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: Buoyant
providerId: buoyant
generated: '2026-09-04'
method: searched
source: >-
  https://www.buoyant.io/pricing, https://docs.buoyant.io/, https://linkerd.io/docs/reference/,
  and a live probe of https://api.buoyant.cloud (2026-09-04).
created: '2026-05-04'
modified: '2026-09-04'
limit_count: 0
note: >-
  REPLACES a 2026-05-04 bulk-sweep scaffold that invented free/professional/enterprise
  request-per-minute quotas and X-RateLimit-* response headers for a surface that does
  not exist. Buoyant publishes NO rate limits for any of its own endpoints, and no
  rate-limit response headers were observed. The published contract is in-cluster gRPC
  (see grpc/) — a customer runs it on their own infrastructure, so there is no shared
  metered endpoint to throttle. api.buoyant.cloud is a gRPC agent endpoint whose only
  documented callers are the customer's own Buoyant Cloud agents; it publishes no
  documented limits. An honest zero, not an omission.
limits: []
headers:
  published: false
  observed: []
responseCodes:
  published: false
inverse_capability:
  note: >-
    Buoyant sells rate-limit HANDLING rather than publishing rate limits. Linkerd
    provides HTTPLocalRateLimitPolicy for meshed services, and Buoyant Enterprise for
    Linkerd 2.20.0 added rate-limit-aware load balancing in which circuit breaking and
    endpoint selection account for HTTP 429 responses returned by the proxied service.
  sources:
    - https://linkerd.io/docs/reference/rate-limiting/
    - https://linkerd.io/docs/tasks/rate-limit-aware-load-balancing/
    - https://docs.buoyant.io/release-notes/buoyant-enterprise-linkerd/enterprise-2.20.0/
maintainers:
  - FN: Kin Lane
    email: kin@apievangelist.com