generated: '2026-08-13'
method: searched
source: >-
  https://developer.clari.com/documentation/external_spec (Clari Revenue API v5.0.0
  description + operation responses) and https://api-doc.copilot.clari.com/spec.yaml
  (Clari Copilot API info.description, section "# Rate limit")
name: Clari Rate Limits
description: >-
  Clari publishes limits in prose inside its OpenAPI descriptions and returns NO
  rate-limit response headers. The two APIs are limited differently. Critically, the
  two headline Revenue API limits — concurrent export jobs and the monthly export
  quota — are TENANT-SPECIFIC values that Clari does not publish a number for; they
  are set per organization and adjusted through a CSM, and the only way to learn yours
  is to call GET /admin/limits.
specificationVersion: '0.1'
created: '2026-06-13'
modified: '2026-08-13'
limit_count: 6

response_headers:
  published: false
  x_ratelimit: false
  ratelimit: false
  retry_after: false
  note: >-
    Five operations declare HTTP 429 (ingestBulk, ingestIncremental, ingestAddPicklist,
    ingestGetPicklist, ingestJobStatus) and NOT ONE declares Retry-After, X-RateLimit-*
    or RateLimit-*. A client has no runtime budget signal and must implement blind
    exponential backoff. This is the most consequential gap in Clari's rate-limit
    posture — the numbers exist, the runtime signal does not.
  programmatic_signal:
    operation: limits
    path: GET /admin/limits
    returns: monthly quota plus all active and completed jobs contributing to it
    note: Must be polled deliberately; nothing pushes it.

status_codes:
  rate_limited: 429
  quota_exhausted: 400
  note: >-
    Quota exhaustion on export queueing is reported as 400, NOT 429. Clari's own text:
    "Before creating an export job request, make sure you have available quota. If you
    have completed your monthly quota, your requests to Clari API will fail."

rateLimits:
- name: Revenue API — concurrent export job limit
  api: Clari Revenue API
  scope: Organization
  type: Concurrency
  limit: null
  unit: concurrent jobs
  published_value: false
  source: OpenAPI info.description
  note: >-
    A concurrent job limit exists and is enforced, but Clari publishes no number.
    Verbatim: "Use the /admin/limits endpoint to find out which limit you have hit. If
    you find yourself hitting the concurrent limit often, try adjusting your request
    parameters. Specifically, set SCHEDULE to ONCE or MAX. This parameter will initiate
    retries without extra effort."
- name: Revenue API — monthly export quota
  api: Clari Revenue API
  scope: Organization
  type: Quota
  limit: null
  period: rolling 30 days
  unit: export jobs
  published_value: false
  source: OpenAPI schema description on GET /admin/limits
  note: >-
    Verbatim: "The maximum number of exports that can be run in a rolling 30 day
    period. The monthly quota is at the organization level and can be adjusted as
    needed. Please work with your CSM if you'd like to adjust your quota." Cancelling
    an in-progress job via PATCH /export/jobs/{jobId} frees quota back up.
- name: Revenue API — ingestion request rate
  api: Clari Revenue API
  scope: Per API Token
  type: Rate
  limit: 100
  period: per second
  unit: requests
  published_value: true
  source: 429 response description on GET /ingest/job/{jobId}
  note: 'Verbatim: "API rate limit has been exceeded. The api limit is 100 requests per second."'
- name: Revenue API — ingest records per request
  api: Clari Revenue API
  scope: Per Request
  type: Payload
  limit: 100
  unit: records
  published_value: true
  source: Ingestion API description
  note: 'Verbatim: "It supports uploading up to 100 records per request."'
- name: Revenue API — audit events page size
  api: Clari Revenue API
  scope: Per Request
  type: Pagination
  limit: 1000
  defaultValue: 100
  unit: records per request
  published_value: true
  source: '`limit` parameter description on GET /audit/events'
  note: 'Verbatim: "Integer between 1 and 1000. Defaults to 100." A `nextLink` is provided for navigation.'
- name: Copilot API — request rate
  api: Clari Copilot REST API
  scope: Per API Key
  type: Rate
  limit: 10
  period: per second
  unit: requests
  published_value: true
  source: Copilot OpenAPI info.description, "# Rate limit"
  note: 'Verbatim: "Default rate limit is 10 per second."'
- name: Copilot API — weekly request quota
  api: Clari Copilot REST API
  scope: Per API Key
  type: Quota
  limit: 100000
  period: weekly
  unit: requests
  published_value: true
  source: Copilot OpenAPI info.description, "# Rate limit"
  note: 'Verbatim: "Also 100k requests per week is also imposed. Week starts every Sunday 0 GMT." The reset boundary being an absolute wall-clock instant rather than a rolling window is unusual and worth planning around.'

authentication:
  revenueAPI:
    type: API Key (header)
    header: apikey
    partner_header: partnerkey
    notes: >-
      Generated in Account Settings > API Token. Shown once. Revoking breaks every
      integration using it; deactivating a user revokes all tokens they created.
  copilotAPI:
    type: API Key (dual header)
    headers:
    - X-Api-Key
    - X-Api-Password
    notes: Both required. Found in Workspace Settings > Integrations > Clari Copilot API.
  see: authentication/clari-authentication.yml

x-provenance-note: >-
  This file previously asserted a concurrent-job limit of 3 and a monthly export quota
  of 1000. Neither number appears anywhere in Clari's published documentation — both
  limits are tenant-specific and unpublished. They were replaced with null plus the
  verbatim provider text on 2026-08-13.