generated: '2026-07-19'
method: searched
source: https://docs.getfrontline.ai/docs/rate-limits.md
scope: per-account
note: >-
  Requests are rate-limited per account (a fixed 60-second bucket that resets on the first request
  of each window). Requests from any key under the same account share the bucket. Unauthenticated
  requests are limited per IP at the same threshold.
rate_limits:
- name: Per account
  limit_count: 1000
  window_seconds: 60
  window: 60s fixed bucket
headers:
- name: X-RateLimit-Limit
  meaning: Total requests allowed per window (currently 1000).
- name: X-RateLimit-Remaining
  meaning: Requests remaining in the current window; drops to 0 at the limit.
over_limit:
  status: 429
  code: internal_error
  message: Rate limit exceeded. Please try again later.
  guidance: Wait until the next 60-second window and retry; use exponential backoff.
best_practices:
- Cache reference data (/ai-models, /billing) instead of polling.
- Use the highest pageSize you can tolerate on list calls (default 20).
- Run scheduled syncs off-peak.
- Contact support for a higher limit for legitimate integrations.