generated: '2026-08-14'
method: searched
source: https://developers.facebook.com/docs/graph-api/overview/rate-limiting/
docs: https://developers.facebook.com/docs/graph-api/overview/rate-limiting/
also: https://developers.facebook.com/docs/marketing-api/guides/lead-ads/retrieving
note: >-
  Meta publishes rate limits as FORMULAS, not fixed numbers — the quota scales with the
  size of the thing you are operating on. For Facebook Lead Ads the governing bucket is
  the LeadGen business-use-case limit, which scales with leads generated: a form that has
  captured few leads has a correspondingly small call budget, which is the opposite of the
  usual "big customer, big quota" intuition and the single most important operational fact
  about this API.
limits:
  - id: leadgen-buc
    scope: business use case — LeadGen
    applies_to: Lead ads leads retrieval (this product)
    window: 24 hours
    formula: Calls within 24 hours = 4800 * Leads Generated
    header: X-Business-Use-Case-Usage
    error_code: 80005
    primary: true
  - id: pages-buc
    scope: business use case — Pages (Page or System User access token)
    applies_to: Page-scoped edges such as /{page-id}/leadgen_forms and /{page-id}/subscribed_apps
    window: 24 hours
    formula: Calls within 24 hours = 4800 * Number of Engaged Users
    header: X-Business-Use-Case-Usage
    error_code: 80001
  - id: pages-buc-user-token
    scope: business use case — Pages (User access token)
    window: 24 hours
    error_code: 32
    header: X-Business-Use-Case-Usage
  - id: ads-management-standard
    scope: business use case — Ads Management, Standard Access
    window: 1 hour
    formula: Calls within one hour = 300 + 40 * Number of Active ads
    header: X-Business-Use-Case-Usage
    error_code: 80004
    error_subcode: 2446079
  - id: ads-management-advanced
    scope: business use case — Ads Management, Advanced Access
    window: 1 hour
    formula: Calls within one hour = 100000 + 40 * Number of Active ads
    header: X-Business-Use-Case-Usage
    error_code: 80004
    error_subcode: 2446079
  - id: ads-insights-standard
    scope: business use case — Ads Insights, Standard Access
    window: 1 hour
    formula: Calls within one hour = 600 + 400 * Number of Active ads - 0.001 * User Errors
    header: X-Business-Use-Case-Usage
    error_code: 80000
    error_subcode: 2446079
    note: >-
      User errors REDUCE the quota. Sloppy clients are penalised, which makes error
      handling a rate-limit concern on this platform.
  - id: ads-insights-advanced
    scope: business use case — Ads Insights, Advanced Access
    window: 1 hour
    formula: Calls within one hour = 190000 + 400 * Number of Active ads - 0.001 * User Errors
    header: X-Business-Use-Case-Usage
    error_code: 80000
    error_subcode: 2446079
  - id: platform-app
    scope: per app (platform rate limit)
    window: rolling 1 hour
    formula: Calls within one hour = 200 * Number of Users
    users_basis: daily active users (weekly/monthly used during low-activity periods)
    header: X-App-Usage
    error_code: 4
  - id: platform-user
    scope: per user (platform rate limit)
    window: rolling 1 hour
    limit: undisclosed
    note: Meta does not publish the per-user value; it is shared across every app the user uses.
    header: X-App-Usage
    error_code: 17
  - id: ad-account-legacy
    scope: per ad account (Ads API v3.3 and older)
    header: X-Ad-Account-Usage
    error_code: 17
    error_subcode: 2446079
    note: Legacy surface; retained because older SDK pins still hit it.
headers:
  - name: X-App-Usage
    encoding: JSON object
    fields:
      call_count: percentage of calls used within the rolling one-hour window
      total_cputime: percentage of the CPU-time allotment used
      total_time: percentage of the total processing-time allotment used
    note: Any field reaching 100 throttles the app.
  - name: X-Business-Use-Case-Usage
    encoding: JSON object keyed by business ID
    fields:
      business-id: the associated business identifier
      call_count: percentage of allowed calls used
      total_cputime: CPU-time percentage
      total_time: processing-time percentage
      type: the rate-limit category applied (e.g. leadgen, pages, ads_management)
      estimated_time_to_regain_access: minutes until throttling stops
      ads_api_access_tier: 'access tier — ads_insights / ads_management only'
    note: >-
      estimated_time_to_regain_access is the closest thing this API has to Retry-After.
      It is inside a JSON header value, not a standard header, so a generic HTTP client
      will not find it.
  - name: X-Ad-Account-Usage
    encoding: JSON object
    fields:
      acc_id_util_pct: percentage of calls used before the limit is reached
      reset_time_duration: seconds until the rate limit resets
      ads_api_access_tier: development_access | standard_access
    note: Ads API v3.3 and older.
retry_after_header: false
rate_limit_standard_headers: false
standard_headers_note: >-
  Meta implements neither the draft IETF RateLimit-* headers nor Retry-After. Every signal
  is a proprietary X-* header carrying a JSON document, so an agent must parse the header
  VALUE as JSON to get the remaining budget.
status_on_exhaustion:
  http_status: 400
  body_type: OAuthException
  note: >-
    Throttling does NOT return 429. It returns HTTP 400 with the throttle encoded in
    error.code. A client that keys retry behaviour on 429 will never back off against
    this API.
lead_retrieval_note: >-
  The lead-ads retrieval guide states the bulk-read limit as "200 multiplied by 24 then
  multiplied by the number of leads created in the past 90 days" — the same 4800 x leads
  formula expressed per-90-day-window.
limit_count: 10