generated: '2026-08-26'
method: searched
source: https://www.aifordatabase.com/docs/api-overview/
sources:
- https://www.aifordatabase.com/docs/api-overview/
- https://www.aifordatabase.com/docs/api-errors-and-retries/
api: AI for Database API
limit_count: 6
enforcement_scope: per-organization
window: 1 minute
rate_limits:
- scope: per-account
  plan: Free
  class: general requests
  limit: 60
  window: 1m
  burst: null
- scope: per-account
  plan: Free
  class: chat requests
  limit: 20
  window: 1m
  burst: null
- scope: per-account
  plan: Pro / Max
  class: general requests
  limit: 300
  window: 1m
  burst: null
- scope: per-account
  plan: Pro / Max
  class: chat requests
  limit: 100
  window: 1m
  burst: null
- scope: per-account
  plan: Enterprise
  class: general requests
  limit: 1000
  window: 1m
  burst: null
- scope: per-account
  plan: Enterprise
  class: chat requests
  limit: 500
  window: 1m
  burst: null
exhaustion:
  status: 429
  error_code: RATE_LIMITED
  body: '{"data": null, "error": {"code": "RATE_LIMITED", "message": "..."}, "meta": {...}}'
  guidance: Retry with exponential backoff and jitter.
response_headers:
  documented: []
  observed: []
  retry_after: false
  evidence: >-
    Live unauthenticated GET https://app.aifordatabase.com/api/v1/connections on
    2026-08-26 returned HTTP 401 with headers age, cache-control, content-type, date,
    server, strict-transport-security, vary, x-matched-path, x-vercel-cache,
    x-vercel-id — no X-RateLimit-*, no RateLimit-*, no Retry-After.
gap: >-
  This is the runtime-signal gap for this provider: the numbers are published clearly
  in the docs, but no response header carries limit / remaining / reset, so an agent
  can only discover exhaustion by being 429'd. Adding RFC 9331 RateLimit-* headers
  (and Retry-After on 429) would close it.
other_limits:
- limit: 'Direct SQL query results cap at 500 rows'
  source: 'aifordatabase-mcp run_query tool description (npm aifordatabase-mcp@0.1.2)'
- limit: 'Workflow drafts: 1-5 query steps, max 10 delivery actions, max 20 headers per webhook action'
  source: openapi/aifordatabase-openapi.yml (WorkflowUpdate, WebhookActionConfig)
- limit: 'Free plan: up to 3 workflows'
  source: https://www.aifordatabase.com/pricing/
- limit: 'AI credit budget per billing period, readable via GET /usage/budget (402 on exhaustion)'
  source: openapi/aifordatabase-openapi.yml