generated: '2026-08-26'
method: searched
source: https://docs.rampnetwork.com/rate-limiting
docs: https://docs.rampnetwork.com/rate-limiting
limit_count: 2
scope_model: per origin IP address (not per API key and not per account)
limits:
  - scope: origin-ip
    window: 1 minute
    limit: 100
    burst: null
    applies_to: all Host API endpoints
  - scope: origin-ip
    window: 15 minutes
    limit: 1000
    burst: null
    applies_to: all Host API endpoints
    penalty: >-
      Exceeding the 15-minute limit can result in the origin IP being blocked for
      the following hour, per the published documentation.
response_headers:
  observed: []
  note: >-
    NO rate-limit response headers are published, and none were observed. A live
    unauthenticated GET of https://api.rampnetwork.com/api/host-api/v3/currencies on
    2026-08-26 returned HTTP 200 with content-type, content-length, etag, date,
    x-envoy-upstream-service-time, server (istio-envoy), via, vary, cache-control
    and alt-svc — no X-RateLimit-*, no RateLimit-*, no Retry-After. An agent
    therefore has NO runtime signal for remaining quota and can only discover the
    limit by tripping it. This is the single largest agent-readiness gap in the
    surface.
exhaustion:
  status_code: null
  note: >-
    The docs do not state the status code returned on exhaustion, and the limit was
    deliberately not tripped during profiling.
increase_path:
  contact: partner@ramp.network
  note: >-
    "If you are generating valid traffic exceeding the defined limits, please
    contact us at partner@ramp.network for higher allowance."
evidence:
  - url: https://docs.rampnetwork.com/rate-limiting
    status: 200
  - url: https://api.rampnetwork.com/api/host-api/v3/currencies
    status: 200
    detail: live response header inspection — no rate-limit headers present