-schema concurrency was exhausted.
runtime_mechanism:
  name: Kubernetes API Priority and Fairness (APF)
  owner: the cluster operator
  configurable_by_provider: false
  description: >-
    APF classifies requests into flow schemas and priority levels and sheds load with 429 +
    Retry-After when a level is saturated. Every Calico API call is subject to whatever the cluster
    has configured, and none of it is set or published by Calico.
  agent_guidance: >-
    Honour Retry-After. Back off exponentially with jitter on 429. Do not assume any published
    numeric ceiling exists to reason about — there is none.
commercial_editions:
  note: >-
    Calico Cloud is priced per vCPU-hour, not per API request (see
    plans/calico-plans-pricing.yml). No request-rate limits are published for it either.
policies: []
maintainers:
  - FN: Kin Lane
    email: kin@apievangelist.com