Back to documentation

API Reference

Rate Limits

Plan quotas, token caps, and API throttling.

Plan quotas

  • Starter — 100K AI tokens/mo, 1 server
  • Professional — 1M tokens/mo, 5 servers, full agent network
  • Enterprise — unlimited tokens and servers (custom)

Usage gates

Every AI request passes through UsageService.checkBeforeRequest(). Requests are blocked at 100% quota with an upgrade prompt. Alerts fire at 80% and 90%.

Per-server caps

Professional plans enforce a per-server token cap (40% of plan quota) so one guild cannot exhaust your entire allowance.

API rate limits

REST endpoints are rate-limited per user and IP. Burst limits apply to PRIME commands and automation triggers. Retry after the Retry-After header on 429 responses.