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: Cvent Event Cloud
providerId: cvent-event-cloud
generated: '2026-08-13'
method: searched
source: https://developers.cvent.com/docs/rest-api/guides/handling-rate-limits
sources:
  - https://developers.cvent.com/docs/rest-api/guides/handling-rate-limits
  - https://github.com/cvent/rest-sdks/blob/main/cvent-public-spec/openapi.yaml
created: '2026-05-04'
modified: '2026-08-13'
supersedes: >-
  The 2026-05-04 bulk-sweep scaffold, which asserted invented per-tier limits (10 rpm free, 100 rpm
  professional, 1000 rpm enterprise) and headers Cvent does not return. Every value below is now read
  from Cvent's own published usage-tier table and rate-limit guide.
limit_count: 6
description: >-
  Cvent enforces two independent limits per account: a daily call quota and a per-second call rate,
  both set by the account's usage tier. Every request counts against the daily quota — including
  requests that are themselves rejected with 429.

tiers_published: true
tiers_source: >-
  info.description of Cvent's published OpenAPI ("Usage Tiers" table) and the Handling Rate Limits guide.
self_service_discovery:
  operation: getUsageTier
  method: GET
  path: /ea/usage/tier
  description: Returns the usage tier assigned to the calling account.
  usage_metrics:
    method: GET
    path: /ea/usage
    description: Poll current usage to monitor quota consumption.

headers:
  limit: X-RateLimit-Limit
  remaining: X-RateLimit-Remaining
  reset: X-RateLimit-Reset
  retryAfter: null
  policy: null
  note: >-
    Cvent documents these three headers as present on every response, not only on 429s, so a client
    can watch its own headroom. Cvent returns NO Retry-After header — the back-off interval must be
    computed by the client from X-RateLimit-Reset or from Cvent's published back-off recipe.

responseCodes:
  throttled: 429
  quotaExceeded: 429
  note: >-
    Both failure modes share status 429; they are distinguished by the message body. "Limit Exceeded"
    means the daily quota is gone, "Too Many Requests" means the per-second rate was exceeded.

limits:
  - tier: Free
    name: Free tier daily quota
    scope: account
    metric: calls_per_day
    limit: 1000
    timeFrame: day
    reset: 00:00 GMT (midnight UTC)
    applies: [Cvent REST APIs]
  - tier: Free
    name: Free tier call rate
    scope: account
    metric: calls_per_second
    limit: 2
    burst: 1
    timeFrame: second
    applies: [Cvent REST APIs]
  - tier: Standard
    name: Standard tier daily quota
    scope: account
    metric: calls_per_day
    limit: 15000
    timeFrame: day
    reset: 00:00 GMT (midnight UTC)
    applies: [Cvent REST APIs]
  - tier: Standard
    name: Standard tier call rate
    scope: account
    metric: calls_per_second
    limit: 10
    burst: 10
    timeFrame: second
    applies: [Cvent REST APIs]
  - tier: Premium
    name: Premium tier daily quota
    scope: account
    metric: calls_per_day
    limit: 500000
    timeFrame: day
    reset: 00:00 GMT (midnight UTC)
    applies: [Cvent REST APIs]
  - tier: Premium
    name: Premium tier call rate
    scope: account
    metric: calls_per_second
    limit: 25
    burst: 25
    timeFrame: second
    applies: [Cvent REST APIs]

policies:
  - name: Daily quota exhaustion
    description: >-
      On a daily-quota 429, stop making requests and wait until after midnight UTC. Retrying does
      not help and each rejected request still consumes quota.
  - name: Per-second back-off
    description: >-
      Cvent's published recipe: wait 2 seconds plus a random 1–1000 ms, then double the wait for each
      retry up to a 16-second maximum. After five failed attempts, stop and log the error for review.
    initial_wait_seconds: 2
    jitter_ms: [1, 1000]
    multiplier: 2
    max_wait_seconds: 16
    max_attempts: 5
  - name: Rejected requests still count
    description: >-
      "Every request, even those that return a 429 error, still counts against your quota." A naive
      retry loop therefore burns the remaining daily allowance faster than doing nothing.
  - name: Proactive monitoring
    description: >-
      Cvent recommends reading the X-RateLimit-* headers on every response and setting usage alerts,
      rather than discovering limits by hitting them.

sdk_support:
  automatic_retries: true
  note: >-
    All three first-party SDKs ship configurable retry/back-off; the retries behaviour is applied to
    the generated clients by Cvent's own retries_overlay.yaml
    (overlays/cvent-event-cloud-retries-overlay.yaml).

spec_coverage:
  operations_declaring_429: 456
  operations_total: 458
  note: >-
    456 of 458 operations declare a 429 response in the OpenAPI, so a generated client has typed
    handling for throttling on nearly the whole surface.

maintainers:
  - FN: Kin Lane
    email: kin@apievangelist.com