generated: '2026-08-14'
method: searched
source: https://developers.crossbeam.com/
docs:
  - https://developers.crossbeam.com/
  - https://www.crossbeam.com/legal/security-policy
  - https://www.crossbeam.com/pricing
provider: Crossbeam
providerId: crossbeam
limit_count: 0
note: >-
  Crossbeam publishes NO numeric API rate limits. Its Partner API documentation says verbatim:
  "Rate limits are defined in our security policy and are subject to change with our own
  discretion." The referenced security policy (https://www.crossbeam.com/legal/security-policy)
  was read in full and contains no rate, quota, throttle, header or status-code detail — it is
  a security-controls document. No X-RateLimit-*, RateLimit-* or Retry-After header is
  documented anywhere, and no 429 behaviour is described for the REST API. An agent therefore
  cannot know its budget before it starts, and cannot read a runtime signal to back off.
  Recorded as an honest zero rather than omitted.
headers:
  documented: false
  observed: null
  note: >-
    Not observable without credentials — every Partner API endpoint requires an OAuth bearer
    token plus an Xbeam-Organization header, so no unauthenticated 200 exists on which to read
    live limit headers.
response_codes:
  throttled: undocumented
  note: >-
    The only retry-status list Crossbeam publishes is for WEBHOOK delivery, not for inbound API
    calls: Crossbeam retries a webhook POST to your receiver on 408, 429, 500, 502, 503 and 504,
    then expects you to recover the missed message from the /v1/signals/* REST endpoints. That
    is Crossbeam retrying YOU; it says nothing about limits it applies to your calls.
limits: []
related_quotas:
  - name: Unique record exports
    scope: account (per plan, per year)
    kind: commercial-quota
    values:
      connector: 5000
      supernode: 25000
      enterprise: 100000
    note: >-
      Not a rate limit, but the only published consumption ceiling that API traffic draws
      against — Crossbeam's docs state REST reads and webhook deliveries both count toward the
      Record Export allowance. See plans/crossbeam-plans-pricing.yml.
  - name: Page size
    scope: per request
    kind: pagination-bound
    default: 25
    maximum: 1000
    note: >-
      The `limit` query parameter on the records, overlaps, signals and audit-log endpoints.
      Default 25 (1000 on /v1/audit-logs), maximum 1000. A pagination bound, not a throttle.
      See conventions/crossbeam-conventions.yml.
guidance:
  - Filter server-side with the search endpoints instead of listing and filtering client-side.
  - Subscribe to Signals webhooks rather than polling /v1/signals/* — polling burns record-export allowance.
  - Treat the absence of a published limit as a reason to back off conservatively, not as permission to hammer.
checked: '2026-08-14'