For the complete documentation index, see llms.txt. This page is also available as Markdown.

Rate Limits

Outline:

  • Per-key request budgets (separate for read and write).

  • Headers returned on every response (X-RateLimit-Limit, X-RateLimit-Remaining, X-RateLimit-Reset).

  • Behavior on overflow (429 RATE_LIMITED).

  • Recommended retry strategy: exponential backoff with jitter.

  • Bursts vs. sustained throughput.

Specific limits will be published once the production rollout is complete.

Last updated