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: King's College London
providerId: kings-college-london
created: '2026-06-03'
modified: '2026-08-19'
reconciled: false
tags:
- Education
- Higher Education
- University
- Rate Limiting
description: >-
  King's College London publishes no rate-limit table on any public surface. Re-checked
  2026-08-19: the one institution-operated API, the e-Research AI Hub at
  ai.create.kcl.ac.uk/api/v1, is metered — new accounts get an allocation of free tokens and
  sustained use is billed against a project budget — but the limits, the quota units and the
  429 shape are all documented behind Microsoft Entra ID sign-in at ai.create.kcl.ac.uk/docs and
  could not be read. The 401 responses carry no RateLimit or Retry-After headers. The Pure
  OAI-PMH endpoint at kclpure.kcl.ac.uk/ws/oai answers unauthenticated and publishes no stated
  limit; harvest politely and honor Retry-After. No 429 was observed on any King's host.
notes: >-
  Nothing here is asserted from a specification — King's publishes none. The AI Hub metering is
  read from King's own e-Research documentation; every numeric limit remains unknown.
sources:
- https://raw.githubusercontent.com/api-evangelist/kings-college-london/refs/heads/main/apis.yml
- https://docs.er.kcl.ac.uk/CREATE/ai_hub/
- https://ai.create.kcl.ac.uk/api/v1/models
- https://kclpure.kcl.ac.uk/ws/oai?verb=Identify
responseCodes:
  throttled: 429
limits:
- name: e-Research AI Hub token allocation
  scope: account
  metric: tokens
  limit: unknown
  notes: >-
    New AI Hub accounts receive an allocation of free tokens; beyond it usage is metered against
    a project budget. The numeric allocation is behind institutional sign-in.
- name: Pure OAI-PMH harvesting
  scope: client
  metric: requests-per-minute
  limit: unpublished
  notes: >-
    No published limit and no 429 observed across Identify, ListMetadataFormats, ListSets and a
    94KB ListRecords page. Standard OAI-PMH resumption-token pacing applies.
policies:
- name: Backoff Strategy
  description: Exponential backoff with jitter; honor Retry-After; harvest during off-peak hours.
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com