spec: API Commons Rate Limits 0.1
provider: Ada
providerId: ada
created: 2026-06-12
modified: 2026-08-14
generated: '2026-08-14'
method: searched
source: https://docs.ada.cx/reference/introduction/limits
url: https://docs.ada.cx/reference/introduction/limits
limit_count: 6
scope: per Ada instance
headers:
  published: false
  request: []
  response: []
  retry_after: false
  note: >-
    Re-read 2026-08-14. Ada publishes the numbers but NOT the runtime signal: no RateLimit-*,
    no X-RateLimit-*, no Retry-After, and no remaining/reset field in any response. Neither the
    limits page nor any of the four OpenAPI documents declares a rate-limit response header. An
    agent therefore cannot see itself approaching a limit — it only discovers it has crossed one
    by receiving a 429, and must then guess the wait. Ada's own guidance is "implement retry
    logic with exponential backoff and jitter", which is the workaround for exactly this gap.
  exhaustion_status: 429
  oversize_status: 413
notes: >
  Ada enforces global rate limits across all REST API families to ensure high availability.
  Individual API families may apply their own elevated or additional local limits.
  Exceeding any limit returns HTTP 429 Too Many Requests. Ada recommends implementing
  retry logic with exponential backoff and jitter. Rate limits were increased for the
  End Users API and Conversations API in March 2026 (doubled from 30,000 to 60,000 per day).

global:
  throttledStatus: 429
  retryAfter: true
  retryStrategy: exponential-backoff-with-jitter
  limits:
    - name: Requests Per Second (Global Default)
      unit: requests/second
      limit: 10
      scope: global
    - name: Requests Per Minute (Global Default)
      unit: requests/minute
      limit: 100
      scope: global
    - name: Requests Per Day (Global Default)
      unit: requests/day
      limit: 10000
      scope: global
    - name: Maximum Request Body Size
      unit: MB
      limit: 10
      errorStatus: 413

apis:
  - name: End Users API
    limits:
      - name: Requests Per Second
        unit: requests/second
        limit: 30
      - name: Requests Per Minute
        unit: requests/minute
        limit: 300
      - name: Requests Per Day
        unit: requests/day
        limit: 60000
        note: Doubled from 30,000 on 2026-03-03
    endpoints:
      - pattern: /v2/end-users/{end_user_id}
        methods: [GET, PUT, PATCH, DELETE]

  - name: Conversations API
    limits:
      - name: Requests Per Second
        unit: requests/second
        limit: 30
      - name: Requests Per Minute
        unit: requests/minute
        limit: 300
      - name: Requests Per Day
        unit: requests/day
        limit: 60000
        note: Doubled from 30,000 on 2026-03-03
    endpoints:
      - pattern: /v2/conversations/
        methods: [POST]
      - pattern: /v2/conversations/{conversation_id}/
        methods: [GET]
      - pattern: /v2/conversations/{conversation_id}/end/
        methods: [POST]

  - name: Data Export API
    limits:
      - name: Requests Per Second Per Endpoint
        unit: requests/second
        limit: 10
      - name: Requests Per Day
        unit: requests/day
        limit: 15000