aid: robin-ai
name: Robin AI Rate Limits
specificationVersion: API Commons Rate Limits 0.1
description: >-
  Robin AI's public API documentation enumerates HTTP 429 Too Many Requests
  as a documented error response on Documents and Tables endpoints, but does
  not publish concrete per-second, per-minute, or per-day numerical
  thresholds. Limits are negotiated per enterprise contract. Cursor
  pagination caps at 1000 items per page across all list endpoints.
source: https://robinai.com/robin-api
updated: '2026-05-24'
scope:
  api: robin-legal-intelligence-platform-api
  baseURL: https://api.robinai.com
  authentication: X-API-Key
policies:
  - id: too-many-requests
    type: requestRate
    enforced: true
    httpStatus: 429
    description: >-
      Documented in the OpenAPI spec as a possible response on POST /v1/documents
      and other endpoints. Exact thresholds (requests per second / minute) are
      not published; clients should implement exponential backoff on 429.
    sources:
      - https://robinai.com/robin-api
  - id: pagination-page-size
    type: pagination
    enforced: true
    description: List endpoints (tables, documents, templates, groups, table results) accept `limit` between 1 and 1000 with a default of 100, paged via the `starting_after` cursor.
    parameters:
      limit:
        min: 1
        max: 1000
        default: 100
      cursor: starting_after
    sources:
      - https://robinai.com/robin-api
  - id: payment-required
    type: quota
    enforced: true
    httpStatus: 402
    description: >-
      Documented 402 Payment Required response indicates account-level usage
      cap or billing issue has been reached. Specific quota numbers are not
      published.
    sources:
      - https://robinai.com/robin-api
notes:
  - Robin's surface is enterprise-only; concrete per-account limits are negotiated and provisioned at sales time.
  - 429 backoff is standard; clients should respect any Retry-After header if present.