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: Factiva
providerId: factiva
created: '2026-05-04'
modified: '2026-08-13'
generated: '2026-08-13'
method: searched
source:
  - https://developer.dowjones.com/documents/factiva_integration
  - https://developer.dowjones.com/documents/site-docs-getting_started-api_essentials
tags:
  - Business Intelligence
  - News
  - Research
  - Rate Limiting
description: >-
  Factiva publishes no rate limits. A full-text search of the 337 Factiva documents on the Dow
  Jones Developer Platform returns zero matches for "rate limit", "rate-limit", "throttle",
  "X-RateLimit", "RateLimit-Policy" or "Retry-After", and no operation in any of the three
  published specifications declares a 429 response. An integrator has no published number to
  design against and no runtime header to react to; the only backpressure signal available is a
  5xx. Whatever limits exist are enforced silently or set per contract.
limit_count: 0
limits: []
published: false
headers: {}
header_note: >-
  No RateLimit-* or X-RateLimit-* response headers are documented, and no Retry-After is
  documented. Recorded as absent, not unknown — the docs were searched.
responseCodes:
  throttled: null
  note: >-
    429 does not appear in any published Factiva specification or documentation. Errors observed
    and documented are 400, 401, 403 (gateway, code 1011001), 404, 500 and 503.
documented_quotas:
  - name: Retrieval response_limit
    scope: per-request
    metric: results
    limit: 100
    type: request-cap
    description: Maximum number of results the Factiva Retrieval endpoint will return in one call.
    source: https://developer.dowjones.com/documents/factiva_integration-factiva_retrieval_api-endpoints-retrieval
  - name: Retrieval search filter value length
    scope: per-request
    metric: characters
    limit: 2000
    type: request-cap
    description: Maximum length of a single search_filters value on the Retrieval endpoint.
    source: https://developer.dowjones.com/documents/factiva_integration-factiva_retrieval_api-endpoints-retrieval
  - name: Retrieval metrics identifier length
    scope: per-request
    metric: characters
    limit: 32
    type: request-cap
    description: Maximum length of user_id, work_id and application_id in metrics_data.
    source: https://developer.dowjones.com/documents/factiva_integration-factiva_retrieval_api-endpoints-retrieval
  - name: GenAI article deduplication window
    scope: per-account
    metric: hours
    limit: 24
    type: billing-window
    description: >-
      Duplicate consumption of the same article within 24 hours is deduplicated for usage
      purposes. Note the token-usage endpoint's own caveat that its returned values do not yet
      apply this rule.
    source: https://developer.dowjones.com/documents/factiva_integration-factiva_retrieval_api-endpoints-token-usage
  - name: Retrieval content recency floor
    scope: per-request
    metric: date
    limit: '2025-01-01'
    type: data-boundary
    description: >-
      Retrieval and usage-metrics responses never include items older than 2025-01-01, regardless
      of the requested date range.
    source: https://developer.dowjones.com/documents/factiva_integration-factiva_retrieval_api-endpoints-retrieval
usage_reporting:
  description: >-
    In place of rate limiting, Factiva puts the meter in the customer's hands: token consumption
    is reported through POST /content/gen-ai/token-usage, aggregated twice daily at 02:15 UTC and
    14:15 UTC. It is a reporting surface, not a throttling one — it tells you what you used, not
    what you have left.
  endpoint: https://api.dowjones.com/content/gen-ai/token-usage
searched:
  documents_searched: 337
  keywords: [rate limit, rate-limit, throttle, X-RateLimit, RateLimit-Policy, Retry-After, '429']
  matches: 0
  checked: '2026-08-13'
provenance_note: >-
  Replaces the 2026-05-04 bulk-sweep scaffold (roadmap#35), which asserted 10/100/1000 requests
  per minute across invented Free/Professional/Enterprise tiers and a full X-RateLimit-* header
  set. Dow Jones publishes none of that.
maintainers:
  - FN: Kin Lane
    email: kin@apievangelist.com